Computers and Quantity Surveyors (PDFDrive) PDF

Computers and Quantity Surveyors
Macmillan Building and Surveying Series

Series Editor: lvor H. Seeley
Emeritus Professor, Trent Polytechnic
Advanced Building Measurement, second edition, lvor H. Seeley
An Introduction to Building Services Christopher A. Howard
Applied Valuation Diane Butler
Asset Valuation Michael Rayner
Building Economics, third edition lvor H. Seeley
Building Maintenance, second edition lvor H. Seeley
Building Quantities Explained, fourth edition lvor H. Seeley
Building Surveys, Reports and Dilapidations lvor H. Seeley
Building Technology, third edition lvor H. Seeley
Civil Engineering Contract Administration and Controllvor H. Seeley
Civil Engineering Quantities, fourth edition lvor H. Seeley
Civil Engineering Specification, second edition lvor H. Seeley
Computers and Quantity Surveyors A. J. Smith
Contract Planning and Contract Procedures B. Cooke
Contract Planning Case Studies B. Cooke
Environmental Science in Building, second edition R. McMullan
Introduction to Valuation D. Richmond
Principals of Property Investment and PricingW. D. Fraser
Quantity Surveying Practice lvor H. Seeley
Structural Detailing P. Newton
Urban Land Economics and Public Policy, fourth edition P. N. Balch in, J.
L. Kieve and G. H. Bull
1980 )CT Standard Form of Building Contract, second edition R. F. Fellows
Series Standing Order

If you would like to receive future titles in this series as they are
published, you can make use of our standing order facility. To place a
standing order please contact your bookseller or, in case of difficulty,
write to us at the address below with your name and address and the
name of the series. Please state with which title you wish to begin your
standing order. (If you live outside the United Kingdom we may not
have the rights for your area, in which case we will forward your order
to the publisher concerned.)
Customer Services Department, Macmillan Distribution Ltd

Houndmills, Basingstoke, Hampshire, RG21 2XS, England.
Computers and
Quantity Surveyors
A.J.Smith
Department of Surveying
Trent Polytechnic, Nottingham
M
MACMILLAN
© A. J. Smith 1989
All rl.ghts reserved. No reproduction, copy or transmission

of this publication may be made without written permission.
No paragraph of this publication may be reproduced, copied

or transmitted sa.ve with written permission or in accordance
with the proviSions of the Copyrl.ght Act 1956 (a.s amended),
or under the terms of a.ny licence permitting limited copying
issued by the Copyrl.ght Licensing Agency, 33-4 Alfred Pla.ce
London WC 1E 7DP.
Any person who does a.ny unauthorised a.ct in relation to

this publication may be liable to criminal prosecution and
clvll claims for damages.
First published 1989
Published by
MACMILLAN EDUCATION LTD
Houndmllls, Basingstoke, Hampshire RG21 2XS
and London
Companies and representatives
throughout the world
La.serset by
Panting-Green Publishing Services, London
British Library Cataloguing in Publication Data
Smith A. J.
Computel'S and Quantity Surveyors
1. Great Britain. Quantity surveying.
Appllcations of micro computer systems
I. Title.
624.1'042'0285416
ISBN 978-0-333-49161-4 ISBN 978-1-349-10857-2 (eBook)

DOI 10.1007/978-1-349-10857-2
Cover photograph of a dlgitiser kindly supplled by Techsonlx (U.K.) Ltd.

For Christine, Joanna and Elaine
Contents
Preface ix
Acknowledgements xi
1 GENERAL CONCEPTS 1
1 The Quantity Surveyor and the Computer- a Short History 3
2 Computer Hardware 17
3 The Selection and Management of an Office Computer System 42
4 Computers and the Law 65
5 General Purpose Software 80
6 Public Information Systems 93
7 The Development of In-house Systems 101
8 'Expert systems' -the Future? 115
2 COMPUTERS APPLIED TO QUANTITY SURVEYING 133

9 Computers in Cost Planning 135
10 Computers and Bills of Quantities 144
11 Computers for Post Contract Work 172
12 Quantity Surveyors and Computer Aided Design 183
13 Quantity Surveyors, Computers and Project Management 195
Index 203
It is difficult not to believe that the computer will become the
medium through which increasingly large parts of design and
construction will be done in the future; taking over from the
manual process just as inexorably as steam power and
machines took over from manual labour in the industrial revolu-
tion.
John Chalmers
'Computing for construction in the 'eighties'
Chartered Quantity Surveyor May 1 979
Preface
The quantity surveyor and the computer have been partners in an intense
love-hate relationship for over 30 years; indeed, it would be difficult to
find any other profession in which the relationship between man and
machine has generated, by turns, so much discussion; hope and optimism;
disappointment and disillusion. Even now the rank and file of the profession
is still sharply divided between those who claim 'computer literacy' and
who consequently venerate the computer and all its works and those who,
with apparently equal pride, proclaim themselves to be 'techno-peasants'.
It is not the intention of this book to attempt to persuade students or
fellow practitioners to follow either view. The intention is to attempt to
inform rather than to influence; to try to blow away some of the mystique
which still surrounds the use of computer systems in order that, no matter
which side of the computing fence one eventually elects to stand, at least
the decision is made on some rational basis.
This book is therefore divided into two parts, the first part (chapters 1 to
8) dealing with general concepts of information technology, and the latter
part (chapters 9 to 12) dealing with specific quantity surveying issues. In
many ways this division is, of course, artificial as indeed is the subdivision
in part 2 between the various aspects of quantity surveying, and the notion
that the quantity surveyor stands alone from the rest of the construction
team. I also recognise that there are some areas of work in which quantity
surveyors are employed which are not specifically considered here; none-
theless, the first part of the book should still provide a useful introduction
to the technology.
The book is directed primarily towards providing students with a basic
grounding in the use of computers by quantity surveyors. It covers a wide
range of subjects, and I am only too aware that some areas are treated in a
very superficial way; this is partly deliberate and partly imposed by pressures
of space, but I hope that the references given will aid the research of those
wishing to pursue particular aspects of the subject to a greater depth. I
hope that the book may also be of interest to those quantity surveying
practitioners who have yet to come to terms with the technology, and also
to students and practitioners of other construction disciplines.
A.J. Smith
Acknowledgements
I am indebted to Brian George, Senior Lecturer in the Department of

Building and Environmental Health, Trent Polytechnic for permission to
use extracts from his paper 'Project management and micro computers',
first delivered at the 'Quantity Surveyors and Computers' conference held
at the Barbican in April 1988, in the preparation of chapter 13, and for his
most helpful comments on the draft.
I am also most grateful to Rob Bradgate of the Department of Legal
Studies, Trent Polytechnic for his advice on the legal issues discussed in
chapter 4.
May I also freely acknowledge with much gratitude all of the help and
co-operation given by the many friends and colleagues too numerous to
mention with whom I have discussed ideas and argued principles over the
past twenty years; many of the ideas we discussed have found their way in
one form or another into this book. In particular, thanks are due to my
colleagues from whom I have learned a great deal, especially Dick Massey,
Phil Hawkins, Graham Williams and Dave Rowe at Trent Polytechnic,
Vernon Foster and lvor Neal of Derbyshire County Council and Wally
Harlow, former Chief Quantity Surveyor at Derbyshire County Council.
Thanks are also due to Professor lvor Seeley for his most helpful comments
and to jaqui Hunt for the illustrations.
Finally my grateful thanks to my family for all of the tolerance, help and
encouragement given whilst this book was in preparation.
General Concepts
l The Quantity Surveyor
and the Computer- a Short
History
Objectives
At the end of this chapter you should be able to:

• Outline the history of commercial data processing.
• Outline the major developments which have taken place in the use of
computers by quantity surveyors during the past 25 years.
• Outline the ways in which quantity surveyors are most likely to use
computers today.
Early developments in computing and data processing
The first consideration when discussing the development of early computer

systems is to decide exactly where to begin. Most histories of computing,
for example Evans (1981 ), begin with the development of mechanical
counting machines by people such as Pascal and Babbage, but whilst these
machines have much to do with the use of computing 'engines' to solve
mathematical problems they have little or nothing to do with either the
development of electronic computers or the use of such machines to
process data.
The first electronic, as opposed to electro-mechanical, computers are
now generally acknowledged to have been the 'Colossus' machines first
described in Michie (1973), which were built by the Post Office Research
Station for the Department of Communications of the British Foreign
Office, the first of which was delivered to the Government Code and
Cypher School, Bletchley Park in December 1943. A fuller account of
Colossus following the release of further information by the Ministry of
Defence in 1975 is given in Randell (1980).
Most of the developments in computers at this time were, of course,
highly secret and it was not until after the end of the Second World War
that the use of computers for commercial data processing was considered.
Considerable use had however been made of the electro-mechanical
punched card sorting equipment, developed around the turn of the century
by Hollerith and his contemporaries, for accounting, processing of census
4 Computers and Quantity Surveyors
data and the like. By the end of the second world war the use of such
machines was commonplace and companies such as IBM were already
well established in the data processing field. The origins of data processing
can therefore be said to pre-date the development of the electronic computer
by almost half a century.
It was then against this background of data processing experience and in
the spirit of new optimism just after the war that J. Lyons & Co. began, in
1947, to investigate the use of computers for commercial data processing
and eventually, in consultation with computer specialists from the com-
puting laboratory at Cambridge University, installed LEO, the Lyons
Electronic Office. LEO finally became operational towards the end of
1953, although Lavington (1980:p. 35) points out that LEO was really a re-
engineered version of the EDSAC (Electronic Delay Storage Automatic
Calculator) originally developed in Cambridge in 1949.
Although LEO was the first computer in the world to be used for
commercial data processing, Lavington (1980:p. 41) claims that the world's
first commercially available computer was the Ferranti Mk.1 based on the
Manchester Automatic Digital Machine (MADM) developed at Manchester
University, the first of which was delivered in February 1951. Evans
(1980:p. 92) however claims that the first commercially successful computer
was the UNIVAC 1 the first of which was delivered in March 1951.
It should be pointed out that by no means everyone viewed the future of
the computer with the optimism shown by Lyons. Fisher, McKie and
Mancke (1983:p. 14) quote Cuthbert Hurd, then IBM Head of Applied
Science, as stating that he was told, in 1949, by IBM president Thomas
Watson Snr. that the single IBM Selective Sequence Electronic Calculator
then in existence could solve all of the important scientific problems in the
world requiring calculation. Hurd is also reported as saying that a sub-
stantial body of opinion within IBM at the time:
... could not imagine that enough problems or applications could ever be
proposed by IBM's potential customers to keep a computer busy because
such machines were to have the capability of processing several thousand
operations per second.
The quantity surveyor and data processing
In the beginning...
The origins ofthe quantity surveyor's love-hate relationship with computing

can be traced back to the mid 1950s, when an article, Stafford (1957), was
published in The Builder magazine. In the article the author, a practising
quantity surveyor, advocated the use of punched card sorting equipment
The Quantity Surveyor and the Computer- a Short History 5
for the abstracting stage of the bill production process. The true origins
may in fact be even earlier, since Nott (1967) claims that discussions
began in the Architects Department of Hertfordshire County Council as
early as 1955/56.
Stafford's ideas appear to have been first put into practice by quantity
surveyors in the CLASP group attached to Nottinghamshire County Council
working in conjunction with Derby and Coventry City Councils, and the
first successful implementation of a punched card sorting system for this
kind of work was reported in The Builder in August 1958 (Cooke (1958)).
Eighteen months later, in March 1960, Braine and Pritchard (1960)
described another system, again using punched card sorting equipment,
this time developed within the Quantity Surveying Division of the War
Office. It is interesting to note that the publication of Braine and Pritchard's
article provoked a rather acerbic response from the respected quantity
surveyor and author A.J. Willis, Willis (1960), who questioned whether the
reporting of such fringe activities was not a waste of valuable magazine
space.
Willis's reservations were not however shared by all, and it was later
reported in the Chartered Surveyor (RICS (1961 a)) that:
In September 1959 the Quantity Surveyors Committee endorsed the
appointment of a Working Party to consider and report on the possible
place of computers in quantity surveying practice. A further Working
Party was formed in February 1960 to deal with one facet of this
investigation namely 'to look into the possibilities of the use of computers
in connection with the working up of quantity surveyors dimensions'
and to report to the main working party.
The final report of this latter working party, RICS (1961 b), was presented at
a quantity surveyors general meeting held on the 18th October 1961, and
the ideas presented, together with the views expressed in the interim
report, RICS (1961 a), had a profound influence on the development of
computer systems for quantity surveyors for the following twenty years.
Among the main themes were:
i) Bill production by computer was not only possible but relatively easy:
There are unlikely to be any insuperable problems to the use of

computers for the working up of quantity surveyors' dimensions for
the general run of building and civil engineering work
(RICS (1 961 a))
ii) an industry-wide coded standard library of descriptions was essential:
Standard descriptions are inevitable ... Standardisation of descrip-

tions ... would, we feel, help the industry to build more cheaply
(RICS (1 961 a))
We therefore recommend that the RICS should ... consider the produc-
tion of a standard coded library...
(RICS (1961 b))
iii) one result of the use of computers would probably be a co-ordinated
approach to project information and planning:
The building and civil engineering industries ... are being urged, and
are themselves striving to improve their planning and construction
methods... the adoption by the quantity surveying profession of data-
processing techniques might well be the catalyst for speeding up the
desired changes.
(RICS (1961 b))
iv) the re-use of the quantity surveyor's data by the contractor was seen as
a good idea:
Using the data which produced the bill of quantities to provide data
for:
1) planning constructional operations
2) ordering materials
3) forecasting labour requirements
4) bonusing
(RICS (1961 b))
v) computer processing of bi lis of quantity was to reduce the cost of
building:
We forsee the adoption of integrated data processing techniques
speeding construction and lowering costs.
(RICS (1961 b))
With hindsight one could perhaps query the validity of all of the above.
So far as the first was concerned development of a comprehensive,
working and marketable system as opposed to a test or demonstration
system proved, for most developers, to be anything but easy. As an
example consider the system developed by the Development Group of
Chartered Quantity Surveyors, a consortium formed of twelve of the
leading quantity surveying practices in the United Kingdom, who began
work on a billing system before mid 1966 (see RICS (1967a)) with the
intention of having a system operational by the end of that year. RICS
(1967a) however admits that anticipated completion of a basic system
would be delayed by about a year, and the complete system would not
be ready before 1968. By November 1972 an acceptable commercial
system was still not generally available, and general release was not
anticipated until mid 1973. (Development Group of Chartered Quantity
Surveyors (1972)). Whilst part of the delay can undoubtedly be attributed
to the metrication of the Standard Method of Measurement in 1968, it is
The Quantity Surveyor and the Computer - a Short History 7
clear that the early programs were not satisfactory and that the difficulties
of development were severely underestimated.
It could also be argued that the preoccupation with coding of the data
has restricted rather than advanced the march of progress. The impact of
coding was airily dismissed in the initial reports, RICS (1961 b) for example
said:
In the course of time however... the taker-off may write just brief
descriptions plus the code number. Coding may then actually speed
up taking off rather than delay it.
Dent (1964:p99) went even further:
Once it is decided that a modification of existing taking off procedure is
acceptable then the obvious person to do the coding is the taker-off
himself. The pattern which will eventually emerge is probably one in
which conventional descriptions disappear altogether from the dimen-
sions, and codes will take their place.
There is little doubt that the readiness of those quantity surveyors
involved in systems development to accept the necessity of coding, and
their occasionally ingenious attempts to find economic ways to do it,
allowed this archaic, error prone, boring and unpopular chorP. to be
tolerated in quantity surveying systems long after alternative techniques
became available.
The hope that the adoption of data processing techniques by the
quantity surveyor would lead to better co-ordinated project documentation
and planning might have appeared to be convincing at the time in the
light of contemporary developments in information science, but in the
cold light of day it is difficult to believe that quantity surveyors seriously
expected the architectural and engineering professions to change their
ways simply because the quantity surveyor wanted to use a computer.
Perhaps the true reason why so much time, money and effort was
expended on billing systems is none of the altruistic ones given in the
working party reports. Perhaps the true and overriding reason is set out
in the first few paragraphs of RICS (1961 a) and is rather more selfish:
During the past 15 years these traditional methods (of bill production)
have, however, suffered for the following reasons:
(1) the raised status of the quantity surveying profession has attracted
entrants who, in the main, aspire to taking off as quickly as possible
and who are not content to treat working up as a career. This has
created a chronic staff shortage of workers up and
(2) the high levels of salaries payable to juniors by reason of full
employment and the competition from industry and commerce for their
services.
B Computers and Quantity Surveyors
Quantity surveyors find that the above factors cause working up to

absorb an inordinate proportion of fees.
1960-1980- Triumph or disaster?
Following the publication of the RICS working party reports in 1961, a

number of practices and organisations began work on the development of
billing systems and the technical journals of the 1960s were filled with
articles describing the development of one new system after another. In
1967 the RICS hosted a conference on computer techniques, and by this
time the aim of the RICS Computer Applications Steering Committee had
become much narrower than the brief given to the original 1959 group. It
was now expressed as:
... to encourage the development of systems which by the use of the
computer are able to produce quickly and economically Bills of
Quantities to a national standard ...
(RICS (1967b):Foreword)
The conference brought together most of the pioneers of computerised
working up and the publication of the papers presented, RICS (1967b),
which includes a verbatim transcript of the discussions and an excellent
bibliography, provides a comprehensive survey of the computer tech-
niques in use at the time. Overall the tone is one of optimism although
Fletcher (1967) sounds somewhat disappointed with progress when he
says, of the profession's response to the 1961 report:
The profession generally remains largely uninterested and gave no
encouragement to their executive; but several groups of surveyors con-
tinued their isolated efforts and others, inspired by the working party report,
began exploratory work oftheir own. However, the essential spark of unity
was lost and inevitably the developments in this field followed divergent
paths, as a consequence of which none has attained, or even promises
to attain, the national significance that might well have been achieved
by corporate effort.
Some guide to the general degree of interest in the subject at the time of
the 1967 conference can be gained from an analysis of the conference
delegates, divided by types of employment:
Local authority 74
Central government 15
Private practice 35
Computer manufacturers 3
Others (institutions etc.) 11
138
Extracted from RICS (1967b):appendix C

Of the 138 delegates present at the conference over half (53.62 per cent)
came from local authorities and almost two thirds (64.49 per cent) from
either local or central government. Only fractionally over one quarter of
the delegates (25.36 per cent) came from private practice representing 22
firms. Delegates were also asked to complete a questionnaire, part of which
was concerned with their use of computer systems. Of the 45 completed
questionnaires listed only 20 reported that systems were actually in use
(RICS(1967b):appendix A). This high level of central and local government
dominance is to some extent confirmed by a collection of case studies of
Bills of Quantities systems published in 1969 by the Ministry of Public
Buildings and Works (MPBW (1969)); of the 16 case studies included 10
were of systems developed in either local or central government.
In 1973 the RICS again hosted a series of three conferences on computer
techniques, the first of which, on 9th January, dealt with billing systems.
Although many of the systems presented at that conference were simply
updates of those shown in 1967, Smart (1973) felt able to comment:
These years since 1961 have seen a growing awareness and acceptance
of computer techniques throughout the profession ...
Despite this optimism however a survey by the National Consultative
Committee of the Building and Civil Engineering Industries conducted
some five years later, NJCC (1978:p. 17), reported that:
Despite the lead given by the RICS in 1960, the adoption of these
systems has been slow.
and later (at long, long last!):
It is possible that the original enthusiasm for 'integrated systems' obscured
the significant benefits which can flow from more limited applications.
We consider the potential benefits are such that research and develop-
ment should be initiated to provide improved computer aids for other
aspects of the quantity surveyor's role ...
How then should this period be judged? Had the quantity surveying
profession as a whole really begun to enthusiastically embrace automated
techniques as claimed by Smart, or was the profession in general still as
uninterested and ambivalent as Fletcher (1967) had imagined and NJCC
(1978) appeared to confirm?
On the one hand it can clearly be seen that a few very successful
systems were developed. The ideas first proposed in Monk and Dunstone
(1965) for example which were later marketed by Oldacres Computers Ltd.
have clearly stood the test of time, and the system marketed until the end
of the 1970s by LAMSAC, which was itself based on earlier work by
Hertfordshire County Council and CLASP (described in Harlow, Chapman
and Swanston (1972)) was very successful and achieved widespread use,
being installed in some 60 organisations worldwide. Of the other systems
the MPBW 'Enviro' system, first described in Crisp (1967), was a brave
attempt to eliminate the separate coding process and the Development
Group system mentioned earlier, although never truly successful, embodi~d
at least the philosophy of automatic generation of some quantities from
'key' information thus attempting to influence the taking off process.
On the other hand, however, it appears that, apart from a few large
practices mainly based in London, the lack of easy access to computers
restricted the main thrust of development to local and central government
organisations. This process was fuelled as much by the twin enthusiasms of
a fledgeling computer industry eager to sell its wares and local authority
treasurers who were eager to buy them, as it was by the well thought out
plans of quantity surveyors. Looking back on events then it does seem that
in many authorities the computer was used because it was there; to take
advantage of spare time; to ensure full usage of a very expensive piece of
equipment. Of course once the machine was used to full capacity the
treasurer had a very good argument for expanding the facilities to cope
with the inevitable peaks which result from the full utilisation of any
resource. The entire process then becomes a spiral, and the true cost of
using the system for any particular task becomes impossible to calculate
and often irrelevant.
So far as the computers themselves were concerned the machinery
available had evolved from devices such as the Elliot 803B, which had 4K
of main storage, four tape decks and paper tape input and output as used
by Northamptonshire County Council in 1967 (see RICS (1967b):appendix
A), to machines like the IBM 370/3031 in use in Derbyshire County
Council in 1980, which had 3000K of main storage and 9518MB of disk
backing storage serving 56 video display terminals, described in Widdeson
(1980). Despite the dramatic increase in available computer power however
the basic techniques used by quantity surveyors to solve their problems
had, with the addition of a few relatively minor enhancements, remained
virtually unchanged since the early 1960s. Truly innovative research and
development had effectively come to an end by the mid 1960s and was
not to re-emerge until the advent of the microcomputer at the end of the
1970s.
Revolution! The introduction of the microcomputer
Although the principles underlying the development of the modern micro-

computer were first discovered during the early 1970s, it was not until the
later years of the decade that working systems became available to the
general public in the United Kingdom. When the equipment did begin to
appear however it was launched with an intensive advertising campaign
designed to show how friendly the new machines were, even the names
were chosen with the greatest care, 'Apple' and 'PET' fm example. The
publicity was very effective and soon microcomputers were tipped to
become the biggest growth area of the century. Trimmer (1979) was able to
report:
The resurgence of interest in computing continues. The one day seminar
run by the QS (research and development) committee at the institution
on the 28th. March was a sell-out...
and in a paper presented at that seminar Chalmers (1979) said:
It is difficult not to believe that the computer will become the medium
through which increasingly large parts of design and construction will be
done in the future; taking over from the manual process just as inexorably
as steam power and machines took over from manual labour in the
industrial revolution.
The resurgence of interest reported by Trimmer is reflected in a number of
events organised by the RICS and others, all concerned with the new low
cost approach to computing. Unfortunately, although the computers were
readily available, reliable advice from the suppliers was, in general, lacking,
and exasperation with a network of computer distributors who seemed to
know nothing of the possibilities of the computer they were trying to sell,
beyond standard accounting and word processing packages, led to demands
for unbiased advice from quantity surveyors eager to learn about the new
technology. Organisations such as the Construction Industry Computing
Association (CICA) provided (and still do provide) a valuable service. It is
however conceivable that the new enthusiasm did not extend to actually
buying any equipment; Millen wrote, in 1978:
Quantity surveyors, as any word processor salesman will tell you in an

honest moment, have not been the most successful customers of the new
'space age' generation of office equipment. In fact, as far as concrete
sales are concerned, they have been a relative failure.
(Millen (1978))
Whatever the true picture, in response to the quantity surveyors' pleas, the
RICS, in 1981, commissioned CICA firstly to undertake a study of the use
of microcomputers in the profession, and secondly to prepare a handbook
which would act both as a guide to the implementation of microcomputers
and would also highlight some of the pitfalls into which the unprepared
could fall. The completed work, CICA (1981 ), was published in 1981, and
much of the contents are still valid. This publication was revolutionary in
that, for the very first time in a 'text book' about computing for surveyors,
the authors concentrated on the quantity surveyor as a user and manager
of the machine rather than attempting to teach computing. Whilst other
commentators (e.g. Dent (1964), Alvey (1976) and Brandon and Moore
(1983)) wrote volumes about how computers worked and how to program
them, CICA (1981) relegated a brief description of their workings to an
appendix and stated, quite firmly, that relying on quantity surveyors them-
selves to program operational systems was not, generally speaking, a good
idea (CICA (1981 ):p. 16).
The same period began to see developments in the use of the computer
for other quantity surveying activities. Links between bill production
systems and computer aided design systems were perceived to be an
important area, and systems are described in Daniel (1979) and Hawkins
and Davidson (1979). It is however evident that the pace of development
in this direction was painfully slow, and it was not until 1987 that the
profession as a whole made an effort to confront the issue, when the RICS
commissioned Reading University to undertake a study of computer aided
design systems and the quantity surveyor. The results of the study have
been published as RICS (1987). In an article describing the research Edgill
and Atkin (1987) write:
Before the project commenced in earnest it was evident that many
misconceptions were held by QS's. Whatever those amounted to it was
clear that CAD was not well understood.
The use of computers in other areas however has apparently proceeded
apace. In the post contract field the introduction of the NEDO formula for
calculation of increased costs provided a fertile area for development and
many practices developed their own programs. Some were made available
for wider use and a typical example is described in Howard (1979),
although many practitioners have since accomplished the same thing using
a spreadsheet rather than by writing programs. Much interest has also been
expressed in the use of computers for cost modelling, and the 'state of the
art' in this area was explored at a conference held at Salford University in
1987. The papers presented, published in Brandon (1987), reveal the
enormously wide range of research and development in this field.
In the area of bill production, although some of the old mainframe
packages still remain, converted of course to run on microcomputers, the
era of the micro has seen major developments. Hunt (1982) and Hunt
(1983) for example both review systems using digitisers for measurement,
whilst Hunt (1984a) reviews a revolutionary bill production package the
operation of which implies fundamental changes to traditional measure-
ment techniques. The use of computers to access and process data from
remote data bases has become commonplace, and the BCIS On-line
service described in Robertson (1983) is now well established.
In the realm of research the profession has once again provided a lead to
the industry by securing the only construction industry project to be
funded from the Government's Alvey programme of research into fifth
generation computer systems. The project, which is concerned with the
use of knowledge based 'expert' computer systems in quantity surveying is
described in Brown (1987), and further consideration of expert systems is
included in chapter 8.
How then can we sum up the reaction of the quantity surveying profession
to the use of computers in the age of the micro? On the one hand it is
obvious from the foregoing that some practitioners have made enormous
strides in exploiting the power of the machine. On the other hand it would
appear that, after almost 30 years, by no means all quantity surveyors are
convinced of the benefits of computerisation, although one must admit that
it has proved very difficult to establish just what the level of usage really
is. Hunt (1984b) reports that the number of quantity surveying firms using
computers rose from between 3 and 5 per cent in 1982 to 36 per cent by
the spring of 1984; a year later the figure was quoted at 50 per cent (Hunt
(1985)), and by the summer of 1986 useage was reported to be 'well in
excess of 50 per cent and increasing daily' (Hunt(1986)). Lowe (1986) also
quotes a figure of 'circa 50 per cent', although he goes on to say that
'many practitioners remain hostile to computerisation'. Lowe also believes
that increased computer literacy from the new generation of surveyors is
essential for the profession to prosper. All of this seems, at first sight, to
point to a profession in which, although there may still be a few doubters,
at least a majority have come to terms with the machine and yet, in late
1987, Hunt writes:
When I was asked recently how quantity surveyors had progressed in the
use of computers over the past decade or so, I was tempted to say 'very
little' but that would not be quite true neither would it be the whole
story. The acid test of course is determined by the proportion of the
profession actually using computers effectively and my current guestimate

is about 10-15 per cent. There are various figures produced by the RICS
and CICA ranging up to 50 per cent computer useage, but these figures
relate to the number of people who ' ... have access to a computer ... '
rather then those that actually use a computer for more than ten minutes
each day. The percentage also changes dramatically if word processing
is not included.
(Hunt (1987))
It also appears that this malaise is not restricted to the quantity surveyor but
is industry wide. Barton (1987) asks whether the whole construction industry
is still illiterate and ambivalent as far as computers are concerned.
So what is the real truth? Is it necessary, as the proponents of the
computer have contended for 30 years, that the quantity surveyor should
take the technology on board if he is to survive? Will the use of intelligent
computer aided design systems by architects and designers offer a great
opportunity for the quantity surveyor to transcend the routine work which
presently supports a large proportion of the profession, or will it simply
make the quantity surveyor as unnecessary as the typewriter made the
scribe? Will the effect of the current expert systems research be to enhance
and enrich the quantity surveyor's role by making the expert's knowledge
more widely available throughout the profession, thus improving the quality
of the service the individual practitioner can provide to his client; or will
the ready availability of such systems outside the profession result in the
quantity surveyor shooting himself in the foot with the press of a computer
key?
Time, of course, alone will tell, but history shows that on this issue the
pundits have been more often wrong than right!
References
Alvey R.j. (1976) Computers in quantity surveying Macmillan

Barton P. (1987) 'Computers, construction and apathy' Construction Computing Winter 1987
p.40
Braine D.G. and Pritchard D.G. (1960) 'Billing mechanised' The Builder
18.3.60 pp. 567-573
Brandon P.S. (ed.) (1987) Building cost modelling and computers E.&F .N. Spon
Brandon P.S. and Moore R.G. (1983) Micro computers in building appraisal Granada
Brown G. (1987) 'RICS/Aivey expert systems research' Chartered Quantity Surveyor April
1987 pp. 15-16
Chalmers j. (1979) 'Computing for construction in the 'eighties' Chartered Quantity Surveyor
May 1979 pp.131-133
CICA (1981) Chartered quantity surveyors and the microcomputer RICS
Cooke j.E. (1958) 'Unit quantities and the architect' The Builder 22.8.58 pp. 298-299
Crisp P. (1967) 'MPBW computer system for the production of Bills of Quantities' Computer
Techniques RICS
The Quantity Surveyor and the Computer - a Short History 15
Daniel P. (1979) 'Bills of quantity in a multi-disciplined design system' Chartered Quantity

Surveyor May 1979 pp. 141-142
Dent C. (1964) Quantity surveying by computer Oxford University Press
Development Group of Chartered Quantity Surveyors (1972) The D.C. System paper
presented to the RICS conference on billing systems held 9.1.73
Edgill B. and Atkin B. 'CAD opportunities for the Q.S.' Chartered Quantity Surveyor May
1987 pp. 9-10
Evans C. (1981) The making of the micro Victor Gollancz
Fisher F.M., McKie j.W and Mancke R.B. (1983) IBM and the United States data processing
industry, an economic history Praeger Scientific
Fletcher L. (1967) 'Keystone of construction industry data co-ordination' Computer techniques
RICS pp. 7-14
Harlow W.R., Chapman j.C. and Swanston R. (1972) LAMSAC Computer System paper
presented to the RICS conference on billing systems held 9.1.73
Hawkins M. and Davidson j. (1979) 'Computer aids bills of quantities' Chartered Quantity
SurveyorApril1979 pp. 99-105
Howard R. (1979) 'Formula price adjustment of building contracts: why use computers?'
Chartered Quantity Surveyor April 1979 pp. 105-106
Hunt G. (1982) 'Micronotes: measure for measure' Chartered Quantity Surveyor january 1982
p. 161
Hunt G. (1 983) 'Micronotes: innovation in digitised measurement' Chartered Quantity Surveyor
December 1983 p. 187
Hunt G. (1984a) 'Micronotes: now you can measure your modules' Chartered Quantity
Surveyor january 1984 p. 219
Hunt G. (1984b) 'Micronotes: micros: new developments' Chartered Quantity Surveyor April
1984 p.333
Hunt G. (1985) 'Micronotes: taking stock' Chartered Quantity Surveyor May 1985 p.381
Hunt G. (1986) 'Micronotes: quo vadis computers?' Chartered Quantity Surveyor june 1986
p. 13
Hunt G. (1987) 'Micronotes: making progress' Chartered Quantity Surveyor November 1987
p. 13
Lavington S. (1980) Early British computers Manchester University Press
Lowe j. (1986) 'Problems of pre-conception' Chartered Quantity Surveyor October 1986 p. 21
Michie D. (1973) 'The Bletchley machines' The origins of digital computers: selected papers
Randell B. (ed) Springer-Verlag
Millen D. (1978) 'What in the world is a word processor?' Q.S. Weekly 30.11.78 pp. 6-7
Monk K.W. and Dunstone P.H. (1965) 'Significant development in the techniques of quantity
surveying by computer' Chartered Surveyor February 1965 pp. 420-427
MPBW (1969) The preparation of Bills of Quantity with the aid of computers Ministry of
Public Building and Works Directorate of Research and Information
NjCC (1 978) Computing and communication in the construction industry Department of the
Environment
Nott C.M. (1 967) 'Hertfordshire County Council computerised quantity surveying services'
Computer Techniques RICS
Randell B. (1 980) 'The Colossus' A history of computing in the twentieth century Metropolis
N., Howlett H.j. and Rota G. (eds.) Academic Press
RICS (1961 a) 'The use of computers for working up' Chartered Surveyor April 1961 pp. 561-
563
RICS (1 961 b) 'The use of computers for working up' Chartered Surveyor November 1961 pp.
248-251
RICS (1967a) 'The computer and the Bill of Quantities - towards a new era? Chartered
Surveyor july 1967 pp. 13-17
16
RICS (1967b) Computer Techniques RICS

RICS (1987) CAD techniques: opportunities for chartered quantity surveyors RICS
Robertson D. (1983) 'A new cost data bank' Chartered Quantity Surveyor july 1983 p. 461
Smart D. (1973) Computer aided billing systems - an evaluation paper presented to the RICS
conference on billing systems held 9.1.73
Stafford H.]. (1957) 'The standard draft bill of quantities' The Builder 11.1.57 p. 94
Trimmer A. (1979) 'Quantifying' Chartered Quantity Surveyor May 1979 p. 130
Widdeson P. (1980) Computers in public authorities privately published
Willis A.]. (1960) letter to the Editor The Builder 25.3.60 p. 590
2 Computer Hardware
Objectives

• Explain the terms 'hardware', 'software' and 'firmware'.
• Identify and describe the purpose of the main functional components
of a simple office computer system.
• Discriminate between micro, mini and mainframe computers and
identify their salient characteristics.
• Explain the meaning of the terms 'batch processing', 'interactive pro-
cessing' and 'distributed processing'.
• Identify the main types of input and output device commonly used in
quantity surveying, architecture and construction.
• Explain the meaning of some of the more commonly quoted measures
of computer performance.
• Discuss the major types of backing storage device in common use.
• Outline the main factors affecting communication between computers.
Definition of some basic terms
Computer hardware is the term used to describe the actual computer

machinery itself including all of the attendant peripheral devices such as
keyboards, printers, screens, disk drives etc. It therefore pertains to the
computer machinery as opposed to software, which is the term used to
describe the programs which tell the machinery what to do. It follows then
that the software is intangible, the only physical manifestation of software
being the medium upon which it is stored, for example, a floppy disk or a
cartridge of magnetic tape.
All modern computers are based on what has become colloquially
known as 'chip' technology, a method of micro-miniaturisation which uses
photo-engraving techniques to create complex integrated electronic circuits,
each consisting of many different components, 'printed' onto tiny slices of
some suitable material, usually silicon, called 'chips'. It is the development
of this and similar high technology processes, largely fuelled by projects
such as the American space programme, which has led directly to the wide
range of cheap and very powerful computing equipment now available.
In the main, computer programs are stored on some type of magnetic
medium, for example a disk or tape, and are 'read' into the computer's
memory when required. Once the machine is switched off, however, the
contents ofthe memory are normally lost and the program must be reloaded
from the disk or tape the next time it is required. Some machines use special
chips which can be programmed by a manufacturer or a system supplier in
such a way that the stored program remains intact when the power is
turned off, thus obviating the need to load the program from tape or disk
each time the system is to be used. An example of such a chip might be
that used to control an automatic washing machine. Chips such as these
are therefore part hardware, the chip itself, and part software, the stored
program, and are usually termed 'firmware'. Programmable chips may take
a number of forms, those which have the software permanently written to
them by the manufacturer are often referred to as 'read only memory'
(ROM), whilst those which are supplied unprogrammed by the manu-
facturer and are subsequently programmed by a system supplier are referred
to as 'programmable read only memory' (PROM). A special class of
PROM which can be programmed, subsequently erased, usually by means
of intense ultra-violet light, and re-programmed is known as 'eraseable
programmable read only memory' (EPROM). The capacity of ROM type
chips is usually limited and programs are usually restricted to either simple
sequential routines, such as those used in an automatic washer or a
computer controlling some kind of repetitive process plant, or to contain
'start-up' instructions which then themselves initiate the process of loading
larger programs from disk or tape. This latter type of routine is known as a
'bootstrap loader'.
Computer 'generations'
Brandon and Moore (1983:p5) classify modern microcomputers as sixth

generation machines, but the more usually accepted classification,
stemming from Japanese initiatives announced in 1981, (see for example
Feigenbaum and McCorduck (1983)), would classify such machines as
fourth generation computers. The usually accepted definition of computer
generations is as shown in table 2.1.
Some indication of the speed and complexity of computer development
over the first 40 years can be gained from table 2.2, which compares
EN lAC, a first generation machine with a modern 'lap-top' portable.
Computer Hardware 19
Table 2. 1 Computer generations
Generation Technology Comments
1st (1940-52) Vacuum tubes (valves) large, expensive,

unreliable, slow
2nd (1952-64) Transistors Smaller, faster, needed
air conditioning,
maintenance reduced
but still a problem
3rd (1964-71) Integrated circuits Smaller still, lower
power consumption,
more reliable, cheaper,
faster
4th(1971-) large-scale integrated No air conditioning,
circuits cheap, fast, easy to use
5th (late 1980s/early Currently under
1990s) development
Table 2.2 Comparison of EN lAC with DG/1 lap-top portable
Features EN lAC Data General

(1945) DG/1 (1985)
No. of components > 100,000 valves, <100 chips

resistors etc.
Size Approx. 100 ft 2.8 X 11.7 X 13.7 in
x10ftx3ft
Weight Approx. 100 tons 121b
Power consumption 140 KW 240 Vor
rechargeable battery
Main memory 1 K bits 512 K bytes
Backing storage None Two 3.5 in. disks
max. 715 K
Functional components of a computer system
Having defined in very broad terms what is meant by computer hardware,

let us now consider the main functional blocks which make up a computer
system. All modern computer systems from the smallest micro to the largest
mainframe consist of the same four essential types of equipment; namely
some form of input device to allow the user(s) to communicate with the
machine, some form of output device to allow the computer to com-
municate with the user(s), a calculation and control unit (the computer
itself, sometimes referred to as the central processing unit or CPU), and
some form of storage device to permit long-term storage of data and
programs in machine readable form when the machine is switched off
(often called 'backing storage'). The logical arrangement of the units for a
simple small business system is shown in figure 2.1.
BACKING STORAGE
/
IDISK OR TAPE)
CENTRAL PROCESSING UNIT

ICPUl
INPUT
OUTPUT IKEYBOAROl
IVIDEO SCREE )
HARO COPY OUTPUT

D]
I PRINTER!
Figure 2. 1 Functional components of a simple computer system
In physical terms in the smallest computers all of the functional units may
be contained in one box. In this case the input device will generally be a
keyboard, the output device is likely to be some form of screen or liquid
crystal display, the backing storage is likely to be a built-in disk drive and
the computer itself may all be housed on a single printed circuit board. In
contrast a large computer installation may have a whole network of many
video display units each acting as an input and output device, some
perhaps located many miles away from the computer centre and connected
to the machine via telephone or data transmission cables. In addition a
large computer would be likely to have a number of output devices such
as printers, plotters etc. The central processing unit of a large computer
may well take up several large cabinets and the backing storage will
consist of a number of high capacity disk drives. The central processor and
disk units will usually be housed in a large, specially designed room or
suite and the system will require the services of a number of specialist
operators.
Types of computer system
The range of computer systems currently available may be broken down

into three broad classifications commonly referred to as 'mainframe com-
puters', 'minicomputers' and 'microcomputers'. The dividing lines between
the three groups are very blurred and difficult to define with any precision
at least in terms of function. Some of the larger micros perform the same
function as the smallest minis, and some of the larger minis are used for
the same tasks as the smaller mainframes, but nevertheless the classifica-
tions still continue to be used in practice. The salient features of each type
are described below.
Mainframe computers
Machines at the largest end of the scale. They are physically large
and require purpose-built air-conditioned accommodation and a care-
fully regulated power supply. They will require specialist operators to
control them and will usually be geographically remote from their users.
Since the volume of backing storage is huge, often measured in terms of
thousands of millions of characters, they are very good at information
storage and retrieval and are often used to suppport large tele-processing
networks. Such machines excel at tasks requiring the handling of large
amounts of data. They will therefore be used by organisations such as
banks, large local authorities, central government departments such as the
Driver and Vehicle Licensing Centre, or as the central machine of a large
company.
Minicomputers
Minicomputers are perhaps the hardest group to define since at one end of
the spectrum the larger minis overlap both in appearance and function
with the smaller mainframe machines, and at the other end those
computers called small minis are virtually indistinguishable from those at
the top end of the micro band.
Minicomputers will usually perform rather more slowly than mainframe

machines, and are generally aimed at a small number of concurrent uses
rather than large user networks. They can therefore be rather smaller, be
built rather more cheaply and can be made to be much more tolerant of
changes in their operating environment than their mainframe cousins. A
typical mid-range mini therefore might have the central processor and disk
units contained in a box perhaps the size of two filing cabinets and might
serve six or eight remote users. It will not generally require any special
power supplies or air conditioning; it will generally operate from a normal
13 amp socket outlet in a normal office environment.
Minicomputers may well be used by smaller firms for the same purpose
as larger organisations would use a mainframe, although their capacity as
information handling devices is limited by backing storage typically
measured in tens or hundreds of millions of characters rather than the
thousands of millions of characters which might be normal in a mainframe
installation.
Microcomputers
Microcomputers are, in physical terms at least, the smallest end of the

scale and are essentially personal. They can be divided roughly into two
groups, those at the cheaper end of the scale which are aimed primarily at
the hobby or 'games' market, and the remainder aimed at the business
sector. A typical large single user micro complete with disks, screen,
printer and keyboard can be accommodated on a standard office desk,
whilst a small portable system, colloquially known as a lap-top, will have
keyboard, screen, processor and disks all in one box small enough to carry
in a briefcase and powered by re-chargeable batteries. Whilst micro-
computers are essentially single user systems it is possible to link systems
together in a number of ways. Over short distances systems may be linked
in 'local area networks' (LANs), thus allowing conversation and exchange
of data between different machines, and also allowing a number of com-
puters to share central resources such as high capacity central disk drives
or high speed centralised printers. Over longer distances systems may be
linked using either standard telephone lines or dedicated data transmission
cables. The connection of systems in this way will usually necessitate the
use of a modem (modulator/demodulator) at each end of the line to
convert the output signal from the computer into a suitable form for
transmission over a normal telephone line (see also chapter 6). Data
transmission speeds over long distances using telephone lines will be
considerably slower than the speed which could normally be expected
using a LAN.
The range of business microcomputers now available is very wide,
ranging from machines costing only a few hundred pounds at one extreme
to systems costing tens of thousands of pounds at the other. They are very
reliable and much publicity centres around their alleged ease of use. After
a late start IBM have now, by pure commercial muscle, established the
unofficial industry standards, and most micro manufacturers, apart from
one or two that have specialised in particular sectors of the market, offer
'IBM compatibility'. This does not necessarily mean that their hardware is
constructed in the same way or to the same standards as those manufactured
by IBM, they merely claim that their machines will run software written for
the IBM range. A survey conducted by lnfomatics magazine in September
1987, lnfomatics (7 987), found 53 different manufacturers offering IBM-
compatible machines, but beware - some machines are more IBM-
compatible than others!
Methods of processing data
Batch processing
In general, present day commercially used computer systems are only
capable of executing one instruction at a time. In early systems information
was presented to the computer either on punched cards or on punched
paper tape, the whole job being prepared before being presented, as a
batch, to the machine. The machine then processed the data, working
exclusively on that batch until completion before beginning the next task.
This method of processing is known as 'batch processing', and is still in
use for commercial applications where large amounts of data are required
to be processed with little human intervention, for example a large monthly
company payroll or the final printing of a bill of quantities. Such a method
of operation is ideally suited to this kind of work and many mainframe
computers use this approach extensively for commercial data processing.
When using a batch processing approach incoming jobs are 'queued' in
an input queue, and output from completed jobs will be placed onto an
output queue. A diagrammatic representation is shown in figure2.2. Although
in early systems the information was presented on cards or punched tape it
is now much more likely to be submitted via a video display terminal.
Most commercial systems will allow multiple input queues to allow jobs
to be processed according to criteria set by the system operations manager.
Examples of such criteria might include the time of submission (a simple
first come first served approach), but this is likely to be modified by factors
such as the priority requested by the user (higher priorities often incur a
premium charge for faster turnaround), and the level of processing
resources required to do the job (processes expected to generate large
amounts of print or jobs requiring special arrangements may be given a
lower priority by the system manager because of the resources required).
USER 2
us ER 1 USER 3
NEW JOBS
ADDED TO
QUEUE
' JOB 11
I
JOB 10
JOB 9
8
--
JOB
JOB 7
JOB 6
INPUT QUEUE
COMPUTER
SYSTEM
(PROCESSING JOBS)
OUTPUT QUEUE
JOB 4
OUTPUT RETURNED
JOB 3 TO USERS WHEN
PROCESSING
JOB 2
COMPLETE
JOB 1
USE R 1 USER 3
USER 2
Figure 2.2 Batch processing

Interactive processing
Batch processing is clearly suited to extensive data processing operations

where the user is prepared to submit his information for subsequent
processing and does not expect an instantaneous reply. It is plainly not
suited to operations such as the interrogation of a database, where a
speedy response is required. In order to cope with this type of processing an
alternative approach, called 'interactive processing', is often employed.
For this type of work the time taken by the operator in keying in the
enquiry and the time taken by the computer in passing the result back to
the user is very long when compared with the speed at which the computer
can actually process a request. It therefore follows that if batch processing
techniques were to be. used the processor would be idle for a substantial
proportion of the total time. In this case therefore the computer no longer
concentrates on one job to the exclusion of all others, but instead divides
the available processor time cyclically among the users, giving each user
exclusive use of the computer for a fixed period of time before saving the
task being processed and proceeding to the next user. Each user is therefore
allocated a 'slice' of the available processing power, and the process is
known as 'time slicing'. A diagrammatic representation is shown in figure
2.3.
The time slicing approach is a great improvement on batch processing
techniques, and in a well balanced installation gives each user the illusion
of having exclusive use of the machine. The system is however critically
reliant upon the correct balance being maintained between the number of
users and the central processing power available if response times for
individual users are not to become unreasonably long. Once the number
of users passes the critical level the performance of the system degrades
rapidly, and the only solution is to add more processing power. In addition
the amount of disk storage available may well need to be increased as the
number of users increases. The approach therefore lacks a certain amount
of flexibility and requires very careful management if it is to be successful.
Many minicomputer systems are based upon this philosophy.
Distributed processing
In an attempt to overcome the difficulties posed by the above systems, and

to meet the increasing demand from users for 'desk-top' computer power,
the approach now adopted by many large organisations has been to
distribute processing power from the centre to the users in the form of
USER 1 USER 2
PROCESSOR CYCLES
ROUND ALL USERS
CENTRAL BY ALLOCATING EACH
USER A 'TIME SLICE"
COMPUTER
BEFORE SAVING THE
SYSTEM
TASK AND MOVING
ON
USER t. USER 3
Figure 2.3 Interactive processing

microcomputers used as intelligent terminals to the central machine. Such

a system is known as 'distributed processing', and is illustrated dia-
grammatically in figure 2.4. Terminals may be equipped with their own disk
drives and printers, and such systems will normally allow authorised users
the right to both process data locally and also transfer information to and
from the centre, subject of course to operational controls. Systems such as
this will also usually act as a kind of local area network allowing users to
communicate with each other via the central machine.
USER I US ER 2
0 0
CENTRAL MAINFRAM E
MASS STORAGE AND
CENTRAL SPECIALISEO OUTPUT
MAINFRAME
FACIL ITIES [PLOTTER
COMPUTER LASER PRINTER ETC I
USER < USER 3
Some commonly quoted measures of size and performance
Microcomputer manufacturers usually describe the 'size' of a system by

defining the storage capacity of the central processing unit, sometimes
called the main memory, and the total amount of backing storage (usually
disk space) accessible by the processor at any one time. The 'size' of the
central processor is usually quoted in 'bytes', where 1 byte generally
equals 1 character, but because the numbers tend to be large it is
conventional to give the information in 'K' bytes (KB) where K=2 10 or
1024. Sometimes KB is colloquially abbreviated simply to K, so a machine
described as having 256K (256KB) of main memory will therefore have a
central processing unit with the ability to store 256 x 1024 characters of
information. This measure is important since information held in main
memory can be manipulated much more quickly than information held on
disk, and it is therefore likely that the larger the main memory the faster the
machine will be able to process complex programs using sizeable volumes
of data.
The amount of backing storage available is also usually given in bytes,
but because the numbers tend to be even larger than for main memory,
figures tend to be quoted in 'M' bytes (MB) where M = 220 or 1 048 576.
1MB therefore= 1K2 bytes. As an indication of what this means in terms of
text, Gooch(1986:p13) estimates that 1MB is approximately equivalent to
400 average A4 pages. Backing storage figures for larger computers,
particularly mainframes, may be given in Giga-bytes (GB) where 1GB =
230.
Whilst such measures are useful in gauging the capacity of a system they
say nothing at all about its performance. It is rather like saying a car has
four seats. In addition to the above therefore some manufacturers quote the
speed at which the processor operates, usually in cycles per second quoted
in mega-hertz (MHz), but this again tells a prospective user little about
how well the machine will actually perform a particular task in a working
environment; to return to the car analogy the measures we have so far are
equivalent to the number of seats and the maximum speed of the engine.
What the prospective buyer really wants is some way of comparing how
well each ofthe machines on offer will do the job, and the most often used
way of attempting this is to construct some hypothetical test data consisting
of a mix of representative operations which can then be processed on each
machine, the results being compared one with another. Such tests are
usually termed 'benchmarks', and as Jones (1987) says:
Good benchmarks should compare systems doing the same tasks, and
ideally those tasks should closely resemble the ones users will actually
want to carry out.
One set of benchmarks developed to test the calculation speed of micro-
computers has been developed by the magazine Personal Computer World
and is described in Walker (1986). Benchmarks can of course also be
developed to compare the performance of a number of software packages
on the same machine, rather than of the machines themselves, and an
example of a set developed to compare the relative performance of micro-

computer based data manipulation packages is shown in Lang (1986).
The problems of comparing computer performance are not, of course,
new or limited to microcomputers. With larger computers the problems
become more complex and the units of measurement more numerous.
Among the best known are 'Mips' (millions of instructions per second),
'Mflops' (millions of floating point operations per second) and the Whetstone
test, a mix of typical programming instructions. A table of some of the
more widely used measures for large computers is given in Jones (1987),
although Jones is rather sceptical both about the way in which some
manufacturers express the test results and even about the usefulness of the
measures themselves:
But there is a basic problem with all types of benchmarks. Their
relevance to any user's problem depends on how closely the processing
task used for the benchmark matches the task the user wants to carry
out
Nevertheless Jones reports a senior executive employed by a large computer
manufacturer as saying:
What our sales force and customers want is a number however good or
bad.
Input devices
Video display terminals
These devices currently provide the most popular and widely used method
of interactive communication between the user and the computer, and all
modern computers will have at least one. The unit is connected to the
computer by a cable, and information typed via the keyboard is transmitted
to the processor. Terminals may communicate with the host computer
either on a character-by-character basis or the input may be 'buffered'
within the terminal until some special key, often the 'carriage return' or
'enter' key is pressed. Responses from the computer are displayed on the
screen.
Most keyboards will have the main character keys set in the normal
typewriter style (often called the QWERTY pattern after the arrangement of
the characters on the top row), although other patterns have been tried
from time to time. All terminals will also have other special purpose keys
as well, and many terminals have 'function' keys, the meaning of which
can be controlled from within particular programs.
Screen displays are available in a range of colours, green characters on a

black background perhaps being the most often used although yellow or
orange on black and either black on white or white on black are also
popular. Some computers can use colour display terminals which allow
output to be displayed in a range of colours although the resolution
particularly with the cheaper models is often poor. Some special (and
usually expensive) terminals have the ability to display high resolution
computer generated drawings, and such screens are commonly used for
computer aided design applications.
Optical mark recognition (OMR), magnetic ink character recognition

(MICR) and optical character recognition (OCR)
These systems, as the names suggest, are all attempts to 'read' information
either written or printed on paper.
Optical mark recognition relies on the location by a scanning device of
written marks (generally indicating a 'yes' response) or the absence of
marks (usually indicating a 'no' response) on a pre-printed form. It is
obvious that OMR techniques are therefore very useful for jobs such as the
evaluation of data gathered during market research studies. The limitations
of OMR techniques are the limited range of input options, and the fact that
the forms must be manufactured to reasonable tolerances and must be kept
clean if spurious entries are to be avoided, although some systems do
overcome the dirt problem with the use of special graphite pencils. Typical
data capture speeds are between about 25 000 and 30 000 characters per
minute.
The logical progression from OMR is the recognition of actual characters
rather than just marks on a piece of paper. The first method evolved
entailed printing the characters in a special magnetic ink in a very stylised
type font, the characters then being scanned electronically. This method is
known as magnetic ink character recognition, and is perhaps most widely
known for its use by banks for sorting and processing cheques.
MICR is however only of use where the characters can be printed on the
page in magnetic ink. A different approach is required to recognise
characters written on an ordinary document, and this has led to the
development of optical character recognition (OCR) systems. Most OCR
readers examine each character as if it were made up of a pattern of tiny
dots. When the whole pattern has been scanned the pattern detected is
compared with a set of standard patterns held in the machine and when a
reasonable match is found that character is considered to be the one read.
If no match can be achieved within the tolerances allowed then a read
error occurs and the character is flagged as an error. Two standard
character sets were originally used, OCR 'A' and OCR 'B', but OCR
readers are now available which can cope with typewritten documents.
Much more development is required before OCR techniques can be applied
to hand-written documents. Modern OCR readers will operate at about
2400 characters per second but are conventionally operated at much
slower speeds for greater accuracy. A potted history of OCR and an
indication of future research in this field was given in Howard (1986).
A number of companies have experimented with OCR systems to read in
bills of quantity for subsequent processing by contractors estimating
packages and an example of such a system is described in Lane (1985).
Perhaps the most promising initiative in this area is the development, in
America, of a hand-held scanning device reported by Bajarin (1987) and
now available in the United Kingdom. Such a device will vastly increase
the scope for this type of data input. However whilst the technical prob-
lems of transferring text from typewritten copy into the computer may soon
be solved, the software required to interpret the words is very complex.
Digitisers and graphics tablets
A digitiser (and its smaller brother the graphics tablet) is a form of input
device which permits co-ordinate information to be taken from a drawing
directly using an electronic pen to either trace around shapes or 'pick off'
particular points. Conventional digitisers work by having a very fine matrix
of wires embedded in the board which register a signal in both horizontal
and vertical directions when the board is touched with the special pen.
The computer can therefore calculate the X and Y co-ordinates of the point
selected by the pen, and suitable software can then calculate distances
between points using conventional co-ordinate geometry. Such devices are
of obvious use in computer aided measurement (see also chapter 9).
An unusual form of digitiser is that used by Techsonix in connection
with their computer aided measurement software. This system uses ribbon
microphones positioned along the X and Y axes of the board and the signal
is produced by a special pen which generates an ultra-sonic sound when
pressed onto the board.
Digitisers are available to accommodate A3, A2, A1 and AO size drawings,
and smaller graphics tablets are available to accommodate A4 size sheets.
The price increases rapidly wlth the size.
Pressure pad and similar devices
These devices usually consist of either a pressure pad or a solid state board
similar to a digitiser. Information is written by hand onto a sheet of paper
laid over the board, the characters written are deciphered by a small in-
built microprocessor and the resultant code is transmitted to the host
computer. Such devices are of obvious benefit although the units available
at the time of writing tend to be expensive, are limited in size and can only
'read' capital letters.
Hand held data collection devices
An interesting recent development is the introduction of the 'pocket

computer' which can be used out of the office as a hand held data
collection device. One such machine was reviewed in Gross-Niklaus
(1987), who describes a unit 140mm x 80mm x 30mm containing 32K
ROM and 16K RAM with a two-line, 32-character Iiquid crystal display.
Backing storage is provided by exchangeable 32K plug-in RAM packs. The
unit also has an optional communications link allowing data gathered in
the field to be down-loaded to a larger machine upon return to the office.
Devices such as this offer great potential for use on site, perhaps during
the post contract stage, and their use can be expected to spread rapidly.
The introduction of a modified version of such a device used as an
electronic field book for land survey data is described in Computer Weekly
(1988).
Output devices
Printers
Printers comprise a very wide range of machines indeed. Beginning at the

largest end of the scale, the printers used by a large mainframe computer
are physically large (typically a free-standing cabinet say 1200mm x
1OOOmm x 1500mm high), generally capable of printing a complete line of
132 characters at once. Mechanical printers are still widely used and are
capable of speeds up to say 2000 lines per minute, but the newer machines
employ a laser printing technology and speeds of over 10 000 characters
per second are claimed.
Until very recently, the vast majority of printers sold for mini and
microcomputers were mechanical, and were either of the daisy-wheel or
dot-matrix type. In recent months, however, small laser printers have
become extremely popular and the market share of daisy-wheel printers
has rapidly declined. Printers used for micro and minicomputers tend to be
much smaller, slower and cheaper than the printers used on mainframe
computers.
Daisy-wheel printers take the name from the type of wheel, a plastic or
metal wheel with the character set arranged on individual stalks radiating
from the centre giving a flower-like appearance. A wide variety of different
fonts is available by changing the print wheel, and the main advantage of
the daisy-wheel printer is the print quality which is indistinguisable from
that produced by an electric typewriter. The main disadvantage is the
speed which is limited to about 55 characters per second.
Matrix printers generate the characters from a pattern of dots formed by
pressing small needles against the paper via a conventional typewriter style
ribbon. Most matrix printers will possess the ability to print either at high
speed in 'draft' mode (say up to about 450 characters per second), or to
print what is known as 'near letter quality' (NLQ) by building up more fully
formed letters from multiple passes across the document. Near letter
quality print is therefore much slower than the draft mode, but nevertheless
does produce a very acceptable typewriter-style text. Most matrix printers
have the facility to print a number of different fonts, at different pitches and
with text in different sizes, even changing from one to another actually in
the body of a document; they are thus much more flexible than the daisy-
wheel type.
As noted earlier, small laser printers developed by companies such a
Xerox and Canon using state of the art electrostatic printing techniques are
rapidly gaining ground in small computer installations, due mainly to their
ability to produce virtually any style of print very quickly, very quietly,
with superb clarity at speeds up to about twenty pages per minute. Laser
printers are becoming very popular when used in connection with 'desk-
top publishing' software (see also chapter 5). The main disadvantage of
laser printers at present is the relatively high cost but no doubt this will be
overcome in time.
Other types of printer, based on alternative technologies, are also
available; for example, some manufacturers use an ink-jet mechanism
whilst others use a thermal technique using special paper. Such machines
are however at the moment very much in the minority. Printer technology
is however constantly changing, but Stobie (1988) contains a good review
of the contemporary scene. Stobie also comments that significant techno-
logical improvements now mean that ink-jet printers may be about to make
a significant impact on the market. Such machines have advantages over
laser printers in that they can print on continuous stationery, can handle
paper of greater width, are as quiet or quieter in operation and can print in
a number of different colours.
Plotters
Plotters are used to enable the output of drawn information from the
computer, and the principal use therefore tends to be in computer aided
design although small plotters are sometimes used for the production of
graphs and charts. Plotters may be of two basic configurations:
i) 'Drum' plotters use a continuous roll of paper moved by rollers, and

pens fixed to a carrier which allows movement across the paper. The
main advantage of this type is that, whilst the width of drawing which
can be produced is limited by the width of the rollers, the length is
unlimited. It is therefore possible to produce very large drawings.
Disadvantages are that the plotting speed tends to be slower than the
'flat bed' type, and it is very difficult to use pre-printed paper.
ii) 'Flat bed' plotters have a table upon which standard size sheets of
paper are secured, and units are available in a range of sizes capable
of handling paper ranging from A4 to double AO. Prices rise sharply as
the size increases. Plotting speeds may be higher than for drum
plotters.
Plotters may be connected 'on-line' to the computer for direct plotting, but
in a large computer system the plotter output is more likely to be spooled
on a magnetic tape or similar medium for plotting 'off-line' (i.e. indepen-
dently of the main computer).
A variety of different types of plotting mechanisms is used, but the most
common has removable pens using either liquid ink, roller-ball or fibre tip
in a range of colours. Plotters using electrostatic technology are however
gaining ground in some areas of work, principally due to their ability to
print 'full colour' drawings.
Backing storage devices
Magnetic tape units
Magnetic tape storage is probably the most widely used form of computer
readable secure storage. Some early micro systems used cassette tapes of
the kind used in portable tape recorders, but modern cassette systems use
special cassettes which are much larger, more expensive, can store much
more information and can be accessed much more quickly. Mini and main-
frame systems use special computer tape reels W' wide with either seven or
nine recording tracks. Information is encoded as patterns of magnetic dots
across all tracks. The principal disadvantage of magnetic tape systems is
that the information must be accessed serially, i.e. to find a specific piece
of information it is necessary to begin at the beginning of the tape and read
each record in turn until the required information is found. This obviously
means lengthy (at least by computer standards) search times if the required
information is near the end of the tape.
Magnetic disk units
The most widely used backing storage devices are magnetic disk units.
Each disk consists of a flat platter coated on both sides with a magnetic
recording medium. Information is encoded on the disk as a pattern of
magnetic dots in concentric circles (called 'tracks') usually on both sides of
the disk. Disks therefore overcome the main problem encountered when
using magnetic tape, that is that information can only be accessed serially
beginning from the start of the tape. With disks, in addition to dividing
each surface into tracks, surfaces may also be divided into sectors, and the
tracks sub-divided into blocks. Direct addressing of information is therefore
possible by quoting the appropriate track/sector/block address.
In large systems several disks may be mounted one above another on a
common spindle and the resultant assembly is known as a disk 'pack'.
Three main types of magnetic disk are in common use and are shown
diagrammatically in figure 2.5.
Fixed head drives

As the name suggests the read/write heads in these units do not move, and
it therefore follows that they need a separate read/write head for each
track. The speed of access is thus very fast indeed, but they also tend to be
the most expensive. Fixed head disks are not generally removable, and
will usually only be found on mainframe or the larger minicomputers.
Storage capacities are likely to be measured in gigabytes.
Hard disk moving head devices

These devices only have one read/write head for each recording surface
rather than a separate head for each track, and the head is therefore
required to move across the disk, usually by means of a stepper motor.
Where the disk is 'fixed', (i.e. non-removable) as for example the
'Winchester' technology disks commonly found in desk-top micro-
computers, the read/write heads will usually be the 'take off and
land' type which rest on the disk surface whilst the disk is stationary and
'fly' as the disk begins to spin. Where the disk is removable however the
head must be initially held away from the disk and loaded towards it when
the disk has reached operational speed. Devices ofthis type are used on all
types of computer, although for mainframe systems they have tended to be
displaced by 'head per track' units because of the faster access times
required where large volumes of data are concerned. Typical storage
capacities might range from 20MB on a desk-top micro to several hundred
MB for a mainframe unit.
DISK "PACK" IS FIXED READ/WRITE

NON-REMOVABLE HEADS. ONE PER TRACK
'Head per track' or fixed disk device
DISK' PACK MAY

BE FIXED OR MOVABLE READ/WRITE
REMOVABLE HEADS
Moving head device
DISK ENCLOSED WITHIN PLASTIC

OR CARDBOARD PROTECTIVE SLEEVE
.../":--=~:::;:;-....,~ MOVABLE READ/WRITE

HEADS ACCESS DISK
THROUGH SLOT IN
PROTECTIVE SLEEVE
RECORDING TRACK
40 - 100 PER DISK
Floppy disk device
Figure 2.5 Magnetic disk storage

Floppy disk or 'diskette' systems

These units comprise the smallest members of the magnetic disk family. A
floppy disk consists of a thin plastic disk coated on both sides with a
magnetic recording medium and contained within a cardboard or plastic
dust cover. A number of different sizes of disk are in common use
including 3", 3.5", 5.25" and 8" diameter. Floppy disk systems are much
slower than hard disk devices and, unlike hard disks which revolve all of
the time, floppy disks generally only spin when they are actually being
accessed. They are extensively used in microcomputer systems, and
storage capacities may vary from about 256KB to around 1MB.
Optical disks
Optical disk technology has been around since the late 1970s, and during
the period since the mid 1980s a number of systems have been built which
integrate video-disk playback systems and contemporary computer
technology. Perhaps the best known application is the BBC
'Doomsday' project, described in Linderholm (1987). Data is stored on
optical disk by 'burning' small pits in the surface with a laser, and may
later be retrieved by scanning again with a laser beam. Such systems have,
for a number of years, been proposed as an alternative to magnetic disk.
The principal advantages are seen as being the ease with which visual
information can be stored and accessed by computer database
management software, coupled with an increase in the amount of
information which could be stored on a disk of any given size. The major
disadvantage however is that once the data is written to the disk it cannot
currently be subsequently erased. The computer industry, with its
insatiable appetite for acronyms, refers to such systems as WORM (Write
Once Read Many times) technology.
A report published in 1987 by a leading research company (Wharton
(1987)) predicts a big growth in systems built around optical disks during
the 1990s, predicting total sales of such systems at £66m by 1991
compared with only £2.5m in 1986. Optical disk units are now offered for
use with personal computers, and a number of manufacturers are believed
to be close to offering systems which will allow information to be erased or
overwritten.
Linking computers together
To the normal person in the street it is unbelievable that, although the

computer industry is considerably less than half a century old, is a child of
this wonderfully co-ordinated twentieth century, and represents the
absolute pinnacle of mankind's highest technological achievement, it is
nonetheless not possible to take a disk containing information prepared on

one type of computer (for example an Apricot Xi) and read it directly into
another type of computer (for example an IBM PC). Not only do computers
use a bewildering range of disk sizes, but even when the disk size is the
same, and even if the same software package is implemented on both
machines, there are any number of different formats in which the data may
be recorded which form a barrier between one machine and another. In
order therefore to transfer information between, for example an Apricot
and an IBM PC, it is necessary to resort to connecting the computers
together with pieces of wire, reading the information from the Apricot disk
into the Apricot computer, using some software in the Apricot to transmit
the data down the wire, using some more software in the IBM machine to
catch it, and finally writing the data to an IBM disk.
In view of the lamentable lack of co-ordination between one
manufacturer and another, it is perhaps fortunate that the computer
industry has been able to develop interface protocols to allow data to be
exchanged electronically, and there are now software packages available
for all of the popular machines which will perform the required
transmission and reception functions. Unfortunately the electronic
transmission of information is still not so easy as it sounds since both the
source computer and the target computer need to have the relevant
software loaded in order that the data can be successfully transferred; the
process is analogous to a telephone call which requires the attention of
both parties if a successful transfer of information is to take place.
Local Area Networks (LANs)
Sooner or later an organisation which uses large numbers of stand-alone

desk-top computers will begin to consider the possiblity of connecting
them together, eitheno allow data to be transferred from one machine to
another, or to allow a number of computers to access common central
facilities such as high capacity disks or perhaps high speed laser printers.
Where a number of computers are to be connected together over short
distances (say within the same building) the most common method is to
construct some kind of local area network or LAN. There are many types of
network, but all are essentially based on either a 'star' format where each
separate user or 'node' is connected directly to a central network server, a
'ring' format where all of the nodes are connected together to form a
continuous endless loop, or a 'bus' format where the two most distant
'nodes' are connected together and all other nodes form branches from this
main cable. Each is shown diagrammatically in figure 2.6. In addition to the
physical cabling required, networks will also require some software
running in all of the separate nodes to control access to and traffic on the
,
USER
MESSAGES PASSED FROM USER -

USER VIA CENTRAL CONTROLLER
Star network
MESSAGES PASSED AROUND THE

RING MAY ONLY BE "CAUGHT" BY
THE USER TO WHOM THEY ARE
ADDRESSED
Ring network
MESSAGES ARE "BROADCAST" ONTO

THE BUS BUT CAN ONLY BE
DECODED BY THE SPECIFIC ADDRESSEE
Bus network
Figure 2.6 Basic network types

network itself. Note that this software does not normally prevent the nodes
from working as stand-alone computers in their own right, but does allow
them to communicate with each other should the need arise. An excellent
introduction to the practicalities of local area networks is given in Malcolm
(1987a) and Malcolm (1987b).
Long distance communication
Where communication is required over long distances, the most common

method is to use telephone lines. The cheapest method, although it does
have disadvantages, is to use the normal public switched telephone
network, but greater transmission speeds can be achieved by using a
private wire.
The main problem with data transmission via the telephone system is
that the signals generated by the computer must be modulated in order to
make them suitable for transmission through the telephone system, and
subsequently de-modulated by the receiving computer. This process
generally involves the use of a supplementary box of electronics known as
a 'modem' (Modulator/De-Modulator) attached to each computer. Another
major problem in using British telephone equipment has been the very
slow data transfer speeds, due principally to the use of electro-mechanical
switching equipment in the telephone exchanges, but these problems are
steadily being overcome by the introduction of electronic telephone
exchanges and fibre optic cables.
There are a variety of different protocols in use for the transfer of data in
this way. Toothill (1987) contains a useful overview of the major public-
domain communication protocols, together with an introduction to the
kinds of problems likely to be encountered in practice. Modems and
communication protocols are considered in more detail in chapter 6.
References
Bajarin T. (1987) 'Microsoft signs up with Alps for handheld scanners' Personal Computer
World Vol.1 0 no.8 (August 1987) p. 90
Brandon P. S. and Moore R.G. (1983) Micro Computers in Building Appraisal Granada
Computer Weekly (1988) 'Survey data captured' Computer Weekly 14.1.88 p. 32
Feigenbaum E. and McCorduck P. (1983) The Fifth Generation Addison Wesley
Gooch B. (1986) The Computer Guide for Builders Privately published
Gross-Nicklaus M. (1987) 'Psion organiser II' Personal Computer World Vol.10 no.8
(August 1987) pp. 114-118
Howard A. (1986) 'Character Recognition' Systems International November 1986 p. 69-70

lnfomatics (1987) 'Survey' lnfomatics September 1987 pp. 80-92
jones P. (1987) 'Odious comparisons' lnfomatics December 1987 pp. 60-64
Lane K. (1985) 'Bi lis made easy' Construction Computing Summer 1985 p. 23
Lang K. (1986) 'Screentest- PC File' Personal Computer WorldVol.9 no.12 (December 1986)
pp. 168-171
Linderholm 0. (1987) 'A machine for all times' Personal Computer World january 1987 pp.
160-165
Malcolm P. (1987a) 'Run around the clock' Personal Computer World February 1987 pp.
122-126
Malcolm P. (1987b) 'Caught in the net' Personal Computer World May 1987 pp. 134-138
Stobie I. (1988) 'Silence is golden in the printing marketplace' Computer Weekly 14.1.88
pp. 30-31
Toothill P. (1987) 'Clearing the line' Personal Computer WorldApril1987 pp. 174-176
Walker N. (1986) 'Time's (nearly) up!' Personal Computer World Vol.9 no.12 (December
1986)pp. 164-167
Wharton (1987) Image processing Wharton Information Systems
Widdeson P. (1980) Computers and Public Authorities Privately published
3 The Selection and
Management of an Office
Computer System
Objectives

• Discuss the factors to be considered when choosing an office computer
system.
• Discuss the management problems inherent in the introduction and
subsequent successful day-to-day operation of an office computer
system.
• Develop a management and training strategy to enable new systems to
be introduced as painlessly as possible.
Choosing an office computer system
A common problem with the introduction of computers in a professional

practice is deciding how to go about it. There are pressures from the
press, salesmen, often from staff, and from attitudes of 'keeping up with
the jones'. The first steps are often difficult, usually the most crucial,
invariably the ones with the most pitfalls.
CICA (1981 :p. 8)
Choosing an office computer system, as the above quotation indicates, is
often a difficult business, but in many organisations the process is made
more difficult than it need be because the decision to acquire a computer is
made for the wrong reasons. There are, of course, many reasons why an
organisation might want to introduce computers, but among less valid ones
might be:
i) to keep up with the Jones- 'all our competitors have computers so we
must have one as well'
ii) 'computers are the tool of the future and you ignore them at your
peril'
iii) pressure from enthusiastic staff to 'have a go'
All of the above are based on 'feelings', none is based on any kind of
rational assessment of the contribution which the machine might make to
The Selection and Management of an Office Computer System 43
the organisation as a whole, and whilst the above statements might, at least
in part, even be true, an organisation which installs a computer system
using any of the above for justification is doomed to disappointment. There
is no magic in computer systems; computers will not take an unprofitable
business and make it successful. If computer systems are to be used
effectively then the application of them must be carefully researched and
the costs and benefits must be realistically assessed.
Both Parret & Stead (1986) and Atkin (1987) consider the problem of
choice of an office computer system. Essentially what is required is a three-
stage process to discover:
i) Is there a problem?
ii) If there is a problem is a computer solution appropriate?
iii) If a computer solution is appropriate, how is the most suitable system
to be chosen, implemented and operated?
Stage 1 - Is there a problem?
This first stage is really part of the routine monitoring process which every
business should undertake as a part of its overall business strategy. The
questions to be asked therefore are questions about the performance of the
business itself:
i) What are the objectives of the organisation?
ii) Are they being achieved to the satisfaction of the p~oprietors?
iii) Will the organisation continue, as far as can be foreseen, to achieve
them in the future?
Questions such as these should indicate whether any further investigations
are justified. If the honest answer to all of the above is 'Yes' then the
indications would be that there is no need to undertake detailed investiga-
tion of computer systems which might be costly; if however the answer to
any of them is 'No' then there is a need to progress to stage 2.
The above scenario views the possible selection and implementation of
a computer system very much as the solution to a particular problem. A
further question which may arise, particularly, in a forward-looking organ-
isation, might be:
Although I'm satisfied that the organisation is achieving the required
objectives at the moment, could computers help me to achieve them
more effectively in the longer term?
This question is very much more difficult to answer as it implies the need
for some on-going appraisal, research and development. The extent to
which an individual organisation will consider this to be worthwhile will
obviously depend upon the importance which the organisation attaches to
being at the front of the pack. It must be remembered that research and
development is a high risk venture which can be very costly.
Stage 2- What can I do to resolve the problem?
Is the problem simple or complex7
Stage 1 of the process identifies that a problem exists, the next question to
be resolved is to decide precisely what the cause of the problem is and
what is required in order to overcome it. Some problems may be simple in
that one particular process or activity can be seen to be inefficient, but
others may have a number of causes; a number of contributory factors may
combine together to create the obstruction. Problems such as these may
require one composite solution or they may simply be symptomatic of
other hidden problems. In any event it is vital to be sure that the precise
cause(s) of the problem have been identified before going further.
Having established the precise cause of the problem(s), it is then
necessary to ask two further questions:
(1) Would a computer help?
(2) Would use of a computer be cost effective?
The only way to proceed beyond this point is to:

(1) Appoint someone to find out
(2) Establish his or her brief
(3) Review the findings
The analyst
It is at this point that specialist advice is often sought, although in many

cases there is no fundamental reason why this should necessitate external
consultants. In fact there may be distinct advantages to using someone
from within the organisation, provided firstly that some suitable and willing
person can be found, and secondly that management recognises that
evaluation of this kind may be very time consuming. It is not the kind of
work which can be tackled on a 'part-time' basis in addition to existing
duties. If however:
(i) the problem is very complex in that it affects a range of activities
across the organisation or
(ii) a range of different problems is evident which may necessitate a co-
ordinated solution
then an external consultant may prove to be the better choice although the
cost is likely to be high.
Hunt (1982b) identifies the qualities which an 'in-house' analyst should
possess as:
A suitable candidate for the task must start with an open mind, must
have an analytical approach and sufficient knowledge of quantity survey-
ing procedures to enable extraction of all relevant information'.
Parrett & Stead (1986), whilst in general agreement with the above, go
further:
The person you choose must display the following qualities:
1 Ability to use initiative
2 Self-motivation
3 Willingness to learn and be open-minded- you cannot afford to select
someone with tunnel vision
4 Genuine interest in modern technology
5 Analytical ability and preferably a detailed knowledge of the organ-
isation and its systems
6 A basic concept of computer programming and a broad knowledge of
what computers can do
7 Commitment to the project
It may also be an advantage if the candidate is a user rather than a
computer specialist, because a computer specialist might tend to:
(i) Compromise on management needs in favour of programming needs
(ii) Design systems for ease of programming rather than ease of use
(iii) Fail to analyse the management requirements as well as a user,
basically because a user normally knows the existing formal systems
and the informal systems which surround them
(iv) Have an in-built bias towards hardware and software with which
they are already familiar
(v) Have a lesser sense of urgency and commitment to the project
(vi) Be unable to look beyond the present application
Above all, the analyst must have the authority to do the job and in
particular the authority to investigate and comment on the existing manual
procedures. This will inevitably necessitate complete backing from higher
management, at either director or partner level as appropriate. The analyst
must also of course be given the necessary administrative back-up in the
form of secretarial staff, travel facilities and so on.
The analyst's brief
The analyst should be given a brief detailing exactly what is required

together with a time scale. The detail of the brief will, of course, vary from
organisation to organisation, but remember that this stage is very much a
feasibility study; at this stage we are only concerned with ascertaining
whether or not a computer system would be feasible, appropriate and cost
effective. Remember that the benefits might result either from direct savings
in resources, or be due to a redudion in time, or be the result of producing
a better produd, and the analyst must consider all of these issues. The
likely costs and benefits must then be weighed against all of the other ways
of solving the problem. All costs must include not only the initial cost of
the machine but also the cost of maintenance, staffing, staff training,
accommodation for the computer including possible modifications to services
such as electric wiring, and the likely cost of lost production whilst the
system is installed and commissioned and staff become familiar with its
use. A useful and more detailed discussion of some of the financial issues
involved is contained in DOC (1980) chapter 8.
At the conclusion of the exercise the analyst should present a full written
report for consideration by the board or partners as appropriate and, on
the basis of that report, management must then decide whether or not to
proceed to stage 3.
Stage 3 - Acquisition and implementation
Having decided, on the basis of the report produced during stage 2, to

proceed further, management is now in a position to set an overall budget
for the acquisition of the necessary hardware and software, within which it
must consider the following points:
(i) How is the applications software to be chosen?
(ii) How is the hardware to be chosen?
(iii) Is the equipment to be bought, leased or hired?
(iv) What maintenance cover is required?
(v) What operational procedures are required to ensure the integrity of
the system?
(vi) Who will operate the system?
(vii) What training will be required?
(vii) What accommodation will be required?
(viii) Should future expansion be considered?
(ix) What about the human issues?
How is the applications software to be chosen?
The choice of applications software is by far the most important decision to

be made in the acquisition of a computer system. As CICA (1981) points
out:
Applications software is the link between the problems you want to
solve and their solution by the computer
and Parrett & Stead (1986) stress:
To achieve effective results the user must be able to communicate
effectively with the computer in a language which they both understand.
It is the software, the computer program, that provides this com-
municating link ...
In general a good software package should be:
(i) Robust - the package should be fault-tolerant and it should not be
possible to break it by hitting invalid keys, for example letters when
the system expects numbers or vice-versa, entering an invalid date
etc. A good test is to enter 30 February when the system asks for the
date. If the software does not pick it up as an error then the chances
are the rest of the system is not very robust either.
(ii) Reliable- the package should give the same results consistently and
the answers should always be correct. Do not forget that when you
buy a program you usually sign a software licence agreement which
excludes the vendor from any liability arising from the use of the
program. You therefore take full responsibility for the accuracy or
adequacy of any results. This might be important for example in the
case of programs for the design or selection of structural components.
(See also Hunt (1982a).)
(iii) User-friendly- the system should help you to use it. Menus and error
messages should be clear and unambiguous. Do not rely on the
salesman's demonstration or even your own opinion, get an opinion
either from the person who'll have to use it or from someone who's
already using it, but treat existing users with some care, people are
reluctant to admit they have made a mistake! Very few systems are
truly 'user-friendly':
In the words ofTrevor Nicholas, Head of Information Systems and Resources,
Barclays Bank:
Most (computer) systems are as user-friendly as a cornered rat
(iv) Understandable- the documentation should not only enable you to
see how to use the program, but where the program is to do with
design or opinion the documentation should also explain any under-

lying principles or formulae.
(v) Well supported - for a program of any size there should always be
some form of support service. Ask the vendor what you'd have to do
to get assistance if the program failed in the middle of a vital run at
4pm on Christmas Eve! Ask about future updates to the system.
Carter (1984) reports that the Chartered Institute of Building Services
Computer Applications Panel is to publish algorithms for solution of
building services problems, and argues that the professional institutions as
a whole should take some responsibility for establishing testing and valida-
tion procedures for programs which come within their specialist areas.
Acquiring the software
There are essentially only three ways in which software can be obtained:
(i) 'Off the shelf'
(ii) In-house development
(iii) Specially commissioned from a software house
1 0ff the shelf' software
There is a very wide range of 'off the shelf' software available for many of
the more standard applications, for example word processing, spreadsheet
calculators, data manipulation, accounting, project scheduling etc. If these
are the applications required then it is highly likely that a suitable program
can be found off the shelf.
Hunt (1982a) gives a general checklist of points to examine, and Barrett
(1987b) offers a series of criteria for the evaluation of project management
software. Both lists include the general characteristics described above in one
form or another, and the following general rules may be derived:
(i) How heavy duty is it? Will it contain all the information you need to
process, not only now but in the forseeable future?
(ij) Can it do all of the things you want to do?
able?
(iv) How easy is it to use? How difficult is it to break? Are the error
messages easy to understand?
(v) Does it process the data fast enough?
(vi) How well supported is the package? How good is the documenta-
tion, is there a user 'hotline', how much training is included/
available? Is there an established user base to whom you can talk

before you buy?
(vii) How much does maintenance cost? How much is the next upgrade?
(viii) What machines will it run on? What operating system does it need?
(ix) Will the operators need to be re-trained when the supplier issues the
next release
The correct choice of software supplier is also vitally important if the
installation is to be trouble free. The documentation supplied with many
systems is often difficult for the naive user to understand, and the supplier
is often the lifeline in the event of a problem. He must therefore not only
have a complete understanding of the package himself, but must also be
able to respond quickly and effectively to purchasers' queries and prob-
lems. This may mean either the provision of an adequately manned
telephone 'hotline', or a physical presence sufficiently close to the users'
office to allow a consultant to be able to sort out users' problems in
person.
Finally do not forget that a considerable amount of general purpose
software including word processors, spreadsheets etc. is available either
free of charge or at very low prices through several bulletin boards (see
chapter 5). Whilst such systems can be very good they do, in the main,
tend to be unsupported and their use can therefore be something of a
gamble.
In-house development
For this option to be even considered the organisation must either have
staff already trained who can carry out the work, staff willing to be trained
who can be spared from other duties or be prepared to employ trained
specialist staff to do the work. Software development in-house is a very
expensive activity, and it is most unwise to think that staff who may have
written small programs for themselves in the past are capable of writing
operational systems for use in an office. The only way that this option can
work effectively is for the work to be done by trained computer specialists;
it is therefore an option which should be approached with great caution.
On the other hand however this approach is likely to produce the 'best
fit' solution for any particular organisation. Most of the currently available
quantity surveying packages originated in just this way, and the authors are
now attempting to recoup some of the considerable initial outlay by selling
the product to other practitioners. The quality of such software is very
variable. The development of 'in-house' systems is considered in more
detail in chapter 7.
Software specially commissioned from a software house
As might be expected this is by far the most expensive option and the cost
of the software is likely to exceed the cost of the hardware. It is however
likely to produce a very good 'fit' between the users' requirements and the
final computer solution. An organisation following this route could expect
to shed all of the responsibilty for production of the software, testing and
implementation. The disruptive effects of the installation of the system will
therefore be minimised but the cost is likely to be very high.
The actual process of 'purpose-built' software development is con-
sidered in more depth in chapter 7.
How is the hardware to be chosen?
In order to seled the hardware the organisation's computing philosophy

will need to be determined, particularly in terms of the way in which the
systems are to be used and the anticipated future degree of expansion. For
a small firm purchasing a single machine the problem is fairly simple, but
in the case of a larger organisation a decision must be made as to whether
to use:
(i) a number of stand-alone computers, each used for a separate discrete
task
(ii) a number of microcomputers linked in a network
(iii) a large central computer with a number of terminals
(iv) a distributed processing system with a large central disk
The initial decision will depend largely upon the uses to which the
machine is to be put. For example in the case of a system to be used for
processing bills of quantity where data for jobs needs to be captured from
a number of locations, then a suitable arrangement might be either a large
central computer with terminals or a distributed processing system,
whereas if the system is simply to be used for ad hoc word processing and
spreadsheet applications then a better arrangement might be a number of
separate computers operating either independently or able to communicate
via a local area network. The actual configuration will of course depend
on the particular firm concerned, but a critical factor is the number of
people expected to use each terminal. There is an obvious conflict here in
that, whilst providing a separate terminal on every user's desk might appear
to represent the best solution from the human point of view, it is likely to
be very expensive and none of the terminals is likely to be used to full
capacity. On the other hand if one were to provide a central terminal
resource for a number of people then there are bound to be times of
conflict, for example when several people need to use the terminal at the
same time. Management must therefore decide, as a part of its computing

philosophy, which approach the organisation will adopt, and perhaps
which users have priority in the event of conflict.
In terms of the particular make or type of computer, the choice may
already have been narrowed by the choice of software and the operating
system under which it will run. In the case of small computers it is worth
remembering that more 'off the shelf' software is available for the IBM
range of microcomputers than for any other, but also that some of the
machines which describe themselves as 'IBM compatible' represent far
better value for money than the IBM machines themselves.
Is the equipment to be bought, leased or hired?
The decision about how to pay for the chosen computer can, as Lawrence
(1986) points out, be rather tricky. Whilst the position of any one particular
company is likely to be to some extent unique, and changes in tax
allowances and interest rates can alter the position, Lawrence offers the
following general guidelines:
(i) Only rent or lease for less than two years if you're sure you have no
long-term use for the system and aren't worried about the money
(ii) Buy if you've got the money in the bank
(iii) Lease if you aren't worried about the tax concessions and you
haven't the money to buy
(iv) Don't lease for as long as seven years, over that pe.riod it is cheaper
to buy
(vi) Use an operating lease if you want to keep the expenditure off the
balance sheet
The best advice is to seek the help of an accountant to assess the precise
position at the time.
What maintenance cover is required?
Although modern computers are extremely reliable they do break down

from time to time, and adequate cover for maintenance, and repair of the
hardware in the event of an equipment failure is obviously of prime
importance if the system is to consistently fulfil its purpose.
There are a number of alternative strategies available. It is, of course,
possible to 'go bare', in other words to make no specific provision for
repair and to simply rely on getting the machine repaired if and when it
fails. This is likely, in the long term, to be the cheapest option, but tends to
be a risky business since there is no guaranteed response time and it is
therefore impossible to predict how long the machine will be out of action.
Whilst this might not particularly matter in an organisation using computers
in a non-critical way and where an easy return to alternative manual
methods is possible, for example simple word processing, it is likely to be
of considerable importance where the computer has become established at
the centre of a business and perhaps particularly where the computer is
used to keep financial records. One may also consider 'going bare' in the
case of an organisation using a number of computers of the same type,
particularly if there is some inbuilt redundancy, where one computer can
be be taken out of service for repair and the work load can be redistributed
temporarily onto other machines.
For those organisations adopting a rather more structured policy there
are a number of maintenance plans available depending upon the type of
equipment installed and the level of service required. At the largest end of
the scale, mainframe and minicomputer suppliers will advise on the
selection of a suitable maintenance plan, whilst a number of specialist
companies exist to serve the microcomputer market.
The cost of a maintenance plan will depend basically upon the level of
service required, and companies will prefer to quote for particular installa-
tions depending upon the customer's precise requirements, but the cost will
largely depend upon the required speed of response. A figure of up to 15
per cent of the initial capital equipment cost per year for example might
give a microcomputer installation a guaranteed repair or replacement of
faulty equipment within 24 hours.
What operational procedures are required to ensure the integrity of the

system~
In addition to documentation detailing the use of each application,

operational procedures will also be required in all computer installations
to ensure both the continued security of the machine and the data, and
also to control/regulate the way in which shared computer resources are to
be allocated and the priorities to be given to each task. The formality of
such operational procedures will, to a large extent, depend upon the size
and complexity of the organisation; in a small firm with perhaps one or
two computers the procedures may be very simple, but in a large organ-
isation they will, of necessity, be more complex, but all operational
procedures should be written down in an operations manual. Note that
whilst security of the majority of a firm's data is a purely domestic concern,
where personal data is concerned the Data Protection Act imposes a legal
responsibility on data users to ensure that such data is neither lost,
accidentally destroyed or subject to unauthorised access (see also chapter
4).
Perhaps the most important aspects of operations procedure, which must

be adequately covered regardless of the size of the installation, are those to
do with ensuring the security and integrity of both the machine and its
data. They will also detail the actions to be taken, and the staff responsible
for taking them, in the event of a disaster such as failure of the machine,
loss of data, or perhaps even partial or total destruction of the computer
facilities for example by fire, flood etc.
Day-to-day physical security of the data
The information stored in a computer system, which might well have been
collected over a number of years, may well be much more valuable to an
organisation then the equipment itself. The security and integrity of the
data is therefore of paramount importance and all systems should contain
some procedures to limit the possible consequences of destruction.
Excluding vandalism and fraud, which are discussed later, information
stored in the computer is under threat from a number of sources including:
• physical damage caused by faulty machinery, for example a defective
disk drive may physically damage the disk
• corruption of the information stored on the disk due to a faulty
program
• damage caused by lightning strike, power cut or electrical surge
rendering the data unreadable
• operator error causing incorrect data to be written to the file
• physical damage resulting from some outside cause. In the case of a
hard disk this may be fire, flood etc., but in the case of information
stored on floppy disks may be the result of exposure to extreme heat
e.g. sunlight through a window, use of the disk as a coffee mat etc., or
even loss of the disk itself
The most common method of ensuring day-to-day security of the data is to
take security 'back-up' copies on a regular basis. The frequency and extent
of the copies will depend upon the type of data and the frequency with
which it is usually updated, but most users of hard-disk systems for
example will take a complete 'dump' copy of the whole disk perhaps
weekly onto either floppy disk or magnetic cartridge tape, with regularly
used files perhaps being copied to floppy disk every day. Such a system is
known as a 'father/son' arrangement, and it ensures that information can
be recovered from a known point in time should any of the above
catastrophes occur. Note that with the father/son, or 'single generation'
copy arrangement, the copy tape is overwritten each time the information
is secured; both the original and the copy are therefore in the machine at
the same time, and if a problem occurs during the security operation then
both the master and the copy may be lost. The problem may be overcome
by using a separate security tape each time, but after a time the number of
tapes to be stored becomes unmanageable, and the method tends to be
very expensive. A common compromise is to employ either a two or three
generation technique where either two or three security tapes are used in
rotation. There are therefore always at least two previous copies of the
information available at any one time from which the system can be
restored.
Remember also to store back-up copies in a safe place away from the
computer room; it's not much good having back-up copies of the data
from which to restore the system in the event of fire in the computer room
if the back-up tapes are themselves destroyed in the fire!
Disaster recovery planning
Disaster recovery planning may be defined as the procedures necessary to

allow the organisation to survive partial or total destruction of its computing
facilities. It is an aspect of computing which has often been ignored in the
past, but the problem is that many companies have become slowly
dependent on their computing facilities without realising the implications
or the risk to the business if such facilities were either no longer available
or were seriously disrupted for a significant period of time. Brown (1987a)
claims that, in the financial services sector, computer failure may begin to
affect the business within hours although one to four days is typical, whilst
Donnelly (1986) reports that a survey carried out by Chubb Insurance
Group in America 'records that at least 90 per cent of companies affected
by computer disasters went into liquidation within 18 months'.
The extent to which organisations will need to plan for this type of event
will, of course, depend upon the size of the installation, and how import-
ant that organisation considers its computing facility to be. In the case of a
small firm using the kind of microcomputers available in the high street
stores it will probably be possible to simply replace damaged equipment
with new at very short notice, and provided that the hardware was
adequately insured and the data has been adequately preserved the system
should be operational again very quickly.
The problem becomes more acute as the installation becomes more
complex, and becomes a particular problem when the computer system
becomes centralised in one place, for example with a mini or mainframe
based installation in particular where specialised power supplies or air
conditioning systems are required. In this case a formal disaster recovery
plan is essential, and according to Brown (1987a) should include descrip-
tions of:
.. .facilities necessary to support the minimum processing requirements

after a disaster and a tested set of procedures for either the recovery of
computer systems or manual fallback.
There are a number of options available. At the lowest level one might
make some sort of reciprocal agreement with another company using the
same type of equipment, and this might at first sight appear a very cost
effective option. The major problem is likely to be that when disaster
strikes and you need to make use of the arrangement it may be very
difficult for the other company to make sufficient time available to meet
your needs.
Another option, which may reduce the total recovery time where the
complete computer room is destroyed, is the provision of an alternative
computer room ready equipped with the requisite power supplies and air
conditioning. This is known as a 'cold start' option and may well reduce
the recovery time to weeks rather than months. Other options are 'warm
start' where the stand-by computer room is ready equipped with a selected
'cut-down' version of the computer equipment which can provide a reduced
service within a few hours, or a 'hot start' facility where an outside firm
undertakes to make available a complete installation either immediately or
at very short notice. The use of such a system by P&O Containers Ltd
which provides a complete replica of its installation at eight hours notice
for a cost of £150,000 per annum (1987 prices) is described in Fraser
(1987), and Donnelly (1986) describes the use of a number of similar
systems costing between £5,000 and £200,000 (1986 prices) per year.
Brown (1987a-f} is a useful series of articles giving more in-depth
coverage of the peripheral issues concerned with disaster recovery planning.
Vandalism and fraud
Much has been made in the recent past of the need for computer systems
to be 'user-friendly'; to help users to get the best from the system.
Unfortunately this faoor may also make the systems very vulnerable to
access by unauthorised people and may therefore make them susceptible
to unauthorised alteration of the information which they contain, either
fraudulently or mischievously or both. Whilst the installation is small there
may not be a problem, but the problem assumes greater importance the
more complex the installation and the more widespread the operation
becomes. Norman (1983) classifies a large number of cases revealing the
possible effects of computer insecurity, and also discusses the ways in
which computer managers can minimise the risk involved. It appears
however that many users are unaware of the problem; according to David
Lindsay, Chairman of the British Computer Society security group, reported

in Ernest-Jones (1987):
Most people aren't aware of what security is and that it's needed unless
it hits them individually.
It is very difficult to attempt to quantify the extent of computer related
crime, because the most successful are either never caught or the victims
refuse to make the matter public. Some indication of the size of the
problem can be gathered from reports such as McTaggart (1986) which
reports an estimated loss to United Kingdom companies of at least £40m
per year, and Arthur (1987) which quotes a Confederation of British
Industry estimate of between £20m and £30m per year. Whilst the majority
of this loss can probably be accounted for through organisations such as
the major clearing banks and large public companies, a considerable
degree of computer fraud is undoubtedly committed at much lower levels,
and is a factor which must be considered by any organisation using
computers to handle financial transactions. It has been claimed (Computer
Weekly (1980)) that at least 80 per cent of detected computer fraud is the
result of opportunism
Computer crime isn't the striped jersey and mask type. It's rather like
leaving £5 notes around on the floor and not expecting people to touch
them.
(Brian Hitchin, computer security consultant)
Unauthorised access to computer systems may exist at any level for any
number of reasons. At the mainframe end of the scale, computers are
vulnerable to the kind of attack reported in Holdsworth (1987b), where a
group of 'hackers' based in West Germany successfully breached the
security of a worldwide computer network run by NASA, and claimed to
have examined data stored in 135 different sites, including two in the
United Kingdom, over a six-month period. Closer to home is the case of
the unauthorised access to the Duke of Edinburgh's Prestel mailbox reported
in Holdsworth (1987b). This case also shows the legal problems involved,
since although journalists Schifreen and Gold admitted breaking into the
system and were initially found guilty of electronic forgery under the 1981
Forgery and Counterfeiting Act, the decision was overruled on appeal by
the Lord Chief Justice Lord Lane and the pair were finally acquitted by the
House of Lords. On a more domestic level McTaggart (1986) reports the
case of a Manchester based building materials supplier which, in the early
1980s operated at a comfortable profit, but by 1984 follo»'ing the intro-
duction of a computer based stock control and invoicing system was
operating at a considerable loss. In this case the problem was found to be
unauthorised alteration by counter staff of the prices in the system. Goods
were then supplied to accomplices in the trade at below cost price.
On the larger machines the security and integrity issues are, of course,
taken very seriously by the manufacturers and suppliers, and it is perhaps
microcomputers, and particularly microcomputers connected in networks,
which are most susceptible to the opportunist. Kochanski (1986) sums
up the problem:
Microcomputers are a familiar sight in an increasing number of offices;
but it's still remarkable how little thought is given by the average user to
the sensitivity of information held on the system. People who would not
dream of leaving filing cabinets unlocked have no objection to leaving
computers accessible, with exactly the same information in them.
Kochanski also summarises the main threats to data security and integrity:
First, detectable alteration, where someone erases a report that took days
to prepare; second, undetectable alteration, such as adding false invoices
to the purchase ledger; third, detectable reading, where a disk containing
the customer mailing list goes missing; and fourth, undetectable reading,
where a maintenance engineer makes a copy of the file containing next
year's business plan and takes it with him.
The solution to some of these problems relies on physical security. Floppy
disks, and back-up copies, containing sensitive data should be locked
away, and it is possible to fit the microcomputers and terminals themselves
with a Yale type lock to prevent unauthorised usage.
At the software level the most common protection is the 'password',
some secret combination of characters which must be entered before the
computer will allow access to particular systems or data files. There are
however a number of factors which should be considered if passwords are
to remain effective:
(i) Passwords must, of course, be meaningful to the user if they are to be
easily remembered, but they must be sufficiently obscure that they
cannot easily be guessed. Passwords such as 'FRED', the user's name
or 'chairman' represent no protection at all.
(ii) Passwords must be changed often; the longer a password stays
unchanged the greater the number of people likely to learn it.
(iii) Passwords must be kept confidential to those who need to know
them and this will normally mean that they must not be written
down; stories of passwords being written on the side of terminals are
commonplace, and one computer security adviser is reported in
Ernest-Jones (1987) as knowing of a bank which has a notice in its
payment room saying 'Today's password is... .'
(iv) It is little use having users password-protect their files if the operating
system passwords are left as set by the supplier. Access at the
operating system level will often allow hackers to read a complete

list of user passwords, and it has been estimated (Holdsworth (1987b))
that probably 30 per cent of users leave the manufacturer's password
on their system.
Although passwords are by far the most commonly used security device
there is widespread belief among computer professionals that they are
inadequate for all but the lowest level of security. Other methods which
are now beginning to be used include:
(i) file encryption systems, where the contents of a file are 'coded' and
can only be unlocked with the correct 'key'. One such system is
described in Kochanski (1986).
(ii) 'keyboard dynamics' where the computer operating system establishes
and monitors a user's identity by understanding the unique way in
which that user taps the keys.
(iii) the use of 'mandatory' operating systems, described in Holdworth
(1987a), where the operating system monitors users and decides
whether or not to give access on the basis of a profile of each user
containing details of the type of information he is allowed to access,
the time of day during which he usually works etc.
(iv) the development of automatic 'auditing' routines which record every-
thing which happened inside the computer.
Finally a reminder that computer security does not end even when you sell
the machine. Barrett (1987a) reports the case of a student who purchased a
second hand computer from an army-surplus store, and later found that he
had got 'about 1,500 sensitive files' from the Royal Signals and Radar
Establishment, Malvern, including lists of staff, departmental costings, lists
of equipment and correspondence with suppliers.
Who will operate the system~
Arrangements to actually use the system will, of course depend upon the
software to be installed. In the case of a single-user system which will only
be used by one person, for example a word processing system, the answer
is simple, but the question requires more thought in the case of a multi-
user system or a single-user system which is to be used by a number of
different people. In this case it is essential to have some one person in
control and responsible for the day-to-day running of the computer,
including backing up data in accordance with the procedures laid down in
the operations manual, ensuring that printers are supplied with paper and
ribbons, maintaining stocks of consumables such a floppy disks, liaison
with maintenance contractors etc. A most important function of the
computer operator in this situation is that of supporting other users,

particularly those who are new or those who only have a need to use the
computer occasionally and may therefore need help and guidance.
In the case of a small installation using 'off-the-shelf' software, ex-
perience seems to show that the operator need not necessarily have
specialist computer training, and the size and complexity of the organ-
isation will determine whether or not the supervision of the computer is a
separate full-time job. It should however be remembered that some arrange-
ment needs to be made to cover holiday periods, sickness etc. and there
should therefore be at least one more employee who can take over at short
notice should the need arise.
What training will be required?
Training is one of the most difficult areas to plan satisfactorily, and yet
adequate staff training is essential if an organisation is to get the best from
its computer investment. Experience also appears to show that this is
probably the one area above all others in which management try to
economise. Powell (1987), reporting on research carried out in 1985,
states that on average, for a new system, management allocated only one
day to training and this tended to be largely on a seminar consisting of a
demonstration of the system. Even when more comprehensive training
programmes are available the result is not always totally satisfactory.
Andrew (1987) examines the question of training in a number of quantity
surveying practices which had recently introduced a computerised billing
system, and found that although the firms were not disappointed with the
quality of the training offered, in fact the three-day programmes themselves
were rated by many as very good, nonetheless:
The survey revealed that not all firms feel that they have achieved
extensive utilisation of the system, and that many who have achieved
this level have only done so after a lengthier period than originally
anticipated.
There is an instinctive feeling that the best way to learn to use computers is
by trial and error, and there is no doubt that it is possible to learn this way,
provided that management does not expect the system to be in operational
use at the same time. The basic mistake which is made over and over again
is to expect that a computer system will be working at full capacity almost
within hours of installation. This hardly ever happens, and whilst manage-
ment might well be unwilling to spend money on formal training courses it
must nonetheless be prepared to allow staff the opportunity to gain familiarity
with new systems without the pressures imposed in a production
environment. Adequate allowance must then be included in the castings
for a new computer system to allow for the inevitable loss of productivity
which will occur during the familiarisation process.
Although trial and error can result in very good appreciation of the
workings of a computer system, it is essentially a random process which
relies to a great extent on the availability of a comprehensive manual from
which to learn; something which many computer systems lack. In addition
the time taken to achieve familiarity with a system on a trial-and-error basis
will vary widely from one person to another due to the amount of time
likely to be spent in following 'blind alleys'. The contribution which can
be made by a formal training course in terms of controlled training over a
known period of time should not be overlooked.
Computer-aided training courses, in which the computer itself leads the
user through a series of tutorials, are becoming more popular as an
alternative to organised group training sessions. Whilst they do have
advantages in terms of allowing users to learn at their own pace, they also
have significant disadvantages in that their ability to answer queries raised
by the user during the session are very limited. The absence of the
stimulation which results from the inter-personal contact of a well run
group training session is also very difficult to overcome.
Hunt (1987) has much good advice about the training of quantity
surveyors in the use of computer techniques.
What accommodation will be requiredJ
Regarding accommodation, the question of where to put the new toy is

always a thorny one, and the best solution for any particular organisation
will depend upon the particular computing philosophy adopted and the
type of computer system installed.
Individual microcomputers working either as stand-alone machines or
connected in a network are perhaps the most flexible. Modern micro
computers are very tolerant of widely varying environmental conditions,
and may therefore be placed almost anywhere within a normal office
within reach of an electrical socket, although one should be wary of
enclosed alcoves with poor ventilation where over-heating could occur.
This flexibility is of course a great help in the development of the kind of
user friendly systems now in demand; the executive can have his computer
on his desk, and units can be easily positioned at key points around the
office for those less fortunate who have to share, but there are other factors
to consider, for example:
(i) Microcomputers should be carefully placed to avoid the risk of
damage to the machine from knocks, spilled drinks etc.
(ii) Screens need careful positioning to avoid annoying reflections from

lights or windows.
(iii) Modern microcomputers are very portable and therefore easily
stolen; one therefore needs to make sure they are adequately protected
outside office hours.
(iv) Remember that the users will need space around the machine to
spread out documents etc.
(v) The need for security of the systems, confidentiality of information
etc. might lead to the need to place computers in a separate room
and to restrict access.
(vi) Some types of printer, keyboard and even some computers, are very
noisy and may distract others working in the same room.
(vii) Experience seems to show that central computing facilities, a mini-
computer or a network server for example, are best placed in a
separate room to which access can be restricted if required.
(viii)Aithough modern computers are very tolerant, do not overlook the
possibility that a 'clean' power supply may be required, especially if
there is any equipment connected to the circuit which may cause
'spikes' to occur, for example some types of electric motor.
Should future expansion be consideredl
Yes, but it is difficult to look further ahead than you can see. The planned
operational life of a modern microcomputer system should be taken as no
more than about five years; this is not to say that the equipment will be
worn out after this length of time, but the technology is advancing at such
a rate that systems installed today will be obsolete within this period of
time. It is therefore pointless to attempt to plan too far ahead.
It is however very tempting, following the successful implementation of
one system, to attempt to add more and more to the original system. This
temptation should be viewed very carefully; new applications should be
subject to just as searching a scrutiny as was the original in order to assess
the cost effectiveness of the proposal. In addition the impact of the new
applications on the existing systems must also be carefully considered if
degradation of the existing service is to be avoided.
What about the human factorsl
Much has been written over the years regarding the human aspects of
computing, both from the point of view of human/machine interaction and
also with regard to the health and safety of the operators.
Considerable research is presently being carried out to investigate the

most effective methods of human/computer interaction, and it has been
claimed, see for example Clark (1987a), that many systems operate at only
20 per cent of their potential efficiency because of poor human factor
design. The research is however still in its infancy, and it is likely to be a
number of years before the results of the latest research initiatives are
applied to office systems.
Winfield (1986) offers an excellent introduction to the concepts involved
in the human/computer interface problem, including discussions on human
psychology and systems design, management of computer projects, com-
puter aided learning and the effects of computerisation on social change.
Much more emotive is the subject of the health and safety of the operator,
and concern has been expressed over a number of years about a whole
range of illnesses allegedly suffered by the operators of visual-display units
(VDUs). Issues of concern have included, amongst others, epileptic fits,
radiation, eye problems, effects upon pregnant women, headaches, facial
dermatitis, hypertension, repetitive strain injuries and general fatigue. Whilst
there have been many studies conducted over the years, none appears to
have proved conclusively that the use of VDUs poses any specific medical
hazard, although some scientists claim that there is evidence which seems to
show that some weak electromagnetic fields similar to those which leak
from computers and television sets can trigger allergic symptoms in some
subjects including headaches, nausea, sweating, pallor and indigestion.
More specifically a survey of over 4 000 staff carried out by the North East
London Polytechnic for Newham Borough Council and the local authority
trade union NALGO in 1987 (reported in Clark (1987b)) clearly shows that
VDU users take more sick leave than non-users, the commonest symptoms
being given as muscular aches and pains, eye problems, headaches,
tiredness, dizziness and fainting. It is therefore apparent that management
must tackle this problem if the introduction of computers is to be successful.
Huws (1987) contains much of interest in this area, and she makes
the pointthat many of the medical problems allegedly caused by computers
can be prevented by careful arrangement of the workplace itself. At the
ergonomic level, it is vitally important that the user should be comfortable.
Many eye problems can be avoided if screens are easy to read with good
adjustable contrast and brightness control, adjustable tilt, and are carefully
positioned to avoid annoying reflections and glare. Eye problems can also
be reduced by limiting the periods of time which a user actually spends
looking at a screen. Various limits have been proposed, but a maximum of
about one hour at a time followed by at least a fifteen-minute break seems
to be an acceptable average, although the report on the Newham survey
also recommended that users should spend no more than 50 per cent
of their total working time actually working at a VDU. Muscular aches and
pains can similarly be eased by the provision of suitable adjustable chairs.
On a psychological level there seems to be some evidence in the
Newham survey to support the view that poor job satisfaction often
manifests itself in apparent physical symptoms of tiredness, hypertension
and irritability. There is also evidence that workers feel that in some cases
the computer dictates the pace of the work, that the machine is pushing
them along, that they have in some sense lost control, and that this may
also give rise to a high incidence of hypertension and similar stress-related
illness.
Psychological problems such as these are very complex and in part
may impinge on the human/computer interaction issues discussed earlier.
Management must however involve staff at all stages of computerisation if
feelings of insecurity are to be avoided, and it may be that the kind of
stress-related problems mentioned above can be reduced if workers are
aware of, and in agreement with, the likely impact on their role from the
outset and are actively involved in the whole process from feasibility to
implementation.
A satisfactory resolution of the human issues should perhaps be the
most important consideration of any organisation planning to introduce
computers. The process will be much more successful and the systems will
run much more smoothly if staff are committed to making it so.
References
Andrew J. (1987) Computerising and the need for a training programme Disse rtation
submitted for BSc(Hons) in Quantity Surveying, Trent Polytechnic
Arthur C. (1987) 'Fraud inspires CBI to act' Computer Weekly 22/1/87
Atkin B. (1987) 'Computerising' Chartered Quantity Surveyor March 1987 p. 10
Barrett D. (1987a) 'Secrets sold for £45' Computer Weekly 26/3/87
Barrett D. (1987b) 'Head scratching stuff Computer Weekly 24/9/87 p. 38
Brown J. (1987a) 'When it's more than a movie' Computer Weekly 14/5/87 p. 30
Brown J. (1987b) 'How to assess the impact of disasters' Computer Weekly 28/5/87 p. 31
Brown j. (1987c) 'Fleeing the grim reaper's grasp' Computer Weekly4/6/87
Brown J. (1987d) 'Don't panic as disaster falls' Computer Weekly 18/6/87 p. 17
Brown). (1987e) 'The best laid plans.. .' Computer Weekly 25/6/87 p. 30-31
Brown J. (19870 'Prevention- as good as a cure' Computer Weekly 16/7/87
Carter D.J. (1984) 'Who checks the softwarel' Construction Computing October 1984p. 32
CICA (1981) Chartered Quantity Surveyors and the microcomputer RICS
ClarkS. (1987a) 'Pilot highlights human factors' Computer Weekly7/5/87
ClarkS. (1987b) 'Are you sitting comfortablyl' Computer Weekly 1/10/87
Computer Weekly (1980) '80% of DP frauds result of 'pure opportunismH Computer
Weekly 6/11/80
Donnelly F. {1986) 'Coping with chaos' lnfomatics August 1986 pp. 26--28
DOC (1980) Miaos in construction Design Office Consortium (now Construction
Industry Computing Association)
Ernest-jones T. (1987) 'Why network security needs tightening up' Computer Weekly
26/11/87 p. 11
Fraser G. (1987) 'Making a full recovery' lnfomatics December 1987 pp. 51-52
Holdsworth I. (1987a) 'Working the system' Computer Weekly 23/7/87 p. 20
Holdsworth I. (1987b) 'Hackers' spree spans Europe' Computer Weekly 24/9/87 p. 9
Hunt G. (1982a) 'Micronotes: A question of verification' Chartered Quantity Surveyor
March 1982 p. 235
Hunt G. (1982b) 'Micronotes: Halve labour costs - don't use computers' Chartered
Quantity Surveyor April 1982 p. 275
Hunt G. (1987) 'Micronotes: Get in training' Chartered Quantity Surveyor February
1987 p. 12
Huws U. (1987) The VDU worker's handbook London Hazard !.entre, 103 Borough
Rd., London SE1
Kochanski M. (1986) 'Locking up the data' lnfomatics August 1986 pp. 44-45
Lawrence j. (1986) 'Rent, buy, lease - a beginners guide' Construction Computing
january 1986 p. 8
McTaggart T. (1986) 'Are your security measures up to scratch?' Computer News 13/
11/86
Mulholland KA. (1986) 'finding the right software supplier' Construction Computing
Winter 1986 pp. 15-16
Norman A.R.D. (1983) Computer Insecurity Chapman and Hall
Parrett K. & Stead D. 'Computers: who needs them?' (1986) Construction Computing
Autumn 1986 pp. 34-41
Powell M. (1987) 'Training is all Greek to some' Computer Weekly 28/5/87
Winfield I. (1986) Human resources and computing Heinemann
4 Computers and the Law
Objectives

• Outline some of the more important provisions of the principal legisla-
tion concerned with the use of computers in a commercial environment
including:
(i) copyright and similar forms of protection
(ii) the Data Protection Act (1984)
• Discuss the rights and liabilities usually conferred upon the users of 'off
the shelf' software packages.
• Discuss the liability for errors which might arise from the use of
software packages whether bought 'off the shelf' or developed in-
house.
• Discuss the general legal issues surrounding contracts for the purchase
of computer hardware and software.
Introduction
This chapter is intended to provide an outline overview of some of the

more important legal issues which concern the use of those computer
systems likely to be used by a quantity surveyor. More detailed considera-
tion of the issues raised will be found in the publications referenced in the
text. The law concerning computers tends to vary from country to country
and is constantly changing; these notes consider the situation as it currently
applies in the United Kingdom at the time of writing (mid 1988).
The legal protection of computer software
The law concerning protection of computer software is very confused,

essentially because most of the underlying legal principles were developed
either to protect authors and artists from unauthorised copying of literary or

artistic works (copyright) or alternatively to protect inventors against unfair
exploitation of their inventions by others (patents). The advent of computer
programs which might exist only as zones of magnetism on a tape or disk,
or even worse within the body of a 'chip' was, of course, quite unforeseen
and the application of existing copyright and patent law has therefore
proved very difficult. To add to the confusion, although many countries
include similar legal principles covering copyright and patents within their
own legal systems, there have on occasion been significant variations in
the ways in which those principles have been applied, thus ensuring that
the same piece of software may receive different levels of protection in
different parts of the world. This situation has become increasingly import-
ant as communication between computers across the world has become
more commonplace.
In general terms Brett & Perry (1981) identify four areas of legal rights
which might, in theory at least, be used to provide some protection against
copying and the unauthorised use of software. The areas are given as
patent law, copyright law, the law relating to 'trade secrets' and 'unfair
competition', but Brett and Perry admit that, of these four possibilities, the
two most likely areas are those relating to copyright and trade secrets. A
good basic review of the relevant legal principles is given in Nicolle
(1987).
Copyright
The law relating to copyright developed largely during the eighteenth

century to protect the rights of authors against unauthorised copying
following the widespread use of printing. The law has, at various times,
been extended to cover other written material for example shorthand notes
(Pitman v Hine (1884)), football pools coupons (Ladbroke v Hill (1964))
and railway timetables (Blacklock v Pearson (1915)). The present Acts also
cover sound recordings, films and design drawings.
The current law is enshrined in the Copyright Act (1956), as amended by
the Design Copyright Act (1968) and the Copyright (Computer Software)
Amendment Act (1985), and provides protection both during the author's
lifetime and for 50 years after death. Whilst, as Wilson (1981) points out,
the Copyright Acts make no specific mention of tapes, punched cards or
transient displays as being among the possible forms which a literary work
might take, Brett & Perry (1981) argue that since the definition of literary
work has been extended to cover shorthand code there would seem to be
little doubt that, in principle at least, computer programs stored on
punched card or paper tape would come within the definition although
programs stored on magnetic media might pose a greater problem. Written
Computers and the Law 67
descriptions of the programs, operating instructions, user manuals and

similar documents will, of course, be covered by copyright legislation.
Many computer software authors and software supply companies have
long been concerned about the apparent ease with which software 'pirates'
have been able to circumvent what was perceived to be the spirit of the
law. losses to UK software companies caused by software piracy have
been estimated to be as much as £150,000,000 per annum and an
intensive campaign for more definitive legislation has been conducted by
many interested parties, among the most active of which is an organisation
of software suppliers known as the Federation Against Software Theft
(FAST). The FAST initiative has apparently begun to have some effect;
Nicolle (1988) reports on the computer industry reaction to the new
Copyright Designs and Patents Bill presently passing through Parliament,
and Nicolle (1987) reports on a number of successful prosecutions for
software piracy. Ernest-Jones (1988) reports the first ever custodial sentence
for software piracy passed on a software distributor in late 1987 after he
was found guilty of forgery, copyright infringement, the illegal importation
of software and offences against the Trades Descriptions Act.
Even if it can be assumed that copyright protection exists practical
problems remain. One problem is that the intention of the copyright
system is to protect the skill and care of the author in expressing a
particular idea in literary form; there is no attempt, or wish, to provide
protection for ideas themselves although as Wilson (1981) points out it is
sometimes very difficult to draw the line between the tw.o, and this might
particularly apply in the case of computer software where one person
might write a program using an algorithm developed by another which
produces identical results but uses a different language.
A second problem is that of proof. Before a plaintiff can bring a successful
action for breach of copyright he must prove that an infringement has
actually occurred, and this can itself prove to be very difficult. By far the
most widespread abuse of computer software is associated with the large-
scale illegal copying of well known and well used 'industry standard'
programs. Illegal copies are often made in the Far East and may well be
provided to unsuspecting UK customers complete with professionally
reproduced manuals in such a way that the total package is indistinguish-
able from the genuine article. Most major software authors however
maintain a register of purchasers, and it is only when an unregistered
owner of a pi rated package contacts the software author for support advice
that the real nature of the software becomes clear.
In law the user of pirated software could probably be sued by the
copyright owner, but software suppliers have much more interest in
catching the software pirate than in pursuing unwitting customers. One
possible method which has apparently met with some success is reported
in Nicolle (1988). In this case Ashton-Tate, authors of the dBase range of
programs has offered all illegal users of their packages an immunity from
prosecution and the right, for a small fee, to become legitimate and to
receive future updates in return for naming the source of their programs.
Such methods will plainly only work if users are aware that the version
they are using is in fact illegal.
A further area of concern to computer users is the unauthorised copying
and use of the data stored within a computer system rather than the
programs themselves. Whilst Nicolle (1988) reports that data bases and
programs created by artificial intelligence will be covered by the proposed
new Copyright Designs and Patents Bill, the situation under present law is
far from clear-cut, and in any event the burden of proof, particularly for
files such as databases, may present a serious problem. It has been
suggested that users might 'mark' their files with some 'dummy' informa-
tion specifically to allow a positive identification in the event of proof of
theft being required. The position with regard to data is also complicated
by the fact that it was held in Oxford v Moss (1979) that pure 'information'
does not fall within the definition of 'property' which may be subject to
theft under the criminal law. Information alone therefore cannot be stolen.
Perhaps the best advice is to introduce and maintain suitable security
systems to prevent the theft of such information in the first place.
Trade secrets
In addition to any rights of action under the copyright laws, rights of action
may also apply under the law relating to trade secrets. This branch of the
law is not primarily based in statute but is instead based on that principle
of equity which attempts to protect relationships of confidence. The law
will therefore act to prevent or restrict one party from taking unfair
advantage of the relationship by preventing him from using confidential
information in an improper way. In the words of lord Denning, the law is
essentially based on:
....the broad principle of equity that he who has received information in
confidence shall not take unfair advantage of it. He must not use it to the
prejudice of him who gave it without obtaining his consent.
(Seager v Copydex (1967))
The law Commission in 1974 recognised three factors to be considered in
any particular case:
The nature of the infonnation
The information must be confidential and must be seen to be so. The

courts seem prepared to take a wide view; in the words of Megarry J:
If the circumstances are such that a reasonable man standing in the

shoes of the recipient of the information would have realised that upon
reasonable grounds the information was being given to him in confidence
then this should suffice.
(Coco v A.N. Clark (Engineers) Ltd. (1969))
The relationship between the parties
It was established in Saltman Engineering v. Campbell Engineering Ltd.

(1948) that a contractual relationship is not essential, although it was
held in Lamb v. Evans (1893) that if a contract does exist it will be held
to contain implied terms equivalent to the equitable obligation. Whilst
the dictum of Lord Denning in Seager v. Copydex appears to imply a
direct relationship of this kind, Wilson (1981) points out that in Prince
Albert v. Strange (1849) the plaintiff was held to be entitled to an injundion
against a third party who was the recipient of confidential information
initially given by Prince Albert to another, and a similar decision was
reached in Argyl v. Argyl (1969). The crucial point in these cases seems to
be that the third party involved knew that the information was confidential
and that he should not have had access to it, but consider the case of an
employee who copies confidential information belonging to his employer
and subsequently sells the information as his own to an innocent third
party. Would the employer have a right of adion against the third party?
Although such a scenario has yet to be specifically decided by an
English court, in Seager v. Copydex (No. 2) (1969) the defendants were
held to be liable even though they believed their product to be a result
of their own development and not to be based on information obtained
from the plaintiff. Despite this decision, it seems unlikely that a defendant
could be accused of breach of trust if he had no knowledge that the
information in question was confidential. Wilson (1981) points out that it
has been accepted in the USA that once the third party has been made
aware of the the confidence then he becomes bound by it and must not
disclose the information to others. He may however be permitted to
continue to use the information himself if he has substantially changed
his position as a result of it, for example, if he has installed new plant in
order to exploit it.
Misuse of the information
Any unauthorised use of confidential information for profit is a misuse. It

is however for the plaintiff to prove that confidential information has
been misused, either by the defendant directly or by some third party
and this may be extremely difficult to do particularly in the case of
computer software and data.
Remedies for breach of copyright or unauthorised use of confidential

information
Breach of copyright is a criminal offence, and in addition a number of

remedies are also available under the civil law.
The principal remedies in the case of computer software or data will
be either an injunction, damages or an order for delivery up and destruction
of the relevant documents, tapes, disks etc. The Court may also make an
order for the preparation of an account of profits. Damages may be
awarded and will normally be based on the market value of the informa-
tion rather than the loss actually suffered by the plaintiff. The court
however has considerable discretion in deciding the level of damages
appropriate in each particular case.
The Data Protection Act (1984)
One particular kind of computerised information the confidentiality of

which is often specifically protected by law is that from which an indi-
vidual may be identified. Such information is regulated in the United
Kingdom by the Data Protection Act (1984). The United Kingdom is not
alone in this area, indeed one of the main reasons for the introduction of
the Data Protection Act was to comply with the Council of Europe Data
Protection Convention first opened for signature in 1981. The majority of
European and Scandinavian countries either already have or are in the
process of enacting similar legislation, and outside Europe legislation
exists in the USA, Canada, Australia and New Zealand. It is typical of
the British parliamentary process that the need for some form of protection
for personal data was recognised as long ago as 1969, when Kenneth
Baker introduced his Data Surveillance Bill in the House of Commons.
The prime purpose of the Data Protection Act is to regulate the storage
and communication of personal information stored in computer systems.
It does not relate to personal information stored or manipulated in any
other form.
In principle the Act requires users of data (termed 'data users' in the
Act) which might identify particular individuals (termed 'data subjects')
and which is stored in computer systems for whatever purpose to register
with a Government-appointed Data Protection Registrar. The Act allows
certain exemptions from either the whole or parts of the Act for certain
classes of data which do not pose a particular threat to the privacy of
data subjects, or for data collected in connection with national security,
the administration of justice, the detection of crime or for tax collection.

In terms of normal business activity, data relating to payroll and pensions,
and name and address files used only for distribution purposes are
exempt from registration, and certain exemptions also apply to data
stored purely for statistical or research purposes. The quantity of personal
information covered by such exemptions is considerable, and the possible
extent of uncontrolled access to such data has given grounds for concern;
Computer Weekly (1988) reports on a motion in the House of Commons
which claimed that MIS can access information about any adult in the
United Kingdom through the Department of Health and Social Security
computer centre, that the electoral roll is used by police and immigration
officials, can be sold to any organisation, and that such activities are not
regulated by the Act.
All registrations made under the Act are required to give:
(i) the name and address of the data user
(ii) a description ofthe personal data held
(iii) a statement of the purpose for which the data is held
(iv) a description of the sources of the data
(v) persons to whom the data may be disclosed
(vi) countries to which the data may be sent
(vii) an address for receipt of requests from data subjects for access to the
data
The register is maintained by the Registrar and is available for public
examination. Failure to register, or to keep a registration· up to date
constitutes a criminal offence and will result in prosecution. The Registrar
has a number of investigators under his direct control, and if he suspects a
breach of the Act may seek access to a data user's premises. If access iS
denied then the Registrar may apply for a warrant empowering him to
enter, search and seize discs, tapes, documents etc. Kavanagh (1987)
reports the first ever such raid on the home of a Staffordshire special
constable in October 1987.
The Registrar also has the power to refuse an application for registration,
and data users who then continue to process the information commit a
criminal offence.
Data subjects have the right to a copy of any personal data held by any
data user covered by the Act, and data users are entitled to make a small
charge to cover the processing costs involved. Data subjects also have the
right to correction of incorrect or inaccurate data, and also a right to
damages for any loss directly attributable to either the use of incorrect data
or the loss or unauthorised disclosure of the information to third parties.
The Act is underpinned by eight 'data protection principles':

(i) Personal data shall be obtained and processed fairly and lawfully
(ii) Personal data shall be held only for one or more specified and
lawful purposes.
(iii) Personal data held for any purpose shall not be used or disclosed in
a manner incompatible with those purposes.
(iv) Personal data held for any purpose shall be adequate, relevant and
not excessive in relation to that purpose.
(v) Personal data shall be accurate and, where necessary, kept up to
date.
(vi) Personal data held for any purpose shall not be kept for any longer
than is necessary for that purpose.
(vii) Individuals shall be allowed rights of access to personal data.
(viii)Appropriate security measures shall be taken against unauthorised
access, alteration, disclosure or destruction of personal data.
More detailed consideration of the Data Protection Act may be found
in Savage & Edwards (1984) and Sterling (1984).
Contracts for the supply of computer systems
All of the normal rules of contract law governing offer, acceptance and
consideration apply, of course, to contracts for computer systems, but there
are a few additional pitfalls which might arise because of the typical
quantity surveyor's lack of familiarity with the details of computer systems.
Sellors (1984) gives a good introduction to the problems inherent in the
successful negotiation of contracts for computer systems, and Gooch
(1986) includes detailed comment on the successful selection and purchase
of computer systems for small contractors. The major potential problem
areas which must be specifically addressed during the negotiation and
contracting process have been identified by Bigelow & Nycum (1975) as:
Basic requirements
The actual functions which the system is required to perform must be

defined in as much detail as possible. In particular requirements such
as the required standards of accuracy, anticipated volumes and formats of
input and output, acceptable timescales for each of the various processes,
implementation period and training programme etc. must be specified as
accurately as is reasonably possible. It is essential for all parties involved to
have the the same expectations of system performance if future disagree-
ments are to be avoided.
Responsibilities of the parties involved
All parties must be absolutely clear about their respective responsibilities.

It is fundamentally important to the successful implementation of any
computer system that the client, user, hardware supplier, software supplier,
systems analyst and any other specialists involved understand their obliga-
tions and responsibilities to each other, and that these obligations and
responsibilities are written down and mutually agreed. It is equally import-
ant that one party is responsible for managing and co-ordinating the whole
process, and that all of the other parties recognise his authority.
System specifications for hardware and software
The system specifications must be settled before any contracts are signed.
Details may be based upon some sort of performance specification, but in
this case it is important to specify as precisely as possible the circum-
stances under which that optimum performance is to be achieved. Although
this method of evaluation is widely used, for example for bill production
systems, such systems are often purchased on the basis of their performance
in one particular user's office, often the firm which developed the software
in the first place. There is no guarantee that because a system performs
adequately in one environment it will also do so in another; what is
perfectly satisfactory to one office or firm may be quite unacceptable to
another. It is equally important, particularly for systems which mix hard-
ware from one supplier with software from another, to establish exactly
who is responsible for ensuring that the system as a whole meets the
specified criteria. Noise control may also be a problem, and maximum
permissible levels should be stated where appropriate.
Site preparation
Ambient operating conditions for the equipment must be established as

early as possible and arrangements must be made for the provision of any
necessary air-conditioning, clean power supplies etc. Arrangements must
also be made to acquire any necessary specialist furniture etc. required.
Any necessary security precautions should also be implemented to ensure
the integrity of the machine and the data.
Training
Particular attention should be paid to training, particularly who should be

trained to administer the system, and who to use the system. The amount
of training required and the cost must be established and included in the
contract price.
Delivery, installation and acceptance
Detailed schedules of delivery dates must be agreed, together with respons-

ibility for the installation and testing of the equipment. The contrad must
also include details of the tests which must be satisfied both by each
component individually and also by the system as a whole before the
system is accepted.
Financing
The time and method of payment must be agreed, including the cost
of 'extras' such as additional training days, maintenance etc. lawrence.
(1986) gives a good introduction to possible financing options.
Maintenance
Establish the necessary maintenance procedures for both the hardware and
the software (see chapter 3). Do not forget to make some provision to cater
for catastrophic failure.
Modifications and upgrades
Contracts should include some provision for possible upgrades to either

software or hardware or both. A system should be devised for reporting
software faults, and a mechanism should also exist for notification by the
supplier of known 'bugs' together with corrections as required.
Warranties
Warranties must be carefully detailed in the contract. Many manufacturers

of small computers offer warranties as short as 40 days from the date of
delivery, whereas suppliers may well extend the warranty period to 12
months. Pay particular attention to the warranties in the case of systems
built up of equipment manufactured by a number of different firms, and
ensure that someone provides an acceptable warranty on the system as a
whole. Warranties on software may be especially restridive, and this
aspect is discussed in more detail below.
The purchase of computer software
It is important to recognise that, in most cases, the purchase of a software

package does not confer any ownership of the software itself, merely a
licence to use it. Most software packages purchased 'off the shelf' are
supplied shrink-wrapped in plastic, often with the licence agreement
prominently displayed, but sometimes there will only be a notice on the
outside of the package stating that, by opening the package, the user
accepts and becomes bound by the terms and conditions of the licence
which are set out in full inside. Such licence agreements will attempt to
limit things like the extent to which the software may be copied, the use of
the software and the liability of the vendor in respect of errors.
The following clauses are not extracted from any particular licence
agreement, but are typical of those which may be included in a package
designed to run on a microcomputer.
You should read carefully the following terms and conditions before
opening this package. By opening this sealed package you agree to
become bound by all of the terms of this agreement. If you do not wish
to become so bound then return this package unopened to your dealer
for a full refund.
This clause invariably appears either as the first clause of the agreement,
usually printed in bold type in capital letters, or as a notice prominently
displayed on the outside of the package. It attempts· to establish the
purchaser's acceptance of the licence agreement as an implied term in the
contract of sale. There is, at present, considerable concern about the
effectiveness of these so called 'shrink-wrapped' licences, and a number of
lawyers have expressed doubt as to whether they could, in fact, be
enforced. The question has not yet come before an English Court.
This software is licensed for use on a single computer (that is a computer
with a single cpu) at a single location provided that you abide by the
terms of this licence agreement. The software may be moved from one
machine and physically transferred to another machine but shall not be
used concurrently on more than one machine. If you wish to operate this
software on a network then please contact the authors to arrange a site
or multi-user licence.
This clause establishes the licencee's right to use the software but only on
a single computer. One of the major difficulties faced by users who wish to
implement the same software package on a number of machines is that,
under this provision of the licence agreement, a separate copy must be
purchased for each machine. There is the provision for negotiation of a site
or multi-user licence, but this is often not much cheaper than buying
multiple copies of the software. Note also that the licence is only granted
whilst the user abides by its terms, and the implication is that the licence
is automatically revoked in the event of a breach.
This software and the accompanying written materials are the subject of
copyright, and the copying of the software or of the written materials is
strictly forbidden except as expressly provided herein. You may make
one copy of the software solely for back-up purposes provided that the
copyright notice is included on the back-up copy.
Copyright in the material is established, but note that the agreement does
allow copying of the software but only for back-up purposes. Note that this
provision will not be of any value if the original disk is copy-protected.
You may not transfer the licence to use software to anyone without the
prior written consent of the copyright holder. Any authorised transferee
shall agree to be bound by the terms of this agreement.
Software authors attempt to keep track of the authorised users of their
packages by maintaining a register. The aim is to enable the users of
'pirated' software to be discovered, but the efficiency of this method must
be very questionable.
This software and the accompanying written materials including the
instructions for use are provided 'as is' without warranty of any kind.
The authors do not warrant, guarantee or make any representation
regarding the use, or fitness for purpose, or the accuracy of the results
obtained from the use of the software. The entire ris'< as to the results or
performance of the software is assumed by the licensee, and if the
software is defective the licensee will take full responsibility for and
assume the full cost of any repair or corrective action. Neither the
authors nor anyone else involved in the creation, production or delivery
of this product shall be liable for any direct, indirect, consequential or
incidental damages arising from the use or abuse of this product.
The authors attempt to exclude themselves from any liability whatsoever.
Note the attempt to make the customer liable not only for the con-
sequences of the use of the software, but also for the consequences where
the software is defective and even for remedying defects in the software
itself! This provision could have extremely serious consequences particu-
larly, for example, in the case of programs for structural design.
Fortunately, it is extremely doubtful whether such a clause would be
upheld by a Court of Law.
Errors arising from the use of computer software
As illustrated above the authors and vendors of most off the shelf software
packages attempt to severely limit their liability for the effects of 'bugs' in
the product. There is presently considerable argument about whether or
not software of this kind is governed by consumer protection legislation
and, more important, the extent to which it is subject to the Sale of Goods
Acts and the Consumer Protection Act (1988) (see Ernest-Jones (1987),
Thomas (1987) and Arthur, Mehta & Ernest-Jones (1988)). The basic problem
centres around whether or not the software can be classified as a 'product'
within the terms of the Act. Whilst the physical medium upon which the
program is written is undoubtedly a product, and is therefore covered, the
same would not necessarily apply to the program itself. It is however
important to note that the Act covers only personal injury and damage to
private property; economic loss is not recoverable.
Both software purchasers and third parties may have a right of action in
tort against the software author if it can be proved that the author was
negligent. It is important to remember that, although in· contract the
doctrine of privity exists, there is no such obligation in tort. It was
established in Donaghue v. Stevenson (1932) that a duty of care is owed
to:
persons who are so closely and directly affected by my act that I ought
reasonably"'to have them in contemplation.
It is therefore possible that even third party users of a software package
could, in theory at least, sue the author for negligence if it can be shown
that he had been negligent in the writing of the program. The practicalities
of doing so would however be somewhat difficult.
Damages for loss of business property are available in negligence, and
have in the past also been awarded for pure economic loss (e.g. Junior
Books v. The Vietchi Co. Ltd (1982)). The Courts are however backtracking
somewhat from the decision reached in Junior Books, and it is now
unlikely that damages for pure economic loss would be awarded in
negligence.
In the case of software written especially for a particular firm however,
particularly if the contract for the system contained clauses such as those
mentioned earlier (for example governing permissible standards of accuracy
etc.), the situation is likely to be somewhat more straightforward. In this case
it is likely that, although the user will carry the responsibility for the
correctness of the computer's results, if he can show that the software is
defective then he will have a right of action under the law of contract
against the software author and may thus be able to 'step down' his
responsibility.
Negligence
In general a specialist who contracts to provide some skill, for example

carrying out a systems analysis or developing a computer program, is
bound by the provisions of the Supply of Goods and Services Act (1982)
which states that:
In a contract for the supply of a service where the supplier is acting in
the course of business, there is an implied term that the supplier will
carry out the service with reasonable skill and care.
This actual meaning of this requirement was defined as long ago as 1957
as:
Where you get a situation which involves the use of some special skill or
competence, then the test as to whether there has been negligence or
not is not the test of the man on the top of the Clapham omnibus,
because he has not got this special skill. The test is the standard of the
ordinary skilled man exercising and professing to have that special skill;
it is well established law that it is sufficient if he exercises the ordinary
skill of an ordinary competent man exercising that particular art.
(Bolam v Friern Hospital Management Committee(1957))
In other words the court will basically expect that professionals will display
the same standard of skill and care as might be expected from a reasonably
skilled and competent professional using well practised and accepted
professional methods of reaching a conclusion. Note that this does not
imply that a professional person warrants the succesfull outcome of their
endeavours, merely that they will exercise due care in seeking to obtain the
desired result. For example, a doctor could not be sued for professional
negligence following the death of a patient provided that his treatment of
the illness was not at fault, similarly an architect undertaking the design of
a building cannot be held liable in negligence if the client is dissatisfied
provided that he can be shown to have exercised due care in formulating
his design.
References
Arthur C., Mehta A. and Ernest-Jones T. (1988) 'Learning the facts of life' Computer Weekly
7.1.88p.18
Bigelow R.P. and Nycum S.H. (1975) Your computer and the law Prentice-Hall International
Brett H. & Perry L. (eds) (1981) The legal protection of computer software ESC Publishing
Computer Weekly (1988) 'Report by the Parliamentary Correspondent' Computer Weekly
7.1.88 p. 6
Ernest-Jones T. (1987) 'Software bugs can end in court' Computer Weekly 1.1 0.87 p. 3
Ernest-jones T. (1988) 'Software pirate is first in gaol' Computer Weekly 14.1.88 p. 1
Gooch B. (1986) The computer guide for builders Privately published

Kavanagh j. (1987) 'Officers make first strike' Computer Weekly 29.10.87 p. 1
Lawrence j. (1986) 'Rent, buy, lease - a beginner's guide' Construction Computing january
1986 p. 8
Nicolle L. (1987) 'Softly softly through the law courts' Computer Weekly 2.7.87 p. 27
Nicolle L. (1988) 'Nations move in to rid software sea of piracy' Computer Weekly 21.1.88
pp.20--21
Savage N. & Edwards C. (1984) A Guide to the Data Protection Act 1984 Financial Training
Publications Ltd
Sellers j. (1984) 'Contract negotiation' Construction Computing january 1984 pp. 29-31
Sterling j.A.L. (1984) The Data Protection Act 1984- a guide to the new legislation CCH
Editions Ltd
Thomas M. (1987) 'Protect the consumer - and yourself too' Computer Weekly 26.11.87 pp.
34-35
Wilson A. (1981) 'The protection of computer programs under Common Law- procedural
aspects and United Kingdom copyright law and trade secrets' in Brett H. & Perry L. (eds.)
The legal protection of computer software ESC Publishing
5 General Purpose
Software
Objectives

• Outline the advantages and disadvantages to be gained from the use
of word processing systems, spreadsheets, simple database manage-
ment software and integrated software packages.
• Explain the basic principles of operation of word processors, spread-
sheets and database management systems.
• Discuss the ways in which quantity surveyors might use general
purpose software.
General purpose software
In the beginning all computer software packages dealt with specific prob-
lems, for example a payroll program, a program to calculate the size of a
steel beam or a program to process Bills of Quantity. Programs such as
these are commonly written in one of the so-called 'higher level' computer
languages such as COBOL or Fortran, and one's ability to use the computer
to solve new problems using such methods obviously requires familiarity
with a conventional programming language.
During the mid-1970s however the advent of small desk-top computers
led to the development of interactive general purpose computer programs
such as spreadsheets for the manipulation of mathematical data, word
processors for the manipulation of text and data manipulation systems to
allow information to be stored, classified, sorted and retrieved in accordance
with user defined criteria. Such programs can obviously be used to solve a
wide range of different problems, and such 'general purpose' packages
have now been developed to the point where they offer very powerful
tools to the computer user without the need for conventional programming
skills. Collectively they probably now constitute the most widely used
range of computer software.
They are in general extremely reliable and easy to learn. McDonagh and
Daly (1987) gives an excellent account of the use of general purpose
software packages in a private quantity surveying practice.
General Purpose Software 81
Word processors
Word processing essentially means the manipulation of written text by

electronic means, and computer systems of this type are commonly used in
modern offices. Because the use of computers permits the easy alteration
and reorganisation of text stored in the system, word processors are
obviously most useful for documents such as reports which may be subject
to several revisions before the final version is approved. In the past this
process often necessitated retyping the document after each revision, with
the consequent risk of new typing errors occurring in a section of the report
which had previously been 'clean'. The use of electronic word processing
ensures that text can be easily inserted, deleted, altered or moved around
within the body of a document, the current state of the document being
always displayed on the screen and only printed out when all corrections
have been completed.
Word processors may also be very useful for 'standard' documents such
as standard letters or standard forms. In the past the use of 'standard'
documents such as these meant the storage of large stocks of the pre-
printed forms onto which names, addresses and other 'particular' informa-
tion had to be typed, but the use of a word processor not only allows the
standard text to be stored electronically and printed when required but
also allows such documents to be 'personalised' very easily and very
convincingly. The same techniques may also be used for other types of
document for example specifications. In the past specifications were often
prepared by 'cutting and pasting' a specification from a previous scheme
and the draft was then completely retyped. Using electronic word pro-
cessing documents can be 'cut and pasted' electronically and the need for
retyping is virtually eliminated.
Types of word processor
There are essentially two approaches to electronic word processing:

(a) word-processing software which will run on many types of general
purpose office computer
(b) dedicated word-processing computers where both the hardware and
the software are dedicated solely to text processing
Whilst word-processing software for use on general pupose office com-
puters has the advantage that the computer itself may be used for many
other tasks, such systems may provide less sophisticated and com-
prehensive facilities than dedicated systems, although one must hasten

to add that the software presently available is, in general, quite satisfactory
for general office use. Dedicated word processors tend to be more expensive
than a general purpose computer of equivalent power with word-
processing software, but are likely to provide additional facilities such as
customised keyboards with special keys for particular functions, and screens
able to display a complete A4 page rather than the normal 24 lines x 80
characters commonly displayed on a normal VDU. In addition dedicated
word processors may be capable of handling larger documents than can
often be accommodated on a general purpose machine, and dedicated
machines may therefore be preferred if large amounts of text are to be dealt
with on a regular basis.
Basic elements of word processing
There are three basic groups of facilities which will be present in all word-
processing systems:
(i) Facilities for entering and editing text including the ability to move
sections of text around within a document, copy text from one part of
the document to another etc.
(ii) Facilities to control the appearance of the text on the page typically
including control of margin widths, number of lines per page, text
justification etc., and also a mechanism to allow the use of special
printer facilities such as control of the type font, underlining, italics,
sub and superscripting, bold face, line spacing, character pitch etc.
(iii) Facilities for storage and retrieval of standard blocks of text including
the ability to merge into other documents as required.
The above facilities should be present in all modern systems but in
addition particular systems might offer:
(i) A spelling checker whereby the computer will check the spelling of
all words in the document against a stored dictionary and will report
any words not listed; the report will therefore include any mis-spelt
words. Many systems allow correction dynamically during the spell-
checking process, and some systems will even offer to correct
unmatched words with close matches selected from the dictionary.
Associated with the spelling checker there is, of course, a set of
dictionary maintenance routines to allow updating as required, a
most useful facility since most word-processing dictionaries seem to
come with American spellings!
(ii) Automatic indexing facilities are normally provided only on the more
expensive systems, but can be very useful in large documents.
(iii) A thesaurus facility to allow authors to ask the computer to suggest

alternatives to particular words.
(iv) Automatic generation of tables of contents, usually combined with a
system allowing automatic numbering of paragraphs in accordance
with user defined criteria.
(v) 'Macro' facilities allowing commonly used sequences of keystrokes
to be stored in the computer and invoked by the use of 'function' keys
thus mimicking the single key-stroke facility usually offered by
dedicated systems.
(vi) 'Mail-merge' facilities allowing complete runs of standard letters etc.
to be printed automatically with specific information such as names
and addresses being automatically inserted from another file as the
letters are printed. The generation of such documents then becomes
completely automatic once the operator has initiated the process.
(vii) Some systems offer limited mathematical facilities to allow, for
example, columns of figures included in the document to be auto-
matically summed.
Printers for word processing
The characteristics and methods of operation of those types of printer in

common use have already been described in chapter 2, but the selection
of an appropriate printer is obviously of crucial importance in a word
processing context. Whilst suitable printers will almost always be included
in dedicated systems, the choice of printer for word processing done on an
office computer is much more difficult. Many organisations insist on
'typewriter quality' text for the final document, and have in the past been
drawn towards printers of the 'daisy-wheel' type generating fully formed
characters, but the main disadvantage of this type of device is the slow
printing speed, typically no more than about 55 to 60 characters per
second (cps). It has in the past been common to overcome this difficulty
by using a second 'matrix' type printer operating at much higher speeds to
produce the draft documents, using the daisy-wheel only to print the
finished copy.
Small laser printers have recently begun to take over from daisy-wheel
printers and offer an excellent alternative for those who require high
quality print at high speeds. They are however rather more expensive than
the other types of printer in common use, and cannot use continuous
stationery.
Most modern matrix printers can operate in two separate modes, a fast
(say 130 cps) draft mode and a much slower mode which involves
multiple passes over the text to generate much higher quality print for
finished documents. One such machine can therefore offer a useful

compromise in a small institution if the budget is restricted.
'Social' effects of word processing
Perhaps the most important social effect which has typically followed from
the development of word processors, particularly those implemented on
portable lap-top computers has been an increasing tendency for executives
to do much more of their own typing rather than either writing longhand or
dictating either to a dictaphone or directly to a secretary.
In the words of Reis (1984):
Word processing is not just for secretaries. Anybody can use a word
processor, in fact in my office we have no typewriters at all -everything
is done on a word processor and we do most of the typing ourselves.
One of the major advantages of word processing is that mistakes can be
easily rectified; therefore bad typing is not a major problem. I find it
quicker to word process myself than to dictate or write the stuff for a
typist to type it out for me.
Other effects of the word-processing revolution may therefore be:
(a) not only to reduce the demand for pure typists but also to render those
previously highly prized and accurate typists much less desirable than
those who might be less accurate but faster
(b) to ensure that basic keyboard skills form an essential component of
the budding executive's skills portfolio
Spreadsheets
Spreadsheets are to financial calculation what word processing is to text.

Essentially, in its simplest form, a spreadsheet can be imagined as a large
sheet of paper divided vertically into a number of columns and horizontally
into a number of rows to form a grid. In the words of Hunt (1986):
(A spreadsheet) can best be described as an enormous electronic abstract
sheet.
Each 'pigeon-hole' formed at the intersection of a column and a row is
called a 'cell', and may be addressed by quoting the appropriate column
and row numbers (see figure 5.1). Cells may contain either a number, some
text or a formula indicating that the contents of the cell are to be derived
by calculation from the contents of other cells.
A B c D
2 apples 10
3 pears 5
4 total (B2+B3)15
Figure 5.1
In figure 5.1 therefore:

cell A2 contains the text 'apples'
cell A3 contains the text 'pears'
cell A4 contains the text 'total'
cell B2 contains the number 10
cell B3 contains the number 5
cell B4 contains the formula B2+B3 to signify that the cell value is to be
calculated by adding the contents of cells B2 and B3 giving a total value
of 15
Spreadsheets provide a comprehensive range of in-built functions to
allow extensive manipulation of mathematical data, and the facilities
commonly provided may be grouped as follows:
(i) Text entry and editing functions including facilities to move or copy
blocks of information from one part of the sheet to another. Since an
electronic spreadsheet may be very large (perhaps dozens of columns
and hundreds of rows), and a normal VDU can only display 24 lines
x 80 characters of information, the screen is used as a 'window'
which may be moved around the total spreadsheet at will. Most
systems also allow the screen to be split either horizontally or
vertically, or in some cases both, in order to display either two or
four separate 'windows' at the same time.
(ii) Formatting functions to alter the display format of the sheet, for
example some systems allow selected data to be displayed as bar
charts, pie charts or histograms, to alter the widths of columns, to
display information in alternative formats (e.g. integer form, truncated
to two decimal places, scientific notation etc.).
(iii) Mathematical functions which may be included in formulae to allow
the manipulation of data. Typical functions provided in all systems
might include the following although some systems will provide

many more:
+addition
- subtraction
* multiplication
I division
1\ exponentiation (raising to a power)
< is less than

> is greater than
=is equal to
<> is not equal to
<= is less than or equal to
>= is greater than or equal to
SUM totals a list of numbers
COUNT counts the number of entries in a list
AVERAGE calculates the SUM of a list divided by the COUNT
MIN finds the lowest value in a list
MAX finds the highest value in a list
INT truncates a number to integer form
ABS removes any minus sign to give an absolute value
SQRT returns the square root of a number
In addition formulae may also include logical operators, for example if a
cell contains the formula:
IF(expression,formula1,formula2)
then 'formula1' will be used if 'expression' is TRUE, whereas 'formula2'
will be used if 'expression' is FALSE.
Similarly the formula
OR(expression1,expression2)
returns a value of 1 if either 'expression1' or 'expression2' is TRUE, but
returns a value of 0 otherwise, and the formula
AN D(expression 1,expression2)
returns 1 if both 'expression1' and 'expression2' are TRUE, otherwise a
value of 0 is a returned.
The formula
NOT(expression)
returns a value of 1 if 'expression' is FALSE or a value of 0 otherwise.
These particular functions therefore provide the spreadsheet user with a
branching mechanism, thus allowing spreadsheet packages to be used as
programming languages in their own right. Trigonometric functions are

also commonly included (angles may have to be given in radians):
SIN(a) returns the sine of angle 'a'
COS(a) returns the cosine of angle 'a'
TAN(a) returns the tangent of angle 'a'
The ability to use formulae to derive cell values from the values of other
cells obviously means that commonly used calculations can be set up on
standard sheets and the results immediately displayed for any particular set
of input data. Spreadsheet systems have thus achieved great success and
widespread use among accountants and business planners for performing
'what-if' simulations, but it is a mistake to view them purely in this way.
Weber (1987), in using a spreadsheet to create an expert system, draws
attention to the wider possibilities:
All too often spreadsheets are treated as if they were only fit for
accounting and financial forecasting. Even the simplest spreadsheet is a
limited form of language while the more sophisticated ones are powerful
and versatile programming languages in their own right. Their control
structures tend to be limited, but they do offer many advanced functions
that would take a great deal of coding in other languages.
Spreadsheet systems therefore present extremely powerful and flexible
tools for many different types of application. One or two systems even
allow the data to be considered in a three-dimensional form rather than in
the more conventional two-dimensional form, thus allowing the analogy of
a book rather than a single sheet.
Typical applications which have been developed in quantity surveying
include measurement (Smith (1986)), cash flow forecasting (Hunt(1986)),
early design estimates (Williams (1987)) and the cost modelling of building
services (Ellis (1986)). McDonagh & Daly (1987) record the use of spread-
sheets for many aspects of quantity surveying work including cost planning,
estimating and the production of interim valuations, and Cooke & Bala-
krishnan (1985) illustrate examples of use in cash flow forecasting,
accounting, post contract management, life cycle costing and plane sur-
veying. Hawkins & Massey (1985) also give, amongst others, examples of
formula fluctuation calculations, 'ticksheet' valuations, cut and fill calcula-
tions, tender error correction and land valuation.
It seems possible that the spreadsheet concept will continue to develop
even further. lnce (1988) reports that Xerox have developed a package
called the Analytic Spreadsheet Package (ASP) within which cells can not
only contain numbers, text and formulae but also 'objects' such as pictures,
knowledge bases, dialogues, and pop-up menus.
Database management software
Perhaps the most widespread use of computers today, certainly in the

commercial world, is for the storage and manipulation of information. The
storage capacity of modern computer systems means that even small desk-
top microcomputers can store, analyse and manipulate very large quantities
of information, thus providing management with extremely powerful
decision support tools. The use of databases as general purpose informa-
tion systems by quantity surveyors in multi-disciplinary organisations is
well established, but some have used small database manipulation packages
in more specialised quantity surveying practice. The collection and manipula-
tion of cost data is perhaps the most obvious area.
Betts (1987) proposes the use of a micro-based database package for the
collection and analysis of information by building contractors for use
during the tendering process.
The ability to store such large volumes of information poses a number of
problems among the most pressing of which a.·e:
(i) How does the user keep track of all of the information stored in the
system?
(ii) How does the user instruct the system to select and present the
information required in the form in which it is required for a specific
purpose?
(iii) How does the computer manager ensure that the information is
stored efficiently, can be accessed in a wide variety of ways and can
be easily updated if required?
Relational databases
The problems mentioned above obviously become more acute the larger
the collection of information becomes, and the administrative problems
involved in controlling both access to and updating of the information
become more difficult to satisfactorily resolve as the number of users
increases. There is therefore the need to establish some logical structure
within the computer, within which the information can be stored and
which will allow the information to be referenced in such a way as to
permit the scanning, selection, analysis and presentation of relevant informa-
tion in accordance with particular criteria specified by the user.
Collections of data organised in this kind of structured way have become
known as 'databases', and a wide variety of software tools have been
designed specifically to organise and manipulate such databases, ranging
from simple electronic 'card index' systems suitable for use on the smallest
of microcomputers up to very large systems designed for use on large
mainframe machines.
Information systems containing large volumes of data pose particular

problems especially as regards organisation of the information. PMA
Consultants (1 979) is a useful study of the problems of controlling and
integrating construction industry information, and also contains a good
basic overview of the relevant principles of information science.
Such systems also present significant computing problems in order to
ensure efficient storage of the information and optimum speed of access.
Such problems become especially critical as the number of users grows
and their reporting and presentation requirements become more complex,
and some very complex methods of referencing the relationships between
the various pieces of data have been devised. Databases organised on this
kind of model are termed 'relational', and the definitive work in this area is
encapsulated in a set of twelve rules formulated by Dr Edgar Codd. Codd's
rules, summarised in Hares (1 988), form an outline specification for a truly
relational database. Hares (1 988), however, points out that no available
database management system yet complies with all Codd's rules although
the technical solutions to ensure compliance are now well understood and
it is likely that full compliance will be achieved very soon.
lnfomatics (1 988) contains a survey of available relational database
management software packages, ranging in price from £400 for a system to
run on an IBM PC up to systems costing in excess of £300,000 running on
large mainframes. The most popular in the United Kingdom is shown to be
the dBase Ill package from Ashton-Tate with a claimed 200 000 users.
Fourth generation languages (4GL)
One of the fundamental requirements for a good database management

system is that the user should be able to interrogate the database in a
language, similar to that used for natural speech. True natural language
interaction with computers still remains a dream (but see also chapter 8),
although considerable research has been carried out in this field. Each of
the database management systems presently available includes some
interpretation of a natural query language (indeed the language used for
the Software AG product 'Adabas' is even called 'Natural'), and such
programming languages have been termed 'fourth generation languages',
often abbreviated to 4GL. Fourth generation languages as a class are very
powerful and provide extremely clever ways of manipulating the informa-
tion stored in the database with considerably less effort and within a much
shorter time than was possible with languages such as COBOL or PL/1.
Specialist programmers experienced in fourth generation languages are in
great demand, and database query languages are now being used for many
commercial application systems which would, in the past, have been
written in languages such as COBOL. Reed (1987) is a case study of the

problems faced by a major company who, with a need to update its
commercial computer systems, decided to adopt a 4GL solution.
Database management and control
The introduction of a multi-user database poses particular management

problems in that different users might need to be allowed different 'rights'
to access and update the information. Information stored in a database is
likely to represent a considerable investment in terms of time spent
collecting it, and any organisation will naturally want to protect the data-
base from unauthorised alteration by users whilst still allowing the data to
be read. Alternatively, some parts of the data may be particularly sensitive,
and one might need to restrict some users' access rights whilst still allowing
access to other non-sensitive data. Most systems allow the data to be
protected by various levels of security, and management must decide what
database administration and security procedures are appropriate for each
particular installation or organisation (see also chapter 3).
Integrated software packages
From the point of view of the microcomputer user, the three general
purpose software tools so far described, word processors, spreadsheet
systems and information management systems, form three complementary
parts of a complete business system. In the past limitations on the available
memory capacity have meant that it has only been possible to run one of
the three packages at a time, and users have therefore been largely
deprived of the facility to integrate information from one element, e.g. a
spreadsheet, into another element, e.g. a word-processed report. The develop-
ment of much larger microcomputers led directly to the development of
software packages which integrated the three basic elements and attempted
to provide facilities for the interchange of information between one element
and another.
A number of such packages now exist, among the most popular are
'Symphony' from Lotus and 'Framework' from Ashton-Tate, but it should
be noted that such packages are at best a compromise and are unlikely to
provide the power of, for example, a dedicated database package such as
dBase Ill.
Desk-top publishing
Desk-top publishing software is, in one way, a logical extension of word-

processing in that it attempts to place facilities for the production of high
quality printed material within the reach of the microcomputer user. Desk-
top publishing is not however simply enhanced word processing; as Budgett
(1987) explains:
It is essential to recognise at the outset that desk-top publishing software
is totally unlike any other; by itself it doesn't produce anything. It most
closely resembles an infinite supply of blank sheets of paper onto which
items may be attached in order to create a page, which may contain iext
or graphics or both.
Desk-top publishing software is therefore intended to allow users to emulate
electronically the manual cut and paste operations involved in traditional
publishing, as well as providing many alternative typefaces, styles and
sizes of lettering. In essence it is to take the output from other programs, for
example text created using a word processor, diagrams, charts and tables
perhaps produced by a spreadsheet or business planning package, or even
pictures created by a graphics design suite, and lay them out on to an A4
page so as to create a true camera-ready copy.
The process is therefore much more to do with art than technology; in
practice a skilled professional paste-up artist can vary the layout to make
the effect pleasing to the eye, informative, amusing or whatever other effect
is appropriate to the particular work in hand. An understanding of the
fundamentals of design and layout is therefore required if lay users are to
make the best use of the facilities provided.
In practical terms any worthwhile desk-top publishing package will be
able to read information in native format from a wide range of other
packages and should be able to preserve attributes such as underlining,
bold face etc., and should also permit text and graphics to be collected
from different packages and merged onto the same page. Pictures read into
a system using a scanning technique can also be reproduced but the
amount of storage required to hold a reproduction of a black and white
picture with sixteen levels of grey is enormous.
Systems are available to run on a wide variety of computers, but a laser
printer is essential.
References
Betts M. (1987) 'A co-ordinated system of information retrieval for building contractors
tendering' in Brandon P. (ed.) Building Cost Modelling and Computers E. & F.N. Spon
Brandon P. (1987) (ed.)Building Cost Modelling and Computers E.& F.N. Spon
Budgett H. (1987) 'What every desktop publisher should know' Personal Computer World
August 1987 pp. 146-150
Cooke B. and Balakrishnan S.V. (1985) Computer spreadsheet applications in building and
surveying Macmillan
Ellis C. (1986) 'M and E computer spreadsheets' Chartered Quantity Surveyor january 1986
PP· 10-11
Hares j. (1988) 'Tapping the trends' lnfomatics February 1988 pp. 68-70
Hawkins P. and Massey R. (1985) Spreadsheet examples Trent Polytechnic Department of
Surveying
Hunt G. (1986) 'Micronotes: The enormous abstract sheet' Chartered Quantity Surveyor
February 1986 p. 15
lnce D. (1988) 'Defining The Object' lnfomatics July 1988 pp. 66-71
lnfomatics (1988) 'Survey of relational databases' lnfomatics February 1988 pp. 74-78
McDonagh N.H. and Daly J.P. (1987) 'Cost management of construction projects' in Brandon
P. (ed.) Building Cost Modelling and Computers E. & F.N. Spon
PMA Consultants (1979) Relevance of information syntax to computer data structures PSA
Directorate of Architectural Services Data Co-ordination Branch
Reed D. (1987) 'Playing the fourth generation game' lnfomatics October 1987 pp. 57-60
Reis G. (1984) 'All you wanted to know about word processing but were afraid to ask'
Construction Computing April 1984 p. 4
Smith j. (1986) 'Measuring quantities with the aid of a computer spreadsheet' Construction
Computing Autumn 1986 pp. 21-23
Weber J, (1987) 'Working to rule' Personal Computer World June 1987 pp. 130-133
Williams G.R. (1987) 'The development of a spreadsheet application for the cost modelling of
buildings during the early design stages' in Brandon P.S. (ed.) Building Cost Modelling and
Computers E. & F.N. Span
6 Public Information
Systems for Quantity
Surveyors
Objectives
• Explain what a bulletin board is.

• Explain what a public information system is.
• Outline the principal types of public information system which may
be of use to quantity surveyors.
• Explain what is meant by 'public domain' software.
• Explain the hardware considerations which it may be necessary to
take into account to allow access to bulletin boards and public
information systems.
• Comment on the economics of using public information systems and
the constraints which management might need to apply.
Bulletin boards
Despite the problems of communication mentioned earlier in chapter 3

computer systems have, for many years, been used to disseminate informa-
tion. Groups of computer enthusiasts who may be spread over considerable
distances have long been among the most active in this field, using their
home computers to pass both programs, data files and messages via the
public switched telephone network. Many of the early computer groups
established a central data store which could be accessed by anyone else in
the group, and which could be used to 'post' messages and notices of
general interest. Such systems quickly became known as 'bulletin boards',
and Toothill (1987a) states that the first such system was established in
1978 in the United States. The idea quickly spread, and in a very short
time large numbers of open-access bulletin board systems became estab-
lished, mainly among the computer hobbyist community on both sides of
the Atlantic. Toothill (1987b) for example lists details of 130 such systems
available in the United Kingdom in August 1987 including addresses and
telephone numbers.
Electronic mail
A logical development of the early bulletin board services was the develop-
ment of personal communication systems based on computers; essentially
a system in which every user has his own private bulletin board onto
which anyone with the correct telephone number and password can leave
messages. In effect the system provides each user with an electronic mail
box, and such systems have therefore become known as 'electronic mail',
often abbreviated to 'Email'. The principal advantage of the technique is
usually quoted as one of speed since the message is received so soon after
transmission as to be almost instantaneous; devotees of the technique
therefore frequently refer to the more traditional mail services as 'snail
mail'. Many private organisations have their own electronic mail systems,
often extending over considerable geographical distances.
The two best known publicly available electronic mail services in the
UK are probably 'Prestel' and 'Telecom Gold'.
By far the largest and most dramatic growth in the area of electronic
communication has been achieved by Fax (facsimile transmission) systems.
The transmission side of a Fax machine essentially consists of an image
scanner connected via a microprocessor to the telephone network. It can
therefore scan and mail any image whether text or picture. Receipt of
information is also controlled by the microprocessor, and reproduction of
the image is usually by electrostatic means. Whilst the quality of drawn
information is not particularly good at present the system is adequate for
sketches and text. Nonetheless, Fax has revolutionised business communica-
tion in a very short space of time. The vast majority of firms now have a
Fax machine, and it is possible for written and drawn information to be
communicated almost instantly to many parts of the world. Improvements
in the technology will doubtless mean that in the near future electronic
communication of written and drawn information will become the norm
and a reduction in the price of small machines will soon bring them within
the reach of the domestic consumer. It is unlikely, however, that conven-
tional postal systems will completely disappear. Fax is not at present
suitable for documents larger than A4 or consisting of more than a few
sheets, although no doubt even these restrictions will in time be overcome.
The EDICON initiative
Industries such as shipping already make considerable use of electronic

mail, and probably the most important development in the use of the
technology in the construction industry has been the establishment of an
organisation called Electronic Data Interchange (Construction) Ltd,
abbreviated to EDICON. EDICON was established in May 1987 as a non-
Public Information Systems for Quantity Surveyors 95
profit making organisation devoted to introducing electronic data transfer

throughout the whole of the construction industry, and a full report was
published in Stewart (1987). Working groups have been established to
investigate the possibilities in CAD, bills of quantities production, ordering
and invoicing and the production of technical and product information.
Partners in the venture include, amongst others, the Department of the
Environment Property Services Agency, construction groups Trafalgar House
and John Laing, quantity surveyors, architects and leading materials
suppliers and builders merchants.
The intention is to replace the majority of routine postal transactions
with electronic data transfer within a period of five to ten years.
Public information systems
At the same time as the hobbyist community was establishing its open-
access bulletin boards a number of private organisations were also
developing similar systems for information dissemination either for use
within individual companies or for particular restricted groups of people,
and within a very short period of time it became apparent that information
was a highly marketable commodity. Many such systems now exist, and
many are available for general use upon payment of a subscription.
Some information systems, for example Ceefax and Oracle, are free of
charge and do not even require a computer, merely a suitable television
set, although conversation with such systems is of course a 'one-way'
process and the system does not allow for any 'filtering' of the information
by the user. Information is passed in 'pages', and the user must then select
the parts required from the block of information presented.
The use of a computer to interrogate an information system means that
the user can instruct the host system only to provide that part of the data in
which he is specifically interested. The user is therefore able to search very
large databases of information much more selectively.
A number of such systems exist which may be of direct use to quantity
surveyors and which are generally available upon payment of the appro-
priate fee. Examples of such systems include:
• BCIS Online - Direct connection to the Building Cost Information

Service.
• Textline and lnfoline- Databases of news extracted from newspapers
and magazines.
• Context - Information on building products, British Standards and
Codes of Practice etc.
• BRIX - Building Research Station database including information on
BRE Information Digests etc.
An example of a service which attempts to link the concepts of electronic

mail, a bulletin board and an information system specifically for quantity
surveyors is the TP-DOS Online system, an electronic mail and information
service set up and run by the Department of Surveying, Trent Polytechnic.
The system at present includes:
(i) A message area or bulletin board allowing users to communicate
either with each other or to broadcast information to all users
(ii) A file area containing files of information, programs, details etc.
which may be down-loaded to subscribers' machines
(iii) A services area allowing subscribers to remotely run software such as
database enquiry systems which are resident on the host computer
The main disadvantage associated with the use of remote computer based
information systems is seen by some as not the cost of subscription but
rather the cost of connection. Even using the public telephone network
however connection costs should not be too high provided that the host
computer is within local dialling range. Even if the cost of the call were to
be charged at the highest inland rate Toothill (1987c) calculates that the
hourly connect charge should be only £1 0.17p Oune 1987 prices).
Perhaps a greater disadvantage at present is that the whole business, at
least in the United Kingdom, is still rather cumbersome and slow, but
engineering improvements to the telephone network such as the installa-
tion of digital exchanges and fibre optic cables which are presently in hand
should ease the problem in due course.
Public domain software
'Public domain' software is computer software which is readily available,

usually free of charge, upon request. Users are allowed to copy the
software, indeed are even encouraged to distribute it more widely, provided
they do so without charge.
The idea of public domain software was born in the United States where
a number of programmers, mostly hobbyists, decided that microcomputer
software was generally too expensive and they therefore began to distribute
their own programs free of charge. The idea has since grown, and Goodwin
(1986) reported that in October 1986 the largest public software library in
the United Kingdom, the Public Domain Software Interest Group (PDSIG)
'has a library of about 20 000 items on 1 200 disks.' All mat~rial is offered
free to members of the library, upon payment of a small annual sub-
scription. The various catalogues include all types of software including
word processors, spreadsheet systems, data base managers, expert systems,
computer aided instruction systems, communications packages and many
more.
Programmers and software authors are not, generally, paid for software
in the public domain, and whilst the range of public domain software is
very large, the quality tends to be somewhat variable. In most cases the
distribution diskette also includes program documentation and user
manuals. The basic problem is that such software is not in any way
supported and is offered 'as-is'. Users might therefore find bugs and
problems for which there is no easy fix, but as Goodwin (1987) points out
the same criticisms may be applied to a good deal of commercial software.
A variation on the public domain theme is 'shareware'. The shareware
concept again originated in the United States, and is based upon the
philosophy that most people will be prepared to pay something for good
software. Accordingly the software is freely available, initially free of
charge but users are asked to make a donation to the author of whatever they
consider the software to be worth. Most authors undertake, upon payment
of a stated minimum fee, to offer some degree of support including free
updates to the software as and when they become available, and often
access to a help line. Unfortunately the authorsofmostofthe shareware avail-
able in the United Kingdom live in the United States and telephone help lines
may not be all that useful!
A further twist is 'approval-ware'. This concept, common in the United
States but so far uncommon in the United Kingdom, is very akin to the
shareware concept described above, except that in this case users are in
effect sent software 'on approval' through the public domain system. In this
case users who decide to keep the software are asked to pay a stated sum
of money as a licence fee. Similar incentives to those offered by shareware
authors often apply.
Connection to a remote information system
Connecting a computer to a remote information system requires certain

additional hardware and software. As far as the computer itself is con-
cerned virtually any modern micro could be used provided that it has a
serial (RS232) port. In theory all that should be necessary is to connect the
computer to the host, install the necessary software and access the data
base, but in practice, of course, nothing is that simple.
Unfortunately, the electrical signals generated by computer systems are
not generally compatible with the electrical characteristics of the metal
wires and exchange equipment of the normal telephone system, along
which most long distance communications traffic is carried. It is therefore
necessary to interpose some electronics between the computer circuitry
and the telephone system in order to modulate the frequency of the signal
to simulate voice traffic before transmission and demodulate the signal
back to computer frequencies at the receiving machine. Such a device is
called a 'modem' (MOdulator/DEModulator). In most installations the

modem is external and plugged into the computer and the telephone
system but internal modems are available for some machines. A particular
form of the device which contains a special socket into which the handset
of an ordinary telephone is placed is known as an 'acoustic coupler'. Such
devices can be very useful for communicating from locations such as
public call boxes where a direct jack-plug connection to the telephone
system is not possible.
A schematic representation of the arrangement is shown in figure 6.1.
USER"S
MICROCOMPUTER
PUBLIC TELEPHONE
NETWORK
HOST MAINFRAME COMPUTER
Figure 6. 1 Connection to a remote information system

In addition, of course, users will need some suitable communications

software to enable their computer to 'talk' to the host. Both of these
requirements are discussed in more detail below.
Modems
In addition to converting the signal to a suitable frequency for the tele-

phone system to handle, the modem usually also controls the speed of the
transaction, and it is obvious that the modems at each end of the line must
both be capable of operating at the same speed if meaningful communica-
tion is to take place. The speed of transmission is known as the 'baud rate',
and most systems can send information in both directions at the same time;
technically known as 'full duplex' operation. A number of different protocols
exist:
V21 operates at 300 baud full duplex (approximately 30 per second in both
directions simultaneously). This standard is now somewhat old
fashioned and mainly used for acoustic couplers
V22 operates at 1200 baud full duplex, approximately 120 characters
per second in both directions simultaneously. This standard is pre-
sently the most popular choice for communication over the public
switched telephone system since it offers the fastest practical speed at
which reliable communication can take place withoot complex error
checking software. A wide variety of modems is available offering
this standard
V23 operates at a split speed of 1200 baud in one direction and 75 baud in
the other in full.duplex mode. This standard is peculiarly British and
was developed specifically for the Prestel electronic mail service.
Equivalent character transmission speeds are about 120 characters per
second in one direction and 7.5 characters per second in the other
V22bis2400 baud full duplex, approximately 240 characters per second in
both direction simultaneously. Problemsofline 'noise' become critical
at speeds as high as this and complex error checking routines must
be employed. Toothill (1987d) discusses the problem in more detail
Note that the above speed restrictions only apply because of the
deficiencies of the telephone technology. Over short distances very much
faster transmission speeds are commonly used; 9 600 baud full duplex
(960 characters per second) is common with mainframe systems and
speeds up to 19 200 baud full duplex (1 920 characters per second) are not
unknown. In the future long distance communication speeds can be
expected to increase dramatically with the widespread introduction of fibre
optic cables and digital telephone exchanges.
Many modern modems also incorporate facilities to allow automatic

dialling and answering (autodial/auto answer), and there are a number of
ways in which such facilities can be controlled. The most common is a set
of protocols developed by the Hayes company in the United States, and
modems which use such a protocol are therefore designated 'Hayes
compatible'.
Communications software
In order to actually use the computer and modem to transfer information

the computer must be instructed how to transmit and receive data and to
perform some system management functions such as logging the trans-
actions. These functions are accomplished by a communications software
package, and again there is a range of accepted protocols in common use.
The most common in Britain are XMODEM and its derivatives and the
Kermit program now available in the public domain. Toothill (1987d) gives
a brief description of the protocols in common use.
References
Goodwin S. (1987) 'Computer answers "Free software" Personal Computer World October
1986 p. 206
Stewart A. (1987) 'Postage stamped out' Building 22/5/87
Toothill P. (1987a) 'News Bulletin' Personal Computer World May 1987 pp. 174-175
Toothill P. (1987b) 'UK Bulletin Boards' Personal Computer World August 1987 p. 214
Toothill P. (1987c) 'Go by the board' Personal Computer World june 1987 pp. 174-175
Toothlll P. (1987d) 'Clearing the line' Personal Computer World April1987 pp. 174-176
7 The Development of
In-house Systems
Objectives

• Explain the main fat1ors to be considered when contemplating the
development of an in-house computer program.
• Explain the basic principles of systems analysis in the context of the
development of a new computer program.
• Outline the sequence of operations in the development of a new
computer program.
• Explain the roles and responsibilities of the various professionals
involved in the development of a new computer program.
• Discuss the management problems commonly posed by the develop-
ment of new computer systems particularly in the control of time and
cost.
In-house systems development
The development of purpose-built in-house software packages is a sub-

stantial task. It is invariably very expensive and hardly ever completed
within the originally set parameters of time and cost. It should therefore
generally only be considered as a last resort when all other options have
been explored, but unfortunately this exploration process is often merely a
cursory glance, and in many cases companies are discouraged by the
apparent high prices of available software packages; there is always the
feeling that it will be cheaper to do the job in- house using staff already on
the payroll. In many cases this option, although it may appear at first
glance to be more cost effective, proves to be a false economy. In the
words of Gooding (1988):
A great many companies look inside for solutions. They are still prepared
to develop their own systems when a step back from the problem would
tell them that reinventing the wheel is wasteful. Packages are more than
trouble free code; they represent an investment in someone else's ex-

perience.
In the event that it is decided to develop systems in-house, the actual
development work is likely to be done by a professional systems analyst,
but the major stumbling block which often occurs is that although the
analyst knows a great deal about systems organisation and, usually, a great
deal about the available computing options, he or she is much less likely
to have a working knowledge of the client's business. This may pose
particular problems in a specialist area such as quantity surveying where
the technical jargon alone may be difficult for the analyst to understand,
and it may take a considerable length of time before he or she is able to
begin to build an accurate picture of the client's activities, to understand
their particular special requirements.
On the other hand the typical client is equally unfamiliar with the
systems analysis and design process. They know what it is they need to
achieve, but are unable to articulate that requirement clearly in a language
that the analyst can understand. In many cases this uncertain dialogue
considerably lengthens the development process as both client and consul-
tant gradually overcome this communications block. In the very worst
cases both parties appear to have reached a consensus but, because of the
communication problem, have differing perspectives of what has apparently
been agreed with the result that the system, when it finally appears, is
unsatisfactory. It is therefore possible for the client greatly to aid the
development process and enable the analyst much more easily to reach a
correct interpretation of their needs if the client too understands something
of the process of analysis and systems development.
A further point to consider is that of the management of the development
process. It is customary for the systems analyst to also act as development
manager, but there is no reason why this should continue to be so. Indeed
there is no more reason for the systems analyst to manage the development
process than there is for the architect to manage the process of procuring a
building. Many of the management techniques with which the quantity
surveyor is already familiar are just as applicable to the development of a
computer system as they are to the construction of a building or the
installation of process plant.
This chapter then does not attempt to provide a comprehensive course
on systems analysis and design, but rather to give an overview of the
process from the layman's point of view.
The Development of In-house Systems 103
The systems analyst
It is perhaps necessary, before beginning a discussion on the management

of systems development, to establish what is meant by 'systems' and the
'systems analyst'.
NCC (1978:p. 1) defines a system as:
a set of interacting elements responding to inputs to produce outputs
Any system can therefore be considered as having three component parts:
'input(s)', some kind of 'process' and 'output(s)'. Examples of systems might
therefore be the production of a Bill of Quantity, or the construction of a
house. Both take some form of input, process it in some way and produce
some output. Systems may in general therefore be represented as:
Input --> Process --> Output
The above definition, of course, allows some very complex systems. Indeed
the whole of life could be seen as a system, and it is therefore common to
break large systems down into component sub-systems, a process commonly
known as 'factoring'. The system for the traditional production of a Bill of
Quantity could therefore be seen, not as one system but as a collection of
inter-connected sub-systems, for example taking off the dimensions,
squaring, abstracting, billing, editing, typing and reproduction each of
which is actually a system in its own right containing its own inputs,
processes and outputs. Such sub-systems can therefore be arranged in a
hierarchy to give a much more complete picture of the bill production
process; the 'system' as a whole. It is the definition and analysis of such
systems which is the primary task of the systems analyst. In the words of
NCC (1978):
It should be clear that the major task of the systems analyst is to define
the system in terms of its objectives, inputs, processes, outputs, boundary,
and the interfaces between its subsystems.
The second task for the systems analyst is to understand how the the
system behaves.
In the context of the above it is therefore obvious that systems analysis as a
science, although it has in many minds become synonymous with
computer applications, actually owes its origins to the management science
techniques of operational research, the study of organisation and methods,
and work study.
The system life-cycle
The actual process of systems development can be broken down into a

number of discrete stages:
(i) the preliminary assessment
(ii) initial outline study
(iii) detailed study
(iv) detailed systems design
(v) programming
(vi) implementation
(vii) maintenance and review.
Note: Stages (i)- (iii) are sometimes collectively known as 'require-
ments analysis'.
This collection of tasks is in many ways analogous to the tasks involved in
procuring a building and is sometimes referred to as the 'system life cycle'.
The process contains many natural break-points and feedback loops where
management decisions are required and is shown diagrammatically in
figure 7.1.
Note that, as in a building project, radical changes of direction become
progressively more serious as the process proceeds.
Preliminary assessment
The preliminary assessment is a very high level overview of the proposed

system. It will generally consist of an examination of the outlines of the
problem. The objectives of the exercise are to set the bounds of the
problem; to establish the terms of reference, and the outcome will usually
take the form of a proposal for further investigation.
The process is largely investigative in nature and is in many ways
analogous to setting the brief for a construction project. Although there
may be advantages in this process being carried out by a trained systems
analyst, the analyst can for example look at the problem from an unbiased
viewpoint, the learning curve in the case of specialised areas like quantity
surveying generally means that there is often much to be said for this
process being carried out by a quantity surveyor provided that the necessary
objectivity can be guaranteed.
At the completion of this stage the client has the opportunity to accept
the proposal, to suggest possible revisions or to abandon the project
altogether.
The Development of In-house Systems 1OS
PRELIMINARY
ASSESSMENT
l
CONTINUE? REVIEW ABANDON
l tI
I
I
INITIAL STUDY
l
CONTINUE? REVIEW ABANDON
1 t
I
I
I
DETAILED STUDY
1
PRDTOTYPING REVIEW ABAN'DON
l •
DETAILED DESIGN
1
PROGRAMMING
1
IMPLEMENTATION REVIEW ABANDON
l
MAINTENANCE
~
Figure 7. 1 The system life cycle
The initial outline study
Working from the study proposal generated during the preliminary assess-
ment phase the systems analyst will now begin an initial study of the
problem. The study is likely to concentrate on an analysis of the existing
system, inputs, outputs and processes in order that the analyst may formu-
late an understanding of the key areas and in turn develop an overall
assessment of feasibility, time scale and cost. It is perhaps at this early
stage that accurate communication between the analyst and the client is
most important. It is at this stage that the analyst is likely to know least
about the client's business; this stage where inaccurate perceptions of the
problem may be formed; this stage where a complete understanding and
common perception of the problem area is essential ifthe following stages
are to be successful.
It is at this stage that the analyst is attempting to prepare the first
tentative estimates of the scale of the problem, beginning to think about
possible overall strategies and to prepare the first 'ball-park' estimates of
cost. The analyst's job at this stage is to some degree akin to that of a
quantity surveyor faced with a client who knows that their existing accom-
modation is inadequate but is unsure about whether to extend, refurbish or
re.build on a new site.
The outcome of the process should be a report detailing the analyst's
perception of the problem including approximately costed alternative
pro~oosals together with projected time-scales and a note of the advantages
and disadvantages of each. It is vitally important that this report should be
writt,en in a language and style that the client can understand, as it is on
the b•asis of this report that the client will decide whether or not to invest
substamtial sums of money.
Once again the outcome will be either a decision to continue, to review
and revise, which may also at this stage involve a reappraisal of the
original problem assessment, or to abandon the project altogether. Total
expenditure to this point is likely to still be relatively small and abandon-
ment is therefore still a practical proposition. Beyond this point costs begin
to rise s1teeply and the financial consequences of abandonment become
progressively more serious. In the words of Davis (1983:p. 10):
This presentation marks a crucial decision point in the life of the project.
Many projects will die right here; only those promising a sufficient return
on investment should be pursued.
A good example of the kind of problems which can occur when the early
stages are not successful is reported in Ernest-Jones (1988).
The Development of In-house Systems 1 07
The detailed study
Following successful completion of the outline study phase the analyst will
now proceed to a detailed study of the problem and will attempt to
prepare firm proposals for the system with more accurate estimates of time
and cost.
There is a considerable degree of overlap between the detailed study
and the detailed design phase, with no hard and fast rules about which
phase certain processes fall into. I prefer for example to consider the
detailed formulation of input and output mechanisms as a part of the
detailed study phase, whereas other commentators (e.g. NCC (1978))
consider this to be part of the systems design phase. Nonetheless, the
detailed study process is largely analogous to the sketch planning stage of a
building contract with the analyst combining the roles of architect and
quantity surveyor.
It is at this stage that the detailed 'systems analysis', or analysis of the
customer's business is carried out. It is not the purpose of this text to give a
detailed explanation of systems analysis techniques, but NCC (1978),
Davis (1983) and Rothery, Mullally and Byrne (1976) all provide very
readable introductions to the subject.
It is also at this stage that the most contentious decisions may have to be
taken. Many laymen might see the function of the systems analyst as being
merely to computerise existing manual processes, but this may be very far
from the truth. NCC (1978:p. 33) says:
The systems analyst is concerned with analysing systems with a view to
making them more effective either by modification or by substantial
redesign.
and later, p. 43:
The systems analyst is an agent of change ...
The systems analysis is likely to be much more concerned with the input to
a system and the output from it rather than the manual process linking the
two. It is therefore highly likely that the analyst may propose alternative
and more economical ways of accomplishing the particular process, and
this may have a very significant effect on existing jobs, techniques and
work practices and hence upon the degree of acceptance of the final
system by the end user.
As an example, imagine that one were to ask a systems analyst to
investigate the bill production process with a view to maximising the
efficiency of the use of computers and bearing in mind possible future links
with computer aided design systems. The result of such an analysis might
well be a system which might work by constructing a mathematical model
of the building inside the computer from which the computer itself could
measure its own quantities. Such a system would obviously require a

totally different approach to measurement than has been conventionally
used in the past; it might for example require the quantity surveyor to
measure and input to the computer all of the measurements for the
building on a wall-by-wall basis rather than by using the more conven-
tional 'girthing' and elemental measurement techniques commonly used at
present. Such a system might be extremely efficient, but would the changes
to established practice be readily accepted by the rank and file of the
profession among whom more traditional techniques are ingrained? One
must therefore look very carefully at radical solutions, bearing in mind
always that any new system requires the wholehearted support of its users
if it is to be successful.
The detailed study process is essentially an iterative one requiring
considerable discussion between the client and consultant. It can be a
process in which the lay client feels most uncomfortable perhaps because
he has considerable difficulty in visualising how the analyst's promised
vision of the future will eventually appear. Such problems, if not tackled as
they arise, may only surface when the finished system is unveiled and the
clients reaction is 'Oh dear I didn't think it would be like that!' It is in an
attempt to overcome this problem that some commentators recommend
that 'prototypes' or 'mock-ups' of the user interfaces should be constructed
during this stage of the work. Clients can then be shown proposed screen
displays, output and report formats as the design progresses and can
comment as required. The use of techniques such as this can be very
helpful in that they enable the client to feel 'in-touch' with the project and
a part of the project team. The use of prototyping as a technique is
discussed in lnce (1988a), and some of the major problems are summarised
in lnce (1988b). lnce (1987a) discusses the management problems inherent
in the use of the approach.
The result of the detailed analysis phase will be a report detailing one or
more firm system proposals, each complete with forecasts of time and cost.
Davis (1983:p. 12) writes that the analyst may be asked to consider, as a
minimum:
(1) A low cost option which does the job and nothing more.
(2) An intermediate cost solution that does the job well, and is
convenient for the user. This system may include several features
that the user did not specifically request, but that the analyst,
based upon experience and knowledge, knows will prove valuable.
(3) A high cost 'Cadillac' system, with everything the user could
possibly want.
In the final analysis the decision again rests with the client whether to
continue or to abandon the project. If the rlecision is taken to continue
then one of the alternatives will be selected and this conceptual model will
form the blueprint for the physical system.
In practical terms completion of this stage represents the latest point for
the incorporation of major variations to the philosophy of the system.
Major structural changes beyond this point will have disastrous effects
upon the project in terms of time and cost overruns. It is therefore essential
for the philosophy of the system to be fully and finally agreed before
proceeding to the next stage of detailed systems design.
Detailed systems design
Once the detailed study is complete and the system proposal is agreed
then work may begin on the detailed systems design. As mentioned earlier
some commentators prefer to consider this phase as partly creative, but I
prefer to consider the design process to be essentially a technical one
concerned with the design of files, various computer processes, the detailed
selection of equipment etc. The detailed design stage is therefore analogous
to the 'working drawings' stage of a construction projea; it resolves the
technical problems inherent in converting the conceptual design into a
physical reality. The output from this stage is therefore a complete design
for all aspects of the proposed system which can then be passed to a
programmer for conversion into a computer language.
It may still be possible to accommodate minor variations during the
systems design phase but usually, as in a building project, only at the
expense of additional time and cost.
Programming
Upon completion of the detailed design phase, or perhaps with some large
systems overlapping with it, comes the process of converting the analyst's
paper design into a working system; it is therefore analogous to the
construction stage of a building project. The choice of programming
language will already have been made by the analyst and the programming
task is almost entirely a technical one.
In the early days of computing, programming was considered to be
essentially creative. Tiny amounts of memory, extremely slow processors
and very limited backing storage meant that considerable time and effort
were expended in devising ingenious ways to use the computer most
efficiently. Early system design techniques also left much to be desired,
and tended to mean that much of the design was done by the programmer
himself on a trial and error basis, with the result that programming was a
very hit and miss affair with many changes of direction. Such systems thus
11 0 Computers and Quantity Surveyors
tended to be extremely difficult to follow, were usually poorly documented

and were consequently a nightmare to subsequently maintain.
Today however there is less of a need for 'clever' programming of this
kind. The cost of hardware is now so low compared with software develop-
ment cost that, with very few exceptions, complex and ingenious programs
are regarded as a distinct liability. Improved system design techniques also
contribute to better prepared systems and programming today is therefore
no longer the uncertain process of only a few years ago.
Large systems will usually be constructed from a large number of sub-
programs or 'sub-routines' each of which is written and tested individually
by the programmer before being incorporated into the system as a whole.
Unfortunately testing has, in the past, tended not to be one of the high
profile activities of software development. lnce (1988c), in a discussion of
automated testing methodologies, writes:
Testing is the Cinderella of the software project. Whilst text-books and
theoreticians place a large amount of stress on front-end activities, such
as specification and design, those activities traditionally associated with
testing have been ignored.
This is plainly unsatisfactory.
Upon completion of the programming stage the whole system will be
tested by the analyst using test data usually supplied by the client in order
to confirm that the system as a whole meets the originally agreed specifica-
tion. Only when the analyst is completely satisfied that the program
performs correctly in accordance with the agreed design criteria will he
present the system to the client for implementation.
The effect of any variations at this stage will inevitably lead to major
time delays and cost increases. In all but the most minor cases variations
are likely to mean that the programming stage is abandoned in order that
the analyst may reconsider the system design.
Whilst the programmer is actually writing and testing the individual
programs the analyst will usually be writing up the detailed system
documentation. Accurate and complete documentation is of course essential
to ensure ease of maintenance and possible future upgrading.
Completion of the programming stage should therefore result in a
complete working and fully documented system.
Implementation
In a perfect world implementation of a new system would simply be a

matter of switching on the machine and handing the user an operations
manual. In practice of course it doesn't work out that way, and there are
inevitably snags which will require to be resolved before the system settles
down, staff become familiar with it and the process begins to run smoothly.
The management problems associated with introducing new systems have
already been documented in chapter 3.
The ease with which a new system is implemented is a direct reflection
of:
(i) the level of commonality and true agreement reached by the
analyst and the client during the requirements analysis phase,
(ii) the skill displayed by the analyst in translating the agreed require-
ment into a working system and
(iii) the level of management expertise employed during the design and
implementation stages
As noted earlier, it is in the first of these that the analogy with a construction
project begins to break down. Construction projects are three-dimensional
things which can be physically represented both on paper as architectural
and artistic drawings, and in three dimensions by models. It is therefore
comparatively easy to see whether the designer has correctly interpreted
the client's brief. In contrast, computer programs are intangible ethereal
things, and it is only at this final stage that the accuracy of the initial
requirements analysis is eventually revealed.
Maintenance and review
The final stage in the system life cycle is the maintenance and review
phase. Even if the previous phases have been completely successful and no
amendments are required at initial implementation, most computer
systems will need modification at some stage during their life, perhaps
because of changes in legislation, for example a payroll system, or because
of changes in methods of working, for example the introduction of a new
Standard Method of Measurement.
Typically, most well designed systems can cater for some amount of
change provided that the change does not strike at the basic philosophy of
.the system itself, but there will eventually come a point at which the extent
of the changes, or the cumulative effect of a number of changes carried
out over a period of time mean that a redefinition of the problem is
required and the system life cycle must therefore begin again. The critical
management activity must be in recognising when this point occurs; it is
very easy to fall into a position where continued maintenance of an
outdated and over-complex system imposes an ever-increasing main-
tenance burden and yet the very high level of investment in the system
deters management from throwing it away and beginning anew. A good
example has to be the case of the British air traffic control system reported
in Computer Weekly (1987). The article reports that because of a large
number of computer system failures the existing thirteen-year-old system

was to be replaced but:
The Civil Aviation Authority is already committed to spending £200m
refurbishing the West Drayton control centre and a new computer will
be the biggest single item on the shopping list, but the software is so
unwieldy that it is unlikely to be replaced. Independent experts are to
examine the million lines of code which make up the air traffic control
system ... The colossal investment means they will have to find some-
thing seriously wrong with the software before CAA decides to chuck it
out of West Drayton. Starting from scratch is not an attractive idea.
Management of the development process
Bell (1987) and Neale (1987) both point out that, although software is an
intangible thing, the management of software projects is not really that
different from the management of any other project be it the construction
of a building, building a ship or planning a production line. In essence the
same management tasks apply; those of planning and controlling the use of
the available resources in order to achieve successful completion of the
project within the originally agreed parameters of time, cost and quality. It
would therefore follow that the kind of management techniques used in
other industries would be likely to be equally applicable to the develop-
ment of computer software.
We should therefore consider:
(i) forecasting of time and cost
(ii) economic justification
(iii) monitoring techniques
Forecasting of time and cost
Initial forecasting of time (and cost) is even more difficult for a computer
system than it is in the early stages of a building project. Bell (1987:p. 76)
reports that:
Error factors of 200 per cent or 300 per cent are not unknown and worse
figures are known.
Historically, managers have either based their estimates of time on previous
projects or have used some form of formula. The use of previous project
data requires an initial appraisal of the scheme to be completed, together
with an outline assessment of the various tasks to be completed. Data from
The Development of In-house Systems 11 3
previous jobs is then applied and cost, resource and time forecasts are
produced. lnce (1987b) calls this the 'pragmatic method', and it is of
course very similar to the historical cost approach for early estimates of
building cost. This approach has in some cases been computerised and
systems have been built to automatically generate time scales for new
projects based on such historical information. lnce (1987b) reports on the
use of such a tool, called Cocomo, developed in the United States.
Alternative estimating methods use formulae of one sort or another. One
example, the use of which is reported in lnce (1987b), is called 'Function
Point Analysis' and was originally developed at IBM in the late 1970s. The
method works by measuring the proposed system in terms of the functions
required; the number of inputs, outputs, enquiry types, master files and
interfaces required in the complete system. Each total is then given a
weighting and a formula applied and the resultant number gives an indica-
tion of the time a new project will take. It is then a simple matter to
combine the time with existing known staff costs to calculate the total
estimated cost of the project.
Computer software development also has other parallels with the
construction industry in that any time or cost figure given to the client, no
matter how sketchy the information and no matter how approximate the
figure, automatically becomes the maximum allowable and no matter how
accurate subsequent predictions are the first figure given is the only one
which will be remembered as if engraved on a tablet of stone! It is
therefore essential that a sufficient contingency sum is allowed to take
account of problems which might arise.
Bell (1987:p. 44) gives the following list of possible problem areas and
suggests that the more items from the list that apply to the project the
greater the chance that the project will not proceed as forecast:
(i) Large team involved.
(ii) Long time scale.
(iii) Need to sub-contract.
(iv) New skills required.
(v) Senior management commitment unknown.
(vi) Users not experienced in system projects.
(vii) Large change in the user area.
(viii) Several distinct users involved.
(ix) Change in users' power position.
(x) New computer equipment.
(xi) New terminals or communications required.
(xii) New operating system.
(xiii) New language or environment.
(xiv) Several equipment suppliers involved.
Economic justification
The development of a new computer system must be looked upon as an

investment, and will normally need to be justified by the use of normal
investment appraisal criteria. It is not within the scope of this work to give
detailed explanations of investment appraisal techniques, but Bell (1987)
gives examples showing the use of payback analysis, break-even analysis,
net present value and internal rate of return. The economic justification of
projects using techniques of this kind is usually a crucial factor in manage-
ment's decision to abort or proceed.
Monitoring techniques
The successful completion of software development projects requires just as

much carefully planning and monitoring as the construction of a building,
and has historically used many ofthe same tools including bar or Gantt charts
and the many different kinds of network planning in common use in
construction. Neale (1987) however points out that the use of such techniques
has not been entirely successful. As a response to the apparent failure of more
traditional tools the software industry has introduced the concept of the
Integrated Project Support Environment (IPSE) which attempts to provide a
computer system which will itself aid the planning and monitoring of
software projects. The concept is still very new and available systems are thin
on the ground although Jones (1988) reports on the use of two systems
available in the United Kingdom.
References
Bell R. (1987) Management of systems development Hutchinson

Computer Weekly (1987) 'Failures force Hobson's choice' Computer Weekly 20.8.87 pp. 14-15
Davis W.S. (1983) Systems analysis and design- a structured approach Addison-Wesley
Ernest-jones T. (1988) 'Learning the hard way about fourth generation' Computer Weekly
7.4.88 p. 16
Gooding C. (1988) 'Time for the proof of the pudding' Computer Weekly 18.2.88 pp. 26-27
lnce D. (1987a) 'Model answers' lnfomatics September 1987 pp. 61-62
lnce D. (1987b) 'Projecting the future' lnfomatics October 1987 pp. 49-50
lnce D. (1988a) Software prototyping in the Eighties Privately published by the author at 39
High St., Greens Norton, Northants
lnce D. (1988b) 'Moving on from the middle ages' Computer Weekly 11.2.88 p. 32
lnce D. (198&) 'Putting software to the test' lnfomatics March 1988 pp. 52-53
Jones K. (1988) 'IPSEs start to come into their own' Computer Weekly 4.2.88 pp. 24-25
NCC (1978) Introducing systems analysis and design NCC Publications
Neale A. (1987) 'Managing "intangible" projects' .EXE Magazine November 1987 pp. 4-10
Rothery B., Mullally A. and Byrne B. (1976)The art of systems analysis Business Books Ltd.
8 Expert Systems - the
Future?
Objectives
At the end of this chapter you will be able to:

• Explain what is meant by 'artificial intelligence'.
• Explain the meaning of the terms 'decision support system' and 'expert
system'.
• Explain the various terms used to describe different sorts of expert
system.
• Discuss the various kinds of problems which expert systems can help
to resolve.
• Discuss the application of expert systems to quantity surveying, the
construction industry and the related professions.
• Discuss the advantages and limitations of the use of expert systems.
Decision support systems
During the late 1960s -and early 1970s a number of factors came together
to radically change the way in which computers were used in business and
commerce.
In terms of the computer hardware, interactive terminal-driven systems
were becoming the norm, and mainframe computers had grown very large
indeed with the capacity to store and manipulate vast quantities of informa-
tion. Minicomputers were also well established both as stand-alone
machines in smaller companies and also as front-end processors to larger
and more complex mainframe machines. The microcomputer, although
still regarded as a hobbyist's toy, had also begun to appear as a direct spin-
off from the space race of the 1960s. The dramatic increases in both
commercial computer usage and available computer power meant that
virtually limitless quantities of information could be stored, and the use of
terminal meant that the whole of that information could be made available
to executives and managers at the touch of a button in the shape of a so-
called computerised Management Information System (MIS).
The only problem in this corporate garden of Eden was the surfeit of
information, all of which the executive decision makers were expected to

absorb, evaluate, analyse and incorporate into their decision-making
process, an excess of data which most managers found very hard to cope
with. Keen and Scott-Morton (1978:p3) wrote:
During this period the utopian dream (in retrospect, perhaps demonic
would be a better term) of the total integrated database emerged. For
example, some organisations aimed at storing all the company's data in
a huge set of files, that could be accessed through a computer. This 'data
centred' approach, carried to such an extreme, was neither useful nor
feasible.
In short, management was in danger of being totally swamped by the
information revolution; the dilemma is well documented by, among others,
Dearden (1966), Nolan and Gibson (1974) and Kanter (1977).
During the same period a number of management scientists and behav-
ioural psychologists were involved in studying the process of human
decision making, a field of study which developed models attempting to
both explain and assist the human decision making processs. These models
eventually impacted strongly on both operational research and manage-
ment science.
From the synthesis of these models and the enhanced information
storage and recall facilities offered by the new generation of computers
sprang the concept of a computer based Decision Support System (DSS).
Such systems would no longer be simply huge information stores which
overwhelmed the manager with both vital and useless information with
equal zeal, but computer systems which would be specifically developed
to allow the available information to be filtered so that the computer could
actually assist the decision-making process.
Keen and Scott-Morton, building on the work of Anthony (1965) and
Simon (1960), show that all management tasks may be defined as either
'structured' (e.g. linear programming of a construction project), 'semi-
structured' (e.g. the preparation of an early estimate from very sketchy
information), or 'unstructured' (e.g. hiring staff). Note that the taxonomy
makes no implied judgement about the degree of 'difficulty' involved in
each type of task, merely about the level of 'human intuition' which must
be applied and thus, by implication, the presence or lack of 'rules' or
models upon which to base a decision. Obviously the more highly
structured a task becomes the greater the likelihood that a computer system
could be developed which would follow the rules; hence computer systems
should be able to replace managers only in those areas where the task is
highly structured. A good example might be the day-to-day running of a
complex manufacturing plant which used to be considered a difficult and
challenging management task but is now commonly controlled completely
by computers.
Expert Systems - the Future? 11 7
It is therefore in the areas of semi-structured decision making that the use

of a computer based decision support system may perhaps be of the
greatest benefit. Decision support systems therefore concentrate on aiding
the decision maker in tasks where managerial judgement is still an essential
element.
Moore and Chang (1983) point out that, in essence, virtually any
computer system which aids decision making could qualify as a decision
support system. Decision support systems might therefore range from, for
example a spreadsheet model allowing 'what if' simulation to complex
information bases allowing selection of information on the basis of user
supplied criteria. They do however go on to propose a taxonomy which
attempts to place decision support systems in context with other computer
systems. They provide the following definitions:
Transaction Processing Systems (TPS)- pure data processing for gathering,
updating and posting information according to pre-defined procedures.
The TPS therefore involves no management decision making since most
decisions are clerical and routine.
Management Information Systems (MIS) - systems with pre-defined
data aggregation and reporting capabilities. MIS reports are usually
either printed in batch or queried on demand. However if present,
query options normally include only a set of pre-defined data extraction
operators. The BCIS 'On-line' system might therefore be classified as a
management information system.
Decision Support Systems (DSS) - extensible systems with intrinsic
capability to support ad hoc data extraction, analysis, consolidation and
reduction as well as decision modelling activities.
Keen and Scott-Morton (1978) give a number of examples of working
decision support systems, and Bennet (1983) details the factors to be taken
into consideration in creating new systems.
Some researchers, for example Corry and Krumland (1983) have argued
that decision support systems should take more advantage of research into
artificial intelligence in order to promote the use of decision support
systems as consultants to the decision maker rather than merely as passive
information providers. Such research leads directly to one facet of what
have become known colloquially as 'expert systems'.
Artificial intelligence
The concept of 'artificial intelligence', the construction of an 'intelligent'

machine, is a subject which has captured the hearts and minds of men for
hundreds of years. In the main of course, such things lie in the realm of
science fiction, but in the years since the end of the second world war the
pursuit of 'artificial intelligence', AI for short, using computers has become

a widely researched area of study.
In order to make any sense of the concept it is first necessary to define
what 'intelligence' actually is. Typical dictionary definitions speak of 'quick-
ness of understanding', 'intellect', 'capability for rational thought' etc.;
words which even given the most liberal of meanings could not be applied
to any computer system yet in existence or even foreseeable in the near
future. Despite the fact then that a considerable body of so-called artificial
intelligence research now exists, the word intelligence, in this context,
generally implies nothing more than an ability in some sense to 'learn';
that is to say that the computer system is able to evaluate the results of its
own actions, and to modify its own behaviour in order to ensure that
incorrect or invalid actions or responses are not repeated. Simple examples
might for instance include a computer controlled industrial robot equipped
with impact sensors which is required to perform some sequence of
operations in a factory. Such a device could be made to modify its own
behaviour pattern by taking input from the sensors whenever contact was
made with some adjacent object, and using that information could construct
a 'map' of its surroundings in order to prevent accidental collisions. The
machine can thus be said to have 'learned' from its own experience; in
some limited sense to have displayed some intelligence. Another example
might be a computerised version of the guessing game 'Animal, vegetable,
mineral', where the computer attempts, on the basis of answers to questions
asked, to deduce the identity of some object, animal, plant etc. by means
of a decision tree. In the event of an incorrect guess the system can 'learn'
by asking how it could discriminate between the incorrect guess and the
correct answer, and the new information is then added to the computer's
internal decision tree.
Such examples are trivial and yet they do indicate the extremely wide
range of artificial intelligence research. Many of the problems associated
with the expansion of artificial intelligence are technical rather then
theoretical, and are concerned with the simulation of senses which we
take for granted. The construction of manipulative robots with the range of
movement and delicacy of touch of a human hand, or the construction of a
computer system which can process complex visual signals as well as the
human eye and brain are still well beyond the frontiers of modern
engineering, but there is little doubt that the engineering problems will be
solved sooner or later.
It is however very difficult to conceive of a machine capable of original
thought in the human sense. Some eminent AI researchers have suggested
that the phenomenon of original thought, at least in the sense we under-
stand it, may only be relevant to us as human beings within our own frame
of reference, and that even if true intelligence, original thought, of this kind
were possible within a computer system, the conversations which
Expert Systems - the Future? 119
computers might initiate between themselves might be both meaningless

and uninteresting to us. It seems therefore unlikely that the popular science
fiction vision of the all powerful computer ruling the world will ever
actually be fulfilled.
Expert systems - an overview
Perhaps the most widely acclaimed off-shoot from artificial intelligence

research has been the development of a range of computer software, based
on specialist knowledge, which attempt to reach some conclusion, to solve
some problem, by means similar to those employed by a human expert.
Such systems have become known as 'expert systems'.
The concept of expert systems has evolved in response to the need,
partially articulated by researchers engaged in the development of improved
management information and decision support systems, to develop com-
puter systems which can interact with users in a conversational way, and
which can not only recall information from a database but can also
develop and pursue a logical line of reasoning. Such systems ought to be
able to deduce information from combinations of knowledge already
stored in the database and replies given by a user in order to reach
conclusions which may not be readily apparent from the raw data originally
stored in the system.
The first computer program to be explicitly recognised as an expert
system was probably DENDRAL, a system developed at Stanford Research
Institute from about 1965 onwards for the inference of chemical structures
from mass spectrometer readings. Much research has been carried out
since, and Bramer (1982) contains a useful review of the first fifteen years.
The current intense interest in the subject originally stems from the so-
called Japanese Fifth Generation Project so widely publicised by people
such as d'Agapeyeff (1983) and Feigenbaum and McCorduck (1985).
In the United Kingdom the study of artificial intelligence has had a
chequered existence. Whilst research has been on-going in one form or
another since the end of the second world war, development in the United
Kingdom was virtually strangled by the lighthill Report, Lighthill (1973),
which argued that although systems could be constructed which work in a
laboratory, the chances of such systems working satisfactorily in real life
situations was small. It was not until the early 1980s that Britain again
began to take an official interest in a subject which was already hot news
in both America and Japan. The Alvey programme was launched in 1983
in a desperate attempt to catch up with others in the field.
A good 'potted history' of artificial intelligence research is given in
Durham (1984), and a more detailed review of earlier work in Michie
(1982).
Definition of·an expert system
Before attempting to explore the possible uses of expert systems in quantity

surveying and construction, it is first necessary to attempt to define in more
precise terms what an expert system is. As with any developing field there
have been many attempts to provide formal definitions, ranging from the
complex:
An expert system is a means of capturing the knowledge of experts in the
form of programs and data where disagreement among the experts are
settled by mediation and the results refined so as to extract the essence
of their knowledge in such a way that it can be used by less experienced
people within the field ... The expert system is a tool and a means of
coherent communication of the latest views of the experts to the users
who may well be the experts themselves.
(Addis (1981))
to the rather more simple:
[Expert systems are) ... problem solving programs that solve substantial
problems generally conceded as being difficult and requiring expertise.
They are called 'knowledge based' because their performance depends
critically upon the use of facts and heuristics used by experts.
(Stefik (1982))
Perhaps the most definitive definition is that approved by the British
Computer Society Committee of the Specialist Group on Expert Systems
reported in d'Agapeyeff (1983):
An expert system is regarded as the embodiment within the computer of
the knowledge based component from an expert skill in such a way that
the system can offer intelligent advice or take an intelligent decision
about a processing function. A desirable additional characteristic which
many would consider fundamental is the ability to justify its own line or
reasoning in a manner directly intelligible to the enquirer. The style
adopted to attain these characteristics is rule based programming.
although Naylor (1983) sees things in much more simple terms:
To take an everyday example, there were once people who were payroll
experts. In fact some of the most agile minds in the British Empire could
have been found in the Army Pay Corps. Adept at finding loopholes,
special cases, hitherto unknown allowances, these human exf>ertS (some
would say 'superhuman') held sway for years. It was simply the payroll
program which sent pay clerks the way of the dinosaur. The payroll
program can easily be seen as an example of an expert system - it
embodies the total sum of human expertise on the subject of payrolls.
But nobody thinks of it like that anymore.
Since we cannot find a comprehensive definition of an expert system,
perhaps it would be best to attempt to establish what the charaderistics of
such a system ought to be. The following is a shortened and paraphrased
version of the characteristics identified by lansdowne (1982), although he
is careful to point out that not all systems display all of these properties:
(i) They know a great deal about a limited but useful area of interest.
The knowledge may be aquired from experience but, more likely,
from a human expert.
(ii) They give their advice conversationally in the manner of a human
consultant and can understand and respond to simple questions
posed in plain, although somewhat specialised, language.
(iii) Their knowledge is embodied in the form of sets of rules with
corresponding adions. This feature makes for easier correction of
deficiencies or errors in the knowledge base as well as the acquisi-
tion of new knowledge. The knowledge usually exists indepen-
dently of the program and this should make it possible to use
the same program with a variety of knowledge bases able to deal
with different tasks.
(iv) Expert systems often give their advice in probabilistic rather than
absolute terms.
(v) The questions posed by expert systems are limited to those which
are relevant to a particular line of reasoning. Thus if at any time the
system decides that it has sufficient information to arrive at a
conclusion it does not continue to ask questions.
(vi) Expert systems can explain and justify their reasoning in such a
way that experts can accept their credibility and non-experts can
learn from them.
Generally research seems to indicate that expert systems may be of most
uses where:
(i) The performance of the task is more dependent upon factual
knowledge rather than computational method.
(ii) The area of interest is limited and specialised.
(iii) The knowledge base can be built up gradually over a period of
time.
(iv) Consultants exist and are available for consultation.
Types of problem
It has already been shown that management tasks may be sub-divided into
a number of overlapping groups depending upon the degree of human
intuitition needed to resolve them, and that much work has been carried
out in the construction of computer based decision support systems aimed
at helping the manager make the most appropriate decision in each case. It
has also been shown that the more highly structured a task becomes the
more likely it is that a complete computerised solution will be possible.
The types of problem for which computers may provide some assistance
may be classified as:
Problems of classification
In this type of problem something needs to be identified or classified upon

the basis of observable characteristics, possibly without 100 per cent
certainty. This classification will include problems such as medical diag-
nosis, for which some of the most successful expert systems have been
developed (see for example the MYCIN system described in Shortliffe
(1976)). In the construction industry similar problems for which expert
systems have been developed include the diagnosis of faults in air-
conditioning systems (Lansdowne (1982)) and the diagnosis of the causes
of damp in buildings by the Building Research Station.
Problems of selection
In this case the problem is to select components from a wide standard

range to fit a given set of circumstances. In effect this type of problem is
the opposite of the classification problem described above. One well
known example is the XCON system developed to configure DEC VAX
computer systems, and it is not difficult to see that similar systems could be
used for the selection of standard components in prefabricated construction
systems.
Problems of data retrieval and interpretation
Many situations are governed by well defined but often extremely complex
rules, and the solution to problems can often become clear once the rules
are understood. Perhaps the most obvious, and the most complex, mani-
festation of this type of problem is the Law, where considerable volumes of
statute and case law form a large body of rules which guide practitioners in
the resolution of disputes. Oakley (1987) gives an introduction to the use
of expert systems in law, and a more detailed account is given in Susskind
(1987). Oakley (1987) also reports on an expert system developed by
Susskind and Capper called the Latent Damages Advisor which attempts to
unravel the mysteries of the Latent Damages Act 1986. Other similar
obvious contenders in construction include the various Standard Methods
of Measurement, Building Regulations, and the various Planning Acts.
Such systems would differ from simple data retrieval systems by being able
to respond to queries such as 'List the measurement rules in SMM7 relating
to the measurement of concrete formwork' or 'Display the criteria to be
satified for a contractor to succeed in an application for payment of loss
and expense under the JCT80 form of contract'.
Problems of prediction
Given a set of present circumstances and a knowledge base of past events

predict the likelihood of a given event occurring. Naylor (1983) for example
includes a simple weather forecasting system which attempts to predict the
likelihood of rain tomorrow based on historical data. Mathematical pre-
dictive techniques are well known (see for example Green (1980)), and it is
easy to see how systems could be constructed to attempt, for example, to
predict the future cost of buildings.
Structure of an expert system
Figure 8.1 shows a very simplified representation of what actually happens

in real life when a human expert is confronted by a problem. In simple
terms the problem is cross-matched with data held in some knowledge
base and the result of the merger then governs the advice which is given.
Although the above models fairly accurately the way in which a problem is
considered in human terms it is not a good model from which to build a
comparable computer system.
In practice of course the expert is always expanding his or her knowledge
base by reading technical journals, discussions with colleagues etc. This
updating of the knowledge base is likely to take place in very technical
language which would be largely meaningless to a non-expert. On the
other hand however communication between the human expert and the
human enquirer is likely to use much less technical language. The computer
system therefore needs to make some provision for both modes of commun-
ication. In addition human language of any kind is extremely rich and wide-
ranging; similar ideas may be expressed in many different ways and the
computer system must therefore somehow select the important parts of the
problem from the input dialogue. A better representation therefore of an
expert system, adapted from Feigenbaum and McCorduck (1985), is given
in figure 8.2.
0 0 0
Figure B. 1 The human expert
Elements of an expert system
The knowledge base
The knowledge base is the central reservoir from which the system will
ultimately formulate its advice and opinions. The contents of the knowledge
base therefore need to comprise far more than mere facts or information.
According to Feigenbaum and McCorduck (1985:p. 7), knowledge is ' ...
information which has been pared, shaped, interpreted, selected and trans-
formed', and in order to achieve this the knowledge base needs to contain
not only the facts about the relevant 'domain' or sphere of activity, but also
a representation of the 'weighting' which the human expert would give to
those facts; the experience which enables him to provide a balanced
judgement or reach a reasoned conclusion about the facts of any particular
case. In short the knowledge base must firstly contain a set of 'rules'
embodying the facts of the domain, secondly a set of 'goals' denoting
possible results and thirdly some mechanism to relate the 'rules' to the
'goals'. The simplest form is obviously shown by the
IF(rule) THEN(goal)
construct indicating that a certain outcome will occur if a certain rule is
satisfied. It is sometimes extended to
IF(rule) THEN(goal probability x)
indicating that some uncertainty exists. A number of different mechanisms
have been developed and are simply explained in Landowne (1982).
The inference system
The inference mechanism forms the part of the expert system which
connects the problem with the information stored in the knowledge base.
A number of different mechanisms have been developed, but perhaps the
A USERS
(MANY)
---,
KNOWLEDGE INPUT/OUTPUT
~UISITION S'f"STEM • SYSTEM
COMPUTER
SYSTEM
KNOWLEDGE INFERENCE
BASE SYSTEM
L ______ _ _ _ _ _ _ _ _j
KNOWLEDGE
ENGINEER
Figure 8.2 Elements of an expert system

two main methods are known as 'forward chaining' and 'backward

chaining'.
Forward chaining systems are very effective where the process consists
essentially of explaining a complex body of data, for example a system to
interpret contract clauses, methods of measurement etc. The inference
system is then able to ask questions and establish to what extent the
problem data conforms to the rules as declared. In other words, by steadily
building upon the information already established the computer is able to
move forward through the information base until a conclusion is reached.
Forward chaining systems therefore attempt to establish the left-hand side
of the
IF(rule) THEN(goal)
statement. Backward chaining systems on the other hand attempt to discover
whether a specific situation ('goal') exists, and to therefore fire a rule which
would produce that result. They therefore attempt to establish whether the
right hand side of the
IF(rule) THEN(goal)
statement is true. Such systems are often described as 'goal driven', and are
very useful where the observable information comprises the resultant of
some action. Systems attempting to resolve problems of diagnosis often
work by backward chaining.
NCC (1987) contains both written examples and computer software
illustrating several different types of inference mechanism.
The user input/output system
Expert systems are supposed to be friendly, to help people solve problems.

In order to do this effectively people must be able to communicate easily
with them, and the system must be able to explain the reasoning behind its
conclusions or advice. For many specialists in the field this has meant the
use of natural language, and considerable research has been undertaken in
this area.
Basically the problem underlying the machine understanding of natural
language is quite simple. People using everyday speech employ an
enormously wide range of different ways of saying the same thing, of
asking the same question, and the construction of computer programs
capable of parsing language correctly is extremely difficult. Lansdowne
(1982) summarises the main problems very well. In order to overcome
these problems many expert systems resort to the use of questions posed
by the machine, to which the user is required to make only simple, usually
one word, answers.
Most expert systems rely upon the use of a normal keyboard, although
Expert Systems- the Future? 127
other methods of input are possible. Lansdowne (1982) for example

proposes the use of an electronic tablet rather like a digitiser menu, and
one could easily foresee the use of speech recognition devices in the future.
The knowledge acquisition system
Most experts now agree that the initial knowledge acquisition is best
accomplished by a 'knowledge engineer'. Knowledge engineering is a very
new skill which is still not completely defined, but the task of the knowledge
engineer is to extrad the necessary relevant knowledge from the experts
and build up the knowledge base of the system.
In general terms the task can be sub-divided into:
(i) A background study. The knowledge engineer needs to study the
background material on the subject and become familiar with the
jargon in order both that he understands what the expert says, and
also in order that the expert's time is not wasted by needless
explanation.
(ii) The knowledge engineer must discuss the problem domain with
the expert in order to understand the domain boundaries.
(iii) The knowledge engineer must select a problem solving strategy
suitable for the particular domain under consideration.
(iv) The knowledge engineer then 'extrads' the appropriate knowledge
from the experts and develops an early 'test' version of the system.
(v) Ask the expert to 'use' the draft system and determine whether the
output is satisfactory. Ask other experts to use the system and
collect a range of opinion.
(vi) Refine the system by repeating steps (iv) and (v) until the experts
agree that the result is satisfadory.
Feigenbaum and McCorduck (1985:p. 83) report the development of the
following 'rules' for knowledge engineering:
(i) It is very difficult for domain experts to be their own knowledge
engineers; an outsider is required in order to see the problem
clearly.
(ij) The knowledge engineer must be prepared to throw things away;
the first draft will never be acceptable.
(iii) Problems must be carefully chosen and clearly bounded with a
need for specialist knowledge.
(iv) Knowledge engineers must be flexible, and good communicators
on many levels.
(v) If no tools exist to do what is required then build new ones.
(vi) The knowledge engineer needs to develop weighting procedures
to cope with statements such as 'I strongly believe ... '.
(vii) The knowledge base must be readily modified.

(viii) The expert system must explain its own reasoning.
Expert system shells
Although the early expert systems such as MYCIN were special purpose
tools developed to fulfil specific tasks, it quickly became apparent that the
same problem solving process could be used for many different problem
domains; that it was possible for the 'inference system' to be, to a large
extent, independent of both the actual problem data and the domain
knowledge base. Development was therefore concentrated on the con-
struction of 'empty' expert system shells which could, once the domain
knowledge base had been added, be used as general purpose problem
solvers.
A large number of such shells have been developed in recent years
employing a range of different problem solving strategies. As explained
earlier some strategies are better at solving some problems than others, and
Allwood, Stewart, Hinde & Negus (1985) contains an evaluation of the
suitability of various shells for use in the construction industry. NCC (1987)
contains examples of a number of different types of system shell.
Expert systems in quantity surveying and construction
In addition to the work cited above, Wager (1984) and Wager (1985)
contain good reviews of the use of expert systems in the construction
industry as a whole, and both Hamilton (1985) and Shaw (1985) discuss
the use of expert systems in building services. It is however in quantity
surveying that perhaps the most significant strides have been made to date.
In 1983 the government launched a five-year programme of research

under the chairmanship of John Alvey into the possibilities for the use of
fifth generation computer systems in British industry. Research initiatives
were invited from industry under the provisions of the Alvey programme,
and the only successful construction industry submission was made by the
RICS for research into the use of expert systems in quantity surveying.
Brown (1987) and Davis (1988) describe the research, which was based at
Salford University under the supervision of Professor Peter Brandon guided
by a group of selected practitioners.
The intention of the project was to investigate the use of expert systems
in the stategic planning of construction projects. It was decided to consider
four major aspects of the problem viz:
(i) Financial budgeting and initial cost estimating

(ii) Choice of procurement method
(iii) Forecast of construction time
(iv) Development cost appraisal and profitability
The resultant programme called Elsie (from LC - Lead Consultant), was
finally unveiled in the Spring of 1988 for use by members of the Quantity
Surveying Division of the RICS. A draft summary report, Brandon (1987b),
was published at the end of 1987, and the final report by Brandon, Basden,
Hamilton and Stockley (1988) in mid 1988.
The basic operation of Elsie is described in Ashworth (1988 a and b).
Also of interest to those contemplating building expert systems is the
practical experience described in Hamilton (1987) and Stockley (1987).
The effect which the widespread use of expert systems will have upon
the professions has been the subject of much comment and discussion.
Evans (1979) for example wrote at length about the profession's probable
decline of influence as the use of computer based information systems
becomes more widespread. It is also interesting, in the light of the RICS/
Alvey project, to recall that, in 1984, Brandon wrote:
On the face of it, the professions as a whole seem vulnerable. Their
strength lies in being a repository for a unique brand of knowledge. If this
knowledge is enshrined within a computer program it becomes available,
in time, to a much wider audience.
(Brandon (1984))
One might reflect therefore upon whether the RICS/Aivey project will
prove to be truly a success from the point of view of the ordinary quantity
surveying practitioner. Brandon however does offer some hope of a lifeline,
albeit a treacherous one, when he goes on to write:
However some comfort can be derived from the following:
(i) Clients will want someone to take the risk for a decision that is
being made. The professional practitioner will be expected to take
that responsibilty until such time as the risk is negligible or until
the computer can be sued!
(ii) Someone will need to remain the 'domain expert' for updating the
system, although in the very long term it may be possible for self
learning systems to be developed for certain tasks.
(iii) Not all the work of a profession can be enshrined in a machine nor
will it be desirable to do so. Those aspects involving personal trust
and confidence as well as certain creative aspects of judgement
will be difficult to replace. Nevertheless it will probably require
considerably fewer personnel to provide the professional service.
(Brandon (1984))
In the final analysis quantity surveyors and other professionals must accept
and embrace the spread of 'smart' computer systems; to consider doing
otherwise would be as foolish as Canute attempting to halt the tide. If the
quantity surveying profession is to thrive however it must also evolve, must
transcend those highly structured rule based tasks for which the computer
provides an acceptable alternative in favour of those much less structured
tasks involving the exercise of human judgement. In the past a high
premium has been placed on those quantity surveyors who knew intimately
all of the rules of measurement, and could recite forms of contract, but
these are the very people who, in the future, face the prospect of being
upstaged in the same way as Naylor's payroll experts. Perhaps in the
future, although a working knowledge of the relevant rules will still be
essential, much greater emphasis will be placed on the inter-personal skills
of, for example, management, communication and negotiation.
'Intelligent' management information and decision support systems
This chapter began with a discussion on management information and

decision support systems, and attempted to show how this branch of
computer science influenced some early expert systems; the MYCIN
medical diagnosis program for example, described in Shortliffe (1976), is
generally accepted as being an expert system, but since it expresses its
results in probabilistic terms leaving the final decision on interpretation to
the consultant, it may also be seen as a decision support system.
It is perhaps therefore fitting to conclude with the idea that expert
systems may now be combined with management information to provide
managers with 'intelligent' decision support systems which will assist with
those less structured tasks involving large volumes of information to which
expert systems alone are not well suited.
Research into this process of combining artificial intelligence and conven-
tional databases is still very new; in fact as recently as 1984 Thomas
(1984), whilst discussing the theory of such systems, wrote:
There is probably no working example yet of either an intelligent DSS
(decision support system) or a Problem Solver.
More recently however Jones (1987) reports the development of a number
of such systems both in America and in the United Kingdom and it seems
likely that this marriage of artificial intelligence and management informa-
tion systems will make possible the creation of a new generation of
decision support packages to aid the manager in the 1990s.
References
Addis T.R. (1981) 'Definition of an expert system' British Computer Society Specialist Group
on Expert Systems Newsletter no.J Spring/Summer 1981
Allwood R.S., Stewart D.S., Hinde C.and Negus B. (1985) Evaluations of expert systems shells
for construction industry applications Loughborough University of Technology
Anthony R.N. (1965) Planning and control systems - a framework for analysis Harvard
University Graduate School of Business Administration
Ashworth G. (1988a) 'Elsie Plans Her Budget' Building 24/6/88 pp. 60-61
Ashworth G. (1988b) 'Go ask Elsie' Building l/7188 pp. 50-51
Bennett j. (1983) (ed.) Building decision support systems Addison-Wesley
Bramer M.A. (1982) 'A survey and critical review of expert systems research' in Introductory
readings in expert systems D.Michie (ed.) Gordon & Breach
Brandon P.S. (1984) Computers: friend or foe? Department of Surveying, Portsmouth
Polytechnic
Brandon P.S. (1987a) Building Cost Modelling and Computers E. & F.N. Spon
Brandon P.S. (1987b) Draft summary final report, Alvey community club project, the
application of expert systems to quantity surveying Royal Institution of Chartered Surveyors
Brandon P.S., Basden A, Hamilton 1., and Stockley j. (1988) Expert systems: the strategic
planning of construction projects Royal Institution of Chartered Surveyors
Brown G. (1987) 'RICS/Aivey expert systems research' Chartered Quantity Surveyor April
1987 pp. 15-16
d'Agapeyeff A. (1983) Expert systems, Fifth Generation and UK suppliers National Computing
Centre
Davis l. (1988) 'Inside Information' Chartered Quantity Surveyor january 1988 pp. 17-19
Dearden j. (1966) 'Myth of real time management information' Harvard Buisiness Review
Vol.44 No.3 pp. 123-132
Durham A. (1984) 'Fifth generation fever' Practical Computing October 1984 pp. 115-117
Evans C. (1979) The mighty micro Coronet Books
Feigenbaum E. and McCorduck P. (1985) The fifth generation Addison- Wesley
Gorry G.A. and Krumland R.B. (1983) 'Artificial intelligence research and decision support
systems' in Building decision support systems Bennett j. (ed.) Addison Wesley
Green A. (1980) 'Future perfect?' Personal Computer World November 1980
Hamilton G. (1985) Expert systems in building services Paper presented at the 1985
Construction Industry Computing Association Construction Expert Conference
Hamilton I. (1987) 'Developing expert systems for management applications' in Building Cost
Modelling and Computers Brandon P. (ed.) E. & F.N. Spon
jones R. (1987) 'Expert support systems' Computer Weekly 26/3/87
Kanter j. (1977) Management orientated management information systems Prentice Hall
Keen P.G.W. and Scott-Morton M.S. (1978) Decision support systems- an organizational
perspective Addison-Wesley
Lansdowne j. (1982) Expert systems: Their impact on the construction industry RIBA
Conference Fund
Lighthill j. (1973) Report on artificial intelligence- Report of a committee chaired by Sir james
Lighthi/1 Science Research Co unci I
Michie D. (1982) (ed.) Introductory readings in expert systems Gordon & Breach
Moore j.H. and Chang M.G. (1983) 'Meta design considerations in building decision support
systems' in Building decision support systems Bennett j. (ed.) Addison Wesley
Naylor C. (1983) Build your own expert system Sigma Technical Press
NCC (1987) Expert systems starter pack National Computing Centre
Nolan R.L. and Gibson C.F. (1974) 'Managing the four stages of EDP growth' Harvard
Business Review Vol.52 No.1 pp. 76-78
Oakley K. (1987) 'Legal eagles' Expert Systems User November 1987
Shaw M.R. (1985) Expert systems- an objective view Paper presented at the 1985 Conference
on Computers in Building and Services Design, Nottingham University
Shortliffe E.H. (1976) MYCIN: Computer based medical consultations Elsevier North Holland
Simon H.A. (1960) The new science of management decision making Harper & Row
Stefik M. (1982) et al. 'The organisation of expert systems- a tutorial' Artificial Intelligence
Vol.18 No.2
Stockley j.E. (1987) 'Knowledge acquisition for expert systems' in Building Cost Modelling
and ComputetS Brandon P. (ed.) E. & F.N. Spon
Suskind R. (1987) Expert systems in law Oxford University Press
Thomas R. (1984) 'Knowledge base techniques for decision support for management' BCS
Specialist Group on Expert Systems Newsletter No.11 june 1984 pp. 13-15
Wager D.M. (1984) Expert systems and the construction industry Construction Industry
Computing Association
Wager D.M. (1985) The future of expert systems in construction management Construction
Industry Computing Association
Coxnputers Applied to
Quantity Surveying
9 Computers in Cost
Planning
Objectives

• Discuss the use of computers for the collection of cost data and the
preparation of budget estimates.
• Comment on the use of computerised estimating systems by the
quantity surveyor.
• Comment on the use of on-line cost information systems.
• Comment on the use of computers in predicting cost trends.
• Discuss the use of computers in cost modelling
Computers in quantity surveying

Whereas the early part of this book examined a number of issues associated
with the use of computers in an office environment, this second section
now considers the use of computers, in particular quantity surveying activ-
ities. It is however most important to recognise that although, for con-
venience, I have chosen to sub-divide this part into separate chapters
dealing with cost planning, the production of bills of quantities, post
contract work, project management and computer aided design, in reality
each of these areas overlaps with all of the others and it is totally artificial
to consider any one area in isolation. Thus, for example, although a facility
for cost analysis of tender bills is often provided by the vendors of bill
production systems the collection of the data is essentially, by definition, a
post tender process, although the information collected is actually used
during the initial cost advice stage. Close associations also exist between
cost planning, bill production and computer aided design; and project
management in a way embraces the whole spectrum of activities. It is
therefore most important to recognise the links which exist and to appreciate
that each of the separate topics is in fact merely a part of a coherent whole.
Cost planning and cost modelling
Cost planning as a formal process in the United Kingdom began to emerge

during the years immediately following the Second World War, largely
because of an increasing amount of construction activity in the central and
local government funded sectors. Local authorities of the time faced not
only strict centrally imposed cash limits upon annual expenditure, but also
centrally imposed government expenditure ceilings upon the cost of all
new building. Such restrictions required that local authorities should be
able to predict tender values for specific building projects in order to make
sure that the tender eventually obtained would be within the approved
cash limit, and quantity surveyors therefore needed to develop means of
costing the architect's design.
Initially then the process was little more than an estimating function
involving costing the completed design; the easiest method was simply to
measure the quantities of work involved and price them in accordance
with rates derived from previously tendered schemes which had them-
selves proved to be within the required cost guidelines. The process was
therefore essentially linear in that the design was completed to some pre-
defined stage, the cost was checked and if found to be excessive the design
was revised.
During the 1960s, however, the search for innovative building forms
within the constraints of rigidly enforced cash limits required these cost
estimating techniques to be refined and adapted in order that the cost of
the design could be assessed as the process progressed; thus quantity
surveyors developed elemental cost checking techniques whereby the total
budget available for a project was broken down into a number of functional
elements such as foundations, external walls, internal walls etc., the cost of
each of which could be checked individually as required throughout the
design process. Cost thus began, in theory at least, to have some effect
upon the development of the design, in that if an architect wished to
overspend the budget on one part of the scheme other compensating
savings must be made elsewhere or the client must agree to revise the
budget. It was therefore considered that the distribution of cost throughout
the design could be planned. Quantity surveyors still however relied
heavily upon measurement of the work and the use of cost data derived
from previously let schemes as their major cost control tools, and in fact
Morrison and Stevens (1980) report that their 1979 survey of current cost
planning practice revealed that these techniques of measurement and
valuation were still the most commonly used. A fuller account of con-
temporary cost planning practice is given in PSA (1981 ).
During the late 1970s and early 1980s a number of practitioners began
to realise that these traditional methods of cost planning left much to be
desired. Smith (1980) summed up the problem when he wrote:
Computers in Cost Planning 137
One weakness of cost planning as traditionally practiced is that it starts

too late in the design process to have maximum influence. Often it is
little more than an estimating service, costing design decisions being
already taken.
In addition clients had become aware that the tender price of a project in
many cases no longer represented the most significant proportion of the
project cost. Huge rises in the cost of energy during the 1970s caused
building owners and developers to consider not only the capital cost of the
building but also the on-going revenue cost, and such factors of course
began to have a profound effect not only upon the design of individual
elements but more so upon the initial conception of building form. It was
plainly apparent that cost planning had to begin even before sketch plans
were prepared if the client was to achieve maximum value for money, and
furthermore some mechanism had to be developed to allow a number of
alternative design concepts to be evaluated and compared, not only in
terms of capital cost but in terms of total cost profile. A cost model was
required which would permit 'what-if' simulations in order to establish the
'best fit' solution for each individual client and each individual project.
Maver (1979) describes a number of such 'cost models' developed in the
Abacus unit at the University of Strathclyde, and Smith (1980) describes
the use of such techniques in local authority school building.
Despite early enthusiasm in some quarters, however, modelling tech-
niques such as these have not achieved general acceptance within the
profession as a whole. Brandon and Newton (1986) comment that:
For twenty years huge amounts of time and money have gone into the
development of models to improve building cost forecasting and manage-
ment. But an investigation of quantity surveying offices would reveal an
almost exclusive concentration on approximate quantities in various
stages of refinement, often used too late in the design process to be
effective in influencing the design.
Despite this lack of acceptance some commentators such as Beeston
(1983) and Newton (1986) contend that the quantity surveyor should make
more use of even more sophisticated mathematical and statistical tools in
the modelling process.
Computers and cost planning
Cost planning is not an exact science, and most quantity surveyors

specialising in cost planning would regard a subjective 'feeling' for the
project to be a valuable tool in their professional tool kit. No two projects
are exactly the same, and the cost planner must somehow take account not
only of the building and its site, the peculiarities of the architect etc., but
also must consider and make some allowance for factors such as market
conditions. Even more difficult to foresee are factors such as the work-load
oftendering contractors and their keenness or otherwise to win the contract;
factors which the cost planner has no way of evaluating with any degree of
accuracy. There is therefore it seems little chance of a completely automated
cost planning system, but computers can nonetheless have a valuable role
to play in cost planning, in the collection and analysis of historical cost
data, in the processing of cost estimates and in the dynamic evaluation of
alternative design solutions through the medium of cost models.
It should however also be pointed out that modern cost planning is
essentially an iterative and interactive process, and that whilst some
attempts were made to use mainframe computer systems it is only really
during the last ten years that computers have become widely used in this
field.
Cost analysis
We have already seen that many of the quantity surveyor's traditional cost
planning processes depend upon the collection and analysis of historical
cost data gathered from accepted tenders, and facilities to carry out cost
analysis of this kind have long been provided as an 'add-on' to Bill
production packages. The scope of such programs tends to vary somewhat
from system to system, but the level of analysis available depends of
course upon the level to which the original information has been coded
(see also the discussion in chapter 10 entitled 'Data labelling').
The extent of data labelling required to analyse a project into, for
example, the form required by the RICS Standard Form of Cost Analysis,
may be very time consuming and may prove not to be cost effective unless
the labelled data is to be used for some other purpose, for example to
resort a work section tender Bill into elemental form for contract manage-
ment purposes. In addition, although most cost analysis data is presented
in the form of elemental cost/square metre of floor area, what is often
required for future cost estimates is the cost/unit area of a given element,
for example the cost/m 2 of external wall where the cost is related to the
area of external wall rather than to the floor area of the building. This cost
is usually called the 'element unit rate' (EUR), and some programs have
provided facilities for this to be calculated. They do however in general
require the input of the quantity of each element, the 'element unit
quantity' (EUQ), and the calculation of element unit rates is often done
more quickly without using the computer at all.
Tender price indexing
One of the major problems encountered in using historical cost data for
the planning of new projects concerns the difficulty of adjusting tender
price data to take account of time. A study carried out in 1970 at the
University of london developed a method of indexing tenders using a
comparison of the 'price important' items in a given scheme with a
standard price schedule. The methodology developed is described in
Mitchell (1971), and the method is now in common use and forms a major
feature of the RICS Building Cost Information Service (BCIS).
The method, although not difficult to apply manually, requires a good
deal of calculation and a number of practitioners have developed small
computer programs of their own to handle the arithmetic involved.
Others have gone further and have used a file of tender price indices as
input to other computer programs employing statistical techniques to
predict future cost trends. Whilst services such as the BCIS do calculate
and publish predicted cost trends, the values calculated are often on a
national basis, and the national trend can sometimes be widely different
from that shown in a particular region. Some practitioners with sizeable
building programs in a small geographical area, for example local author-
ities, have found it useful to develop programs of this type for their own
use, and experience seems to show that fairly reliable forecasts can be
made up to twelve or fifteen months ahead using these techniques.
Cost estimating
The use of computers for cost estimating involves using the machine to
interface with some kind of database of cost information. Cost information
databases have been constructed using both historical cost analysis data
derived from previously tendered projects, and also using rates built up
from basic principles using current prices for labour, materials and plant.
Both methods have their supporters; the users of basic price data point out
that use of historical cost analyses is fraught with pitfalls since the method
and accuracy of pricing of previous schemes is unknown, whilst the
supporters of the historical approach contend that use of basic price data
takes no account of the competitiveness of tenderers or the state of the
market, and is in any event difficult unless some design decisions have
already been taken. Nonetheless both methods continue to be used.
As regards computer systems based on historical data, the majority work
by allowing the user to select from the database one or more cost analyses
from previous schemes which are as much like the present project as
possible in size, function etc. The analyses selected are then adjusted to
take account of time and location to bring them all to a common base, and
are then aggregated in some way to form a cost plan for the new scheme.
The BCIS 'On-line' service (see below) offers facilities of this type.
Those practitioners preferring to work from basic price data have a
number of systems available to them. Systems of this type generally
incorporate files of basic materials prices, labour constants, plant costs and
outputs in much the same way as those included in a builder's price book;
in fact some use systems originally developed to aid contractors in the
preparation of tenders. Such systems tend to be used mainly for cost
estimating once sketch designs have been produced, and the input options
available often include direct entry of dimensions with the computer then
performing an elementary abstracting and billing process. Some systems
include rudimentary libraries of descriptions similar to but less compre-
hensive than those included in packages for the production of Bills of
Quantities. Some systems make provision for digitised input. A more
complex system of this type is described in Holes (1987).
Many computer aided design systems include cost estimating routines,
most of which work by building up prices from basic rate information
stored with the graphical components, and the integration of computer
aided design and the provision of early cost advice is discussed further in
chapter 12.
The BCIS On-line service
The RICS Building Cost Information Service has provided a centralised

information exchange for quantity surveyors' cost data since the early
1960s, and in the early 1980s began to establish a computer based cost
data bank which could be accessed directly by subscribers through their
own office computer systems via the public telephone system. At the same
time an approximate estimating package was introduced as an integral part
of the service in order that subscribers could manipulate the data extracted
from the database on their own office computers, and thus produce cost
plans for new schemes. A more detailed description of the system is given
in Pegg (1987).
The system appears to function adequately from a technical point of
view, although some users have found the speed of data transfer rather
slow. The system offers considerable potential although it appears that at
present comparatively few quantity surveyors use the service on a regular
basis.
Cost modelling
The majority of innovative work associated with the use of computers in

this field has, in recent years, been concentrated on the development of
cost modelling systems in which the quantity surveyor is able not only to
represent the total cost profile of a project but also to interact with the
model; to indulge in 'what-if' simulations as the design progresses in order
to attempt to ensure the best possible fit between the clients' requirements
and the completed scheme. The most complete current account of work in
this field is probably Brandon (1987).
A number of approaches to cost modelling have been explored, and the
rapid development of powerful low cost computer systems capable of
handling large volumes of data has intensified the effort in recent years.
The majority of existing cost models have employed familiar historical
cost data gathered from cost analyses as described above, but have
attempted to use the power of the microcomputer to allow the user to
interact directly with the data. A nice example of this type of model is that
described by Brandon (1985). Whilst many system developers have written
purpose built computer programs for the task, Williams (1987) describes an
example of this type of model developed using a three-dimensional spread-
sheet as the development tool.
Other researchers have advocated other types of model. Brandon and
Newton (1986) for example describe the beginnings of an expert system
christened GUIDE (Graphical User Interrogation of Development Econ-
omics), and this approach has been carried even further through the joint
RICS/Aivey research programme project considering the use of expert
systems in the strategic planning of construction projects. The results of the
research are described in Brandon, Basden, Hamilton and Stockley (1988).
Expert system techniques as applied here however simply attempt to use
the computer to reproduce the reasoning process used by the human
expert in the solution of cost planning problems, and therefore whilst the
research conducted by Brandon et al is a breakthrough in its own right
such systems may prove of limited use if it can be shown that the human
model upon which they are based is itself either defective, inefficient or
flawed.
Some workers in the field have proposed that the intuitive methods
commonly used at present by cost planners are defective in that they make
little attempt to take account of construction management issues; they are
essentially concerned with predicting costs on the basis of an average of
that which has gone before and they make little allowance for the planning
and constructional difficulties of any one particular scheme. It is for this
reason that some researchers in the field advocate radically different types
of cost model. Beeston (1987) for example proposes the use of three
alternative techniques which attempt to model the construction process
itself either by the simulation of construction in detail, attaching costs to

activity networks or by simulation of construction planning. The develop-
ment of this type of system is described in Bennett and Ferry (1987) and by
Bowen, Wolvaardt and Taylor (1987). Techniques such as these have
obvious overlaps with computer software developed for project manage-
ment, and it may well be that closer integration of the management and
cost modelling functions will in due course become inevitable.
Conclusion
It can clearly be seen from the above that the drive to provide the client
with ever more accurate and reliable early cost advice is seen by some to
be one of the major areas of development in the field of 'traditional'
quantity surveying. The development of the necessary techniques is seen
by some to require the use of extremely sophisticated mathematical and
statistical techniques for which the use of computer systems is seen as
essential. On the other hand however the costing tools used most often by
the majority of quantity surveyors in their day-to-day work are still shown
to be those of approximate measurement and valuation, but even in this
case the advice is usually required very quickly if the design process is not
to be disrupted. The drive for greater and greater speed in the pre-contract
process, together with the use of contractual techniques which allow
overlap between the design and construction phases must mean that the
use of computers in this field will continue to increase.
Taking the broader view it can be seen that all of the different tech-
niques discussed above are likely to have some part to play in the costing
process. At the very beginning of a scheme, when a developer is attempting
perhaps to assess the viability of a new building as opposed to a rehabili-
tation, before any design decisions are taken, there is an obvious use for
interactive models, perhaps using more advanced mathematical and stat-
istical methods to manipulate historical data, implemented in the form of
expert systems, but as the design progresses there is an equally obvious
need for more traditional techniques based on the familiar principles of
measurement and valuation. Such techniques must however be imple-
mented in a form which allows design decisions to be evaluated
dynamically as the design progresses. For the majority of schemes such
techniques will probably be sufficient, but for larger and more complex
projects, in particular perhaps those designed using computer aided design
systems, models taking account of the construction process will probably
be required, and one can I think expect to see closer integration between
computer aided design systems, cost models and project management
software.
Whatever the eventual outcome it is obvious that the area of pre-
contract cost advice is one in which there is considerable scope for the
development of new techniques and increasing application of computer
systems if the industry is to be able to provide the clients of the future with
the building they require at the right price, at the right time and to the right
quality.
References
Beeston D. (1983) Statistical methods for building price data E. & F.N. Spon
Beeston D. (1987) 'A future for cost modelling' in Building, cost modelling and
computers Brandon P.S. (ed) E. & F.N. Spon
Bennett]. and Ferry D. (1987) 'Towards a simulated model of the total construction process' in
Building, cost modelling and computers Brandon P.S. (ed) E. & F.N. Spon
Bowen P.A., Wolvaardt j.S. and Taylor R.G. (1987) 'Cost modelling: a process modelling
approach' in Building, cost modelling and computers Brandon P.S. (ed).
Brandon P.S. (1985) 'Data on tap' Chartered Quantity Surveyor March 1985 pp. 300-302
Brandon P.S. (1987) Building, cost modelling and computers E. & F.N. Spon
Brandon P.S. and Newton S. (1986) 'Improving the forecast' Chartered Quantity Surveyor
May 1986 pp. 24-26
Brandon P.S., Basden A., Hamilton 1., and Stockley j. (1988) Expert systems: the strategic
Holes L.G. (1987) 'Holistic resource and cost modelling' in Building, cost modelling and
computers Brandon P.S. (ed).
Maver T. (1979) 'Cost performance modelling' Chartered Quantity Surveyor December 1979
pp.111-115 .
Mitchell R. (1971) 'A tender based building price index' Chartered Surveyor july 1971
Morrison N. and Stevens S. (1980) 'Cost planning in theory and in practice: A construction
cost database' Chartered Quantity Surveyor june 1900 pp. 313-315
Pegg I.D. (1987) 'Computerised approximate estimating from the BCIS On-line database' in
Building, cost modelling and computers Brandon P.S. (ed).
PSA (1981) Cost planning and computers Department of the Environment Property Services
Agency
Smith G. (1980) 'Cost planning the design process' Chartered Quantity Surveyor August 1980
pp. 13-16
Williams G.R. (1987) 'The development of a spreadsheet application for the cost modelling of
buildings during the early design stages' in Building, cost modelling and computers
Brandon P.S. (ed).
l 0 Computers and Bills of
Quantity
Objectives

• Discuss the factors to be considered when considering the purchase of
a bills of quantities system.
• Explain the advantages and disadvantages of standard libraries of
description.
• Discriminate between the various types of billing system available and
identify their strengths and weaknesses.
• Discuss the advantages and disadvantages of computerised measure-
ment systems.
Computers and Bills of Quantities
It has already been shown (chapter 1) that, collectively, quantity

surveyors have, for some 30 years, devoted more time and effort to the
development of computer systems for the production of Bills of
Quantities than to all of the other applications of the technology put
together. The struggle to develop computerised billing systems which
are both cost effective and easy to use has over the years assumed
something of the fervour of the quest for the Holy Grail; unfortunately
cynics, of whom the quantity surveying profession seems to possess a
large number, would no doubt say that to date both enterprises have
met with approximately equal success in that none of the billing
systems presently available has proved sufficiently attractive to per-
suade a majority of practitioners to subscribe to its use. Nor is it that
the rank and file of the profession are averse to change. The rapid
spread of cut and shuffle techniques during the early 1960s is ample
proof if proof were needed that practitioners are only too willing to
change their methods if it can be seen that there are clear advantages
in doing so. One is therefore forced to assume that since only a
minority of firms use computerised billing techniques there are, despite
continuous and no doubt highly expensive advertising campaigns,
presently no systems available for purchase which can clearly be
shown to be sufficiently cost effective for general use.
Computers and Bills of Quantity 145
Although it is true that there has been no widespread adoption of

computerised billing techniques across the profession as a whole, and
despite the fact that some commentators continue to predict the
imminent demise of the Bill of Quantities as a contract document,
there are nonetheless a substantial number of billing systems available
on the market, and a substantial number of adive users. It is not the
intention of this chapter to present any kind of survey or review of the
packages available, rather to explore the kinds of facilities commonly
offered, to consider the strengths and weaknesses of the various
approaches used and to discuss those factors which ought to be
considered when contemplating the purchase of a system.
It is also necessary to point out at this point that many of the so-
called computer billing packages available for purchase are aimed not
at the professional quantity surveyor preparing Bills of Quantities in
the accepted sense, but rather at the small builder producing 'builders
quantities' for projects let on a 'plan and specification' basis.
Another important distindion which must be recognised is that
between those computer systems primarily designed for building works
and those designed for use in civil engineering.
Whilst this chapter is primarily concerned with systems for building
work, systems for the production of 'builders quantities' and those for
use in civil engineering are also briefly described.
Computerised billing systems -a comparison with manual methods
The sequence of operations for the manual production of a Bill of

Quantities is shown in figure 10.1, and it has already been shown
(chapter 1) that the original spur to development of the first systems
came largely from the observation that the 'working-up' stage of the
process had begun to become very expensive; to 'absorb an inordinate
proportion of fees' (RICS 1961). The working-up process was seen to
be largely mechanical in nature, and initial development thus concen-
trated on the automation of this process. A number of computer
systems currently available follow this model, which might be defined
as a 'basic' bill production system. The sequence of operations involved
in using a system of this type is shown in figure 10.2, and comparison
with figure 10.1 will reveal that:
(i) The abstracting and billing stage, originally requiring considerable
skilled labour, has now been mechanised thus reducing the amount
of skilled labour required for the process as a whole.
(ii) In order that the computer can collect 'like' items and
"TAKING OFF"
SQUARE ANO
CHECK DIMS
r----- -----,
I I
I I
I I
I I
I
COMBINED IN ''cuT AND
''wORKING UP" I SHUFFLE" SYSTEMS
I
TYPE AND CHECK
OPERATIONS NEEDING ~
SKILLED as INPUT L__j
PRINT
OPERATIONS NOT NEEDING II
SKILLED as INPUT L__j
Figure 10.1 Manual bill preparation
subsequently sort them into the correct bill sequence an additional

semi-skilled 'coding' process has been introduced. Various coding
techniques have been employed from time to time and are con-
sidered in more detail below. A further 'data entry' process then
follows during which the coded information is entered into the
computer.
I MEASUREMENT
f -- ......._rr = ====== ==:;,

II II
+ II SQUARE AND CHECK II
.JI II
- - - L!:-:.-::.:.======:.-!.1
CODE SOME SYSTEMS ONLY
t
OAT A ENTRY
+
ABSTRACT AND
DRAFT BILL
+
EDIT AND AMEND
t
I
• CAMERA READY"
FINAL BILL
+
INPUT RATES
+
I RESORT BILL
l
I SKILLED Q S NEEDED 0
'
COST ANALYSIS
NO SKILLED Q s NEEDED D
C0~1PUTER PROCESSING I I
Figure 10.2 Bill preparation by computer (basic system)

(iii) Some systems require the dimensions to be squared before the

coded information is entered to the computer, whereas some allow
the entry of 'raw' (i.e. un-squared) dimensions with the computer
then doing the necessary calculations. Whilst it might, at first sight,
appear obvious that dimensions should be squared by the machine,
mathematical calculation is after all what the machine is supposed
to be good at, some system developers take the view that it is
quicker and less prone to error for dimensions to be separately
squared and checked by unskilled labour using a pocket calculator.
(iv) A 'feedback loop' has been introduced into the process at the
abstrading stage allowing correction, amendment or alteration of
the data followed by re-processing of the draft bill as many times
as is necessary to ensure absolute corredness before final printing.
(v) The total process has been extended to allow the contractor's
tender prices to be added to the information already stored in the
machine. Some systems also provide the facility to 're-sort' the
information into alternative forms, thus allowing, for example, a
bill in SMM work section order for tender purposes and another in
priced elemental form for post contrad management, valuations
and the produdion of the final account.
(vi) Some systems also provide facilities to allow the automatic produc-
tion of elemental cost analyses from the priced data.
In view of the above it can clearly be seen that computerised bill produc-
tion systems provide not only automated produdion of the bill, but also for
subsequent re-use of the data in both the post-contrad and the pre-contract
fields, and occasionally developers have maximised this facility through the
produdion of other software packages dealing with pre and post contract
fundions. Perhaps the best known of these totally integrated systems is the
Ripac system developed by Rider Hunt and partners reviewed by Hunt
(1985).
The whole process only works however at the expense of additional
'coding' and data entry operations, and there is therefore unlikely to be
any dired saving in time or cost in the actual preparation of the Bill. The
major cost saving will probably accrue from the computer producing a
'camera ready' final print ready for duplication, thus eliminating the costly
and time consuming typing stage required with the manual process. One
must also however add that the development of 'cut and shuffle' tech-
niques, which occurred in parallel with some of the original computer
developments, also enabled quantity surveyors to make considerable savings
in working-up costs and the cost advantages of a computer system such as
that shown in figure 10.2 over a cut and shuffle system must be regarded
as marginal at best.
Much effort has been devoted over the years to improving the cost
effediveness of computerised bill produdion systems, most commonly by:
(i) simplifying the coding process

(ii) development of electronic measurement tools mainly based on
digitisers
(iii) development of 'automated measurement' techniques whereby the
computer 'measures' detailed quantities for a particular assembly
from 'global' information provided by the taker-off. Perhaps the
most common application has been in housing where quantities
for standard house types are stored in computer libraries and may
be recalled and merged with specific project information by simply
quoting the house type reference. Such techniques have been
given a variety of names, but are perhaps best known as 'unit
quantities'. An example of the use of such a technique is given in
Alvey (1976) pp. 99-114.
Each of the above is considered in more detail below.
The coding process
Automation of the abstracting and billing process requires the computer to

perform two quite separate and distinct functions:
(i) 'like' items must be collected together and the quantities summed
(ii) the cumulative items generated in (i) above must be sorted into
an acceptable 'bill' order.
The human worker-up, either abstracting traditionally or using a cut and
shuffle system can manage both of these tasks quite easily; he is able to read
and understand the descriptions written by the taker-off and, within limits, is
readily able to compensate for the fact that different takers-off may describe
the same item in subtly different ways and may not always use precisely the
same terminology. He can, in short, use his powers of discrimination to
decide when two items should be added together and when they should be
kept separate. In addition, because the human worker-up is able to under-
stand the meaning of the written description, he is also able to apply a series
of rules enabling him to sort the items into bill order.
This is not the case, however, with the computer. We have already seen
that the computer has no innate intelligence; it must be programmed, must
be explicitly instructed precisely which items are to be added together and
in what order they are to sorted for the final Bill. The application of
traditional computing techniques means that items to be added together
must be identical in every respect for the process to be successful, and the
difficulties involved in using a computer to sort pieces of text of varying
length into an acceptable order using the same rules as those employed by
the human worker-up were seen as unsurmountable by early system
developers. It was therefore considered necessary to convert each item to a

code which the computer could then use both to collect like items together
and also to sort the cumulative items into a predefined order.
This necessity for a code to be attached to each measured item led
directly to the need for some kind of reference manual which the coder
could use, in which every description was listed together with its code.
This manual came to be known as a 'library of descriptions', and the vast
majority of billing systems currently available still use the coded library
approach. Indeed there have been those (e.g. Dent 1964) who have
predicted that, in time, takers-off would simply write down the code rather
than the description thus eliminating the separate coding process. Unfor-
tunately, however, in most systems the code was simply a string of
alphanumeric characters which, with very few exceptions were totally
devoid of human meaning without reference to the coding manual, and
the difficulties inherent in subsequently following taking-off expressed
entirely in codes were too great for the idea to achieve any kind of
widespread acceptance.
Standard libraries of description
It has already been shown that the use of computers for Bill preparation led
directly to the need for a library of standard descriptions, and this was in
reality only a small extension of an already common practice in many
offices. Most firms of any size had an 'office standard' for descriptions,
even if this only consisted of following descriptions used in previous Bills,
and the idea of a formal library of descriptions for computer use was not in
itself remarkable.
What was perhaps more remarkable was the view taken by the RICS and
others, that the use of computers would lead directly to the the use of one
standard library of descriptions for use across the entire industry. The RICS
working party investigating the use of computers for working up said, in
1961:
Standardised descriptions are ultimately inevitable. This is in some
measure already done in some individual offices and departments but in
due course it must come for the profession as a whole.
Standardisation of descriptions ... would we feel help the industry to
build more cheaply. It would do so by assisting work study and thus
keen estimating.
(RICS 1961)
The perceived advantages of industry-wide standard descriptions have
historically been listed as:
(i) Standard descriptions would improve the quality of written com-

munication because the same thing would always be described in
the same way.
(ii) This standardisation would make life easier for estimators in
pricing and for quantity surveyors attempting to compare historical
prices for the same items of work.
(iii) The drafting of descriptions could be simplified, much of the work
could be done by junior staff and editing procedures could be
largely eliminated and replaced by routine checks.
Many quantity surveyors saw in the above an implied criticism of their
much prized skill in accurately describing the work to be carried out. It is
not therefore particularly surprising that, despite the backing given to the
idea by a number of eminent surveyors and the publication of several
standard libraries for manual use, of which the best known is probably that
developed by Fletcher and Moore, the universal use of standard descriptions
has remained a pipe dream. Many practices, it would appear, whilst using
an 'office' standard library for their own convenience still prefer their Bills
of Quantities to have a recognisable 'house style'. Others disagree with the
way in which discursive English in descriptions has been replaced by
staccato descriptions composed of phrases selected from a library and
separated by semi-colons. It is argued by some, convincingly in my view,
that to blindly take options from a library can actually result in a less
accurate description of the work than that which would result from careful
and accurate use of discursive English.
Thus it is that in 1988, some 27 years after the RICS working party
declared that standard descriptions were inevitable, the use of the Library
of Descriptions published for use with SMM7 (HMSO 1988) is still optional.
Contrast this with the civil engineering profession where the CESMM2
method of measurement embodies a mandatory library of standard descrip-
tions.
Types of library and coding structure
The simplest type of standard library would be one in which each complete
description were allocated a code. This type of library is therefore analogous
to the use of a Bill of Quantities from a previous job where the code for a
given item might simply be the page and item reference. If such a bill were
to be stored in a computer one could therefore 'code' dimensions for a
new project by using the page and item reference of the previous Bill, and
the computer would therefore be able to use the codes so allocated both to
collect like items together and to sort the items in the new project into an
acceptable order.
The basic problem with a library of this type is that to offer compre-
hensive coverage of all items, sizes and specifications likely to be
encountered it would need to be very large indeed. In order to be workable
therefore the library must be reduced to a skeleton form, with 'blanks'
inserted for all size also some specification information. This 'missing'
information is then included with items for the new projed as a part of the
code, and the computer substitutes the actual information provided for the
'blanks' in the library. This type of library has been used in a number of
systems; Cyril Sweet & Partners claimed, in 1978, to have developed a
comprehensive library of this type running to only 200 pages. (Sweet
1978).
More common is the 'articulated phrase' library in which descriptions
are de-composed into smaller constituent phrases each of which is then
given a code. Perhaps the most common method of de-composition is to
break descriptions down into a number of functional parts. Such an
approach requires descriptions to be phrased in a structured form, for
example the unstrudured description:
One brick wall in Class B engineering bricks (BS 392l) in cement mortar
(1 :3)
could be expressed in a rather more strudured form, and using SMM7
terminology as:
Masonry ....•....................... Primary SMM7 division
Brick/block wallinq ................ SMM7 sub-division
Enqineerinq bricks BS 3921

Class Bin cement mortar (1:3) ..... Specification
Walls .............................. Description of work
One brick tbick .................... Size information

Once descriptions are organised in this strudured form each component
can be considered separately, and lists may be construded for each in
such a way that 'whole' descriptions may be constr~ded by taking one
seledion from each list. The structural components of the whole description
are often known as 'levels' or 'facets', and part of a simple SMM7 library
construded on this model might look like this:
Primary SMM7 division
Code Description
C Demolition/Alteration/Renovation
D Groundwork
E Insitu concrete/Large precast concrete
F Masonry
G Structural/Carcassing metal/timber
SMM7 sub-division
Code Description
FlO Brick/block walling
Fll Glass block walling
F20 Natural stone rubble walling
Specification
Code Description
FlO Common bricks BS 3921 in cement mortar (1:3)

F20 Engineering bricks BS 3921 Class B in cement mortar (1:3)
F30 Engineering bricks BS 3921 Class A in cement mortar (1:3)
Description of work
Code Description
FlO Walls
F20 Isolated piers
F30 Isolated casings
F40 Chimney stacks
Size information
Code Description
FlO Half brick thick

F20 One brick thick
F30 One and a half bricks thick
The code for the item in question can therefore be formed by taking the
code for each facet and combining them together thus:
F Masonry
FlO Brick/Block walling
F20 Engineering bricks BS 3921 Class Bin cement mortar (1:3)
FlO Walls
F20 One brick thick
The code for the whole item is therefore F F1 0 F20 F1 0 F20.

It is plain that libraries constructed on this 'mix and match' approach
will be much more compact and much more flexible than those constructed
of whole descriptions, although the actual code constructed tends to be
rather longer. In practice of course a library such as that indicated above
would also require the addition of one further facet to deal with 'written
short' items. This type of alphanumeric coding structure is widely used in
many commercial systems, and the standard library published for use with
SMM7 (HMSO 1988) follows this 'articulated phrase' approach, although
with a rather different coding structure.
The coding structure used in the SMM7 Library of Standard Descriptions
is an interesting variation on the technique discussed above. At the heart of
the system is the concept of using the traditional quantity surveyor's
shorthand description as the code instead of the usual meaningless alpha-
numeric string. This concept was first explored by the Ministry of Public
Buildings and Works in the mid 1960s for use with the Enviro computer
system, and has subsequently been developed and refined within the
Department of the Environment into the notation now used in the present
SMM71ibrary.
The following is an example of the technique, and uses codes taken

from the SMM7 library of Standard Descriptions (the I characters simply
indicate the end of each facet):
Coded item - ASP RF/SPEC A/RFG/W+300 FLAT
translates as follows:
ASP RF Mastic asphalt roofing/insulation/finishes
SPEC A Specification reference A
RFG Roofing
W+300 FLAT over 300 wide

It will be interesting to observe how widely the SMM7 library codes are
used in practice. As we have already seen, the initial objective of this
approach in the original Enviro system was to eliminate coding as a
separate operation, but informal discussion with users both within the
Department of the Environment and in private practice appear to show that
this objective was not fully achieved, firstly because the standard codes
became quite convoluted and, secondly, because every taker-off tends to
develop his own set of 'pet' abbreviations with which he is comfortable
and which he is reluctant to abandon.
All coded library systems need to make some provision for the coding of
items not included in the library. Such items, commonly called 'rogues',
are in practice encountered on almost every job, particularly those con-
taining items of demolition or alteration. 'Rogue' items can be extremely
time consuming and inconvenient to incorporate since, in addition to
actually coding the item, the coder may also need to instruct the computer
what the code for the rogue item actually means. This process can often
involve both selecting a suitable unused or 'spare' code and also telling the
computer whereabouts in the Bill the item is to be placed. The ease or
otherwise with which rogue items can be accommodated can be a major
factor in determining the cost effectiveness of any particular system.
Present-day solutions to the coding problem
The most commonly used approach to ease the coding problem centres
around the use of VDU screens both for coding and for the input of the
data. A number of systems use this approach.
In essence the method works by using the computer to display on the
screen all of the descriptions and their associated codes stored in the
library which are appropriate to a particular level of the item description,

and the operator then chooses from the codes displayed. The system is in
effect an automated library and is often an inverted tree structure through
which the user is led step by step.
Early systems using this technique required the user to actually type in
the chosen code on the keyboard, but more sophisticated and user friendly
systems will allow the user to make a selection simply by 'pointing' to the
appropriate item on the screen, either by using the cursor, a mouse, a light
pen or some other electronic device. Such systems obviously hold out the
possibility of direct entry of the dimensions by the taker-off via his own
personal VDU, thus meeting in some degree the early hopes of eliminating
the separate coding process. A printed copy of the dimensions and the
descriptions chosen is usually generated simultaneously on a small printer
attached to the VDU.
Such an approach can be very cost effective in that it eliminates the
separate coding process, although it is apparent that, depending upon the
response speed of the computer the taking off time may well be extended.
The approach does however require that each taker-off has access to a
VDU and this can itself present a problem since the taking off work-load in
a typical office is not usually constant. Neither is it, in general, possible to
even out the taking-off load since generally speaking the preparation of the
Bills of Quantities is the last operation in the pre-tender phase and the
taker-off is often working to a tight time-scale. The combination of these
two factors means that, if the approach is to be successful then each taker-
off must have his own terminal. This can, of course, not only make the
initial cost of the system very high, particularly in an office with a large
number of takers-off, but can also result in considerable under-use of
terminals during slack taking-off periods.
An interesting alternative approach is the use of bar codes. This tech-
nique, used by at least one currently available system, uses specially
printed libraries in which the code for each item is printed in the form of a
bar code similar to that found on many tinned and packaged goods. The
coder or taker-off therefore simply finds the description he requires in the
library, and instead of writing down the description or typing it into a
VDU, simply scans the bar code with a special pointer connected directly
to the computer. One of the main advantages of this system over traditional
coding techniques, as well as that of eliminating the traditional data entry
process, is that the coding error rate is dramatically reduced.
Another alternative approach to solving the coding problem is the
Phraseology Independent Billing System (PHIBS) developed jointly by the
Architect's Department and the Computer Division of the Treasurer's
Department of Derbyshire County Council with assistance from a number
of others. This system has no library as such, the taker-off being completely
free to use any terminology he wishes provided that the description is
structured in levels in the way described earlier. The text is then entered into
the computer by a clerk/typist exactly as written by the taker-off. Once a
batch of data has been input to the machine the computer then scans the
text at each level for 'keywords', from which it constructs an internal code.
This 'hidden' code is then used both to collect like items together and to
subsequently sort the data into Bill order.
The system also includes a number of other facilities to ease the taking-
off process such as the use of 'ditto' for each facet of the description, easy
duplication and re-use of previous descriptions with the option to make
changes, the use of job-specific or global tables of abbreviated descriptions
using 'shorthand' codes such as those described above and the facility to
copy either single items or ranges of items from either previous dimensions
or from previous Bills. The system therefore eliminates coding as a manual
process and makes no additional demands upon the taker-off. There is also
no need for each taker-off to have constant access to a VDU.
Electronic measurement tools
In addition to the 'basic' Bills of Quantities systems described above, a

number of systems exist which also use the computer to assist with the
measurement process, either by generating quantities for groups of items
from one dimension (often termed automated measurement and discussed
in more detail below), or by utilising some form of electronic measurement
de\·ice.
Most common of these is the digitiser (see chapter 2), consisting of a flat
board generally incorporating a matrix of wires and a pointing device of
some kind. The board is able to detect the position of the pointer in both
horizontal and vertical directions and to send this information to the
computer as a series of X and Y co-ordinates usually relative to one of the
corners of the board. Other technologies exist in addition to the wire
matrix, for example the Techsonix system uses a digitiser equipped with
ribbon microphones along the X and Y axes together with an ultra-sonic
pointer. Digitiser tablets are available in a range of sizes from A3 to AO.
The digitiser itself is in concept a fairly simple device and it is easy
to see how, given the X and Y co-ordinates of both ends of a straight line
and some suitable software to handle the co-ordinate geometry involved,
one could use the computer to calculate the length of the line. If the line
were part of a scale drawing and the computer is told the correct scale
then it is also easy to see how the computer could 'take-off' linear
dimensions, and it is this simple concept which is at the heart of all
systems using a digitised approach to measurement.
One of the major problems faced by the developers of systems of this
type centred around the dimensional instability of the drawings. Stretching
and shrinking of the paper often caused inaccurate measurement when

using the scale provided by the operator. It is therefore now common for
the computer itself to establish the true scale of the drawing in each of the
X andY directions. This is generally done by establishing a reference line,
usually the longest dimensioned line, on each of the axes and telling the
computer the intended length of each in 'real-world' dimensions via the
keyboard. The lines are then measured by the computer and the software
can thus establish the true scale of the drawing in each axis.
We have therefore seen how computers can be used to measure the
length of straight lines, and it is obvious that given software only a little
more complex the same technology can be used to measure the plan area
of straight-sided shapes. In addition most digitisers will support a 'stream'
mode of operation in which a continuous stream of X and Y co-ordinates
is transmitted to the computer as the pointer is moved across the board.
This facility thus allows the pointer to be used to trace curves, free and
irregular shapes, and given again more complex software makes possible
the calculation of both the length of curves and also the areas of free and
irregular shapes. The application of such techniques to the measurement
of, for example, floor and ceiling finishings and external works is obvious
and the addition of further dimension information, often entered either
through the keyboard or via a menu mounted on the digitiser board allows
the automatic measurement from plans of wall areas, concrete volumes
etc.
Some developers have constructed highly complex software for specialist
measurement tasks such as the calculation of bulk earthworks volumes
from digitised contour maps of existing and proposed ground surfaces.
Such calculations are extremely tedious when done by hand and there is
no doubt that the use of digitisers for this type of work saves enormous
amounts of time and is very cost effective.
Most billing systems using this approach will use the digitiser board as
an additional input device in conjunction with a VDU.
It should not be assumed however that digitisers offer a complete
solution to the problem of measurement. Whilst, as shown above, the use
of such techniques is very effective for certain types of work it is always as
well to remember that, despite the machine's ability to compensate for the
effects of the dimensional instability of paper, the accuracy of the measure-
ment is only as good as:
(i) the skill ofthe original draughstman and
(ii) the skill of the digitiser operator particularly when following curves
or the boundaries of irregular shapes
In addition it is obvious that the potential error margin will increase the
smaller the scale of the drawing. It is plainly more difficult both to draw
and to accurately trace shapes from drawings at a scale of 1:2500 than it is
to follow similar lines on a drawing drawn at a scale of, say, 1:5, and it is
plainly preferable in many cases to measure using figured dimensions
when they are given. Remember that it is not for nothing that many
architects and engineers add a note to their drawings warning of the
dangers of using scaled dimensions!
Automated measurement
The use of the computer to 'automatically' measure a whole range of items
from one code has its roots in the very beginnings of computer billing
systems, and facilities of this type are described in several of the papers
presented at both the 1967 and 1973 RICS conferences on computer
billing systems. In essence the technique most commonly adopted uses the
computer to store pre-measured quantities for parcels of work each of
which is then given a unique code. The whole parcel may subsequently be
included into a new scheme and multiplied as necessary simply by quoting
the appropriate code and the 'number off', and thus considerable time can
be saved where particular constructional details or component assemblies
repeat from project to project. The most obvious examples occurred in the
1970s in the twin areas of local authority housing schemes and system
building where component assemblies or even whole standard house types
could be stored and recalled in this way. The technique has been given a
variety of names by various system developers, (e.g. 'Biackbox', group
quantities, unit quantities etc.); we shall use the term 'unit quantities'.
Tl-te technique is particularly important since it can also be used to form
a useful link between computer billing systems and architects' computer
aided design systems, and this is discussed further in chapter 12.
Unit quantities
One simple example of the use of unit quantities might be in the case of
standard timber windows, which are commonly coded by the manufacturer
and so identified in trade catalogues. A hypothetical example of such a
window range is shown in figure 10.3, and is used here to indicate the way
in which a unit quantity library could be constructed such that all of the
measured items required for each type of window could be stored in the
computer and associated with the relevant code.
Each window code would therefore need to generate the following items
for inclusion in the Bill:
(i) supply and fixing of the window frame
(ii) glazing
(iii) painting internally and externally
(iv) bedding and pointing the frame.
,f 600mm .~ --,,1'--r_ ___::_;9o:..:_om=m:.:___+,~ .r 1200mm
Height
09 - 900mm A09 B09 C09
12 - 1200mm A12 B12 C12
15 - 1500mm A15 B15 C15
Notes
(i) Height of top light is always 350mm, bottom light varies
(ii) Windows may be glazed with either 4mm clear sheet glass (code
C) or 4mm hammered obscure glass (code 0)
Figure 10.3 Hypothetical range of timber windows
It is important to realise that the facilities available for this type of work
differ greatly from one system to another. The example shown here is
therefore not taken from any particular system but instead indicates the
kind of facilities which have been offered in the past. The unit quantity
library which might be prepared for the example given above is shown in
figure 10.4.
Code I !Measured item
----------------------------------------------------------------
A09-1 1l 1 ITimb wdw/Sw/wdw & fr/ref A09 cramp fxgl 2l
I
A12- 1 ITimb wdw/Sw/wdw & fr/ref A12 cramp fxg
I
A15- 1 ITimb wdw/Sw/wdw & fr/ref A15 cramp fxg
I
B09- 1 ITimb wdw/Sw/wdw & fr/ref B09 cramp fxg
I
I
I
C09- 1 ITimb wdw/Sw/wdw & fr/ref C09 cramp fxg
I
C12- l ITimb wdw/Sw/wdw & fr/ref C12 cramp fxg

I
Cl5- l Timb wdw/Sw/wdw & fr/ref C15 cramp fxg
I
- - - - (1) 2/1 * I 13! Timb wdw/mast/bed & ptg fr
2/1 ** I
I I
A--c14l I 0.501 Gen g1zg/stdd pln glas; 4mm clear sheet/
I .Q...2.5.1 wd puty/in s pane (nr.1)
I I
A--0 I 0.501 Gen glzg/stdd pln glas; 4mm hammered
I .o...221 obsc./wd puty/in s pane (nr.1)
I
B--e 0.801 Gen glzg/stdd pln glas; 4mm clear sheet/
.Q...2.5.1 wd puty/in pane
I
B--0 0.801 Gen glzg/stdd pln glas; 4mm hammered
.Q...2.5.1 obsc./wd puty/in pane
I
c--c 1.10 I Gen glzg/stdd pln glas; 4mm clear sheet/
.Q...2.5.1 wd puty/in pane
I
c--o 1.10 I Gen glzg/stdd pln glas; 4mm hammered
.Q...2.5.1 obsc./wd puty/in pane
I
A--15! 0.501 Gen glzg/stdd pln glas; 4mm clear sheet/
~1(6) wd puty/in pane
I
A--c 0.501 Ddt. Ditto
.ll......!Q.Im
I
A--0 I 0.501 Gen glzg/stdd pln glas; 4mm hammered
I ~I obsc./wd puty/in pane
I I
A--0 I 0.501 IDdt. Ditto
I .ll......!Q.I I
I I I
B--e I 0.801 IGen glzg/stdd pln glas; 4mm clear sheet/
I ~I lwd puty/in pane
I I I
B--e I 0.801 IDdt. Ditto
I .ll......!Q.I I
I I
B--0 I 0.801 IGen glzg/stdd pln glas; 4mm hammered
I ~I lobsc./wd puty/in pane
I I I
B--0 I 0.801 IDdt. Ditto
I .o......!Q.I I
I I I
c--c I 1.10 I IGen glzg/stdd pln glas; 4mm clear sheet/
I ~I lwd puty/in pane
I I I
c--c I 1.10 I IDdt. Ditto
I .o......!Q.I I
I I I
c--o I 1.10 I IGen glzg/stdd pln glas; 4mm hammered
I ~I lobsc./wd puty/in pane
I I I
c--o I 1.10 I IDdt. Ditto
I Q...1QI I
I I I
I *I IPaint/pt wd/glzd wdw etc/g+300 m panes
I ~1(8) I
I I I
I *I Paint e/pt wd/glzd wdw etc/g+300 m panes
I ~I
I I
C12- I !9l I *I Ddt.Paint/pt wd/glzd wdw etc/g+300 m
I I ~I panes
I I I
C12- I I *I Ddt.Paint e/pt wd/glzd wdw etc/g+300 m
I I ~I panes
I I I
C12- I I *I Paint/pt wd/glzd wdw etc/g+300 1 panes
I I ~I
I I I
C12- I I *I Paint e/pt wd/glzd wdw etc/g+300 1 panes
I I ~I
I I I
ClS- I I *I Ddt.Paint/pt wd/glzd wdw etc/g+300 m
I I ~I panes
I I I
ClS- *I IDdt.Paint e/pt wd/glzd wdw etc/g+300 m

~I !panes
I I
C15- *I IPaint/pt wd/glzd wdw etc/g+300 l panes
~I I
I I
C15- *I !Paint e/pt wd/glzd wdw etc/g+300 l panes
.:I I
Figure 10.4 Example unit quantity library
(1) The size of the unit quantity library can be kept to a minimum by
calling forward items using only parts of the total code. In this case
the total number of windows in the range is nine and each
window extracts five items. In addition each window can be
glazed with either clear or obscure hammered glass. The code
therefore needs to consist of the window reference and the type of
glass, thus giving a total of 9 x 5 x 2 = 90 separate items to be
measured. Certain items however are only dependent upon a part
of the code, for example the item for bedding and pointing is
always required, and an item for the window frame is always
required regardless of the type of glass. Using this technique only
38 items actually need to be measured to cover the entire range.
(2) The abbreviations used are based on those included in SMM7
Library of Standard Descriptions.
(3) Another technique which is sometimes used to minimise the size
of the unit quantity library is to generate some dimensions auto-
matically from the code via a look-up table. In this example one
asterisk means that the dimension is generated from the first part of
the code, and two asterisks means that the dimension is generated
from the second part of the code. The code used here consists of
three parts:
(i) a letter denoting the width of the frame
(ii) a two character number denoting the frame height
(iii) a single character denoting the type of glass
The different parts of the code are often referred to as 'parameters',
and in this case look-up tables could be stored in the computer thus:
1st parameter (width) 2nd parameter (height)

A= 0.60 09 = 0.90
B = 0.90 12 = 1. 20
c = 1. 20 15 = 1.50
Thus when the computer finds a letter A as the first part of the code
the* entry in the dimension column is replaced by 0.60, and so on.
(4) This sequence of items measures the glazing to the top lights,
which are always the same size regardless of the window height,
hence only the width and glass type are significant.
(5) These items measure the glazing to the bottom lights.
(6, 7) Here the dimension extracted from the second parameter of the
code is multiplied by a given width to calculate the area of the
bottom pane. Although not good conventional taking-off practice it
again helps reduce the size of the unit quantity library to adopt the
slightly unusual approach of measuring the area of the whole
window and subsequently deducting the area occupied by the top
light.
(8) Here the dimensions for the painting items are both extracted from
look up tables before being multiplied together in the usual way.
(9) Although painting is required on all windows and is measured above,
these two windows only are divided into large panes and thus call
for special treatment. It is therefore necessary to deduct the painting
already measured and add back the correct painting items
If we assume that a given project requires 10 Nr. type B12 windows glazed
with clear glass and 2 Nr. type A09 windows glazed with obscure glass
then the relevant items would be extracted from the unit quantity library,
multiplied by the 'number off', and the dimensions shown in figure 10.5
would be merged with the traditionally measured work.
I
2/1 1 ITimb wdw/Sw/wdw & fr/ref A09 cramp fxg
I I
10/1 1 ITimb wdw/Sw/wdw & fr/ref B12 cramp fxg
I I I
2/1 0.601 ITimb wdw/mast/bed & ptg fr
2/1 0.901 I
10/1 0.901 I
10/1 1.201 I
I I I
2/1 0.501 IGen glzg/stdd pln glas; 4mm hammered obsc./
I .Q......QI lwd puty/in s panes (nr.2)
I I I
2/1 0.501 IGen glzg/stdd p1n g1as; 4mm hammered obsc./
I 0. 90 I lwd puty/in pane
I I I
2/1 0.501 IDdt. Ditto

I 0.401 I
I I I
10/1 0.801 IGen g1zg/stdd pln glas; 4mm clear sheet/
I .o.....l.5.1 lwd puty/in pane
10/1 0.801 I
I l....l.Q.I I
I I
10/1 0.801 IDdt. Ditto
I Q....iQ.I I
I I I
2/1 0.601 IPaint/pt wd/glzd wdw etc/g+300 m panes
I _()_,j_Q I I
10/1 0.901
I l....l.Q.I
I I
2/1 0.601 !Paint e/pt wd/glzd wdw etc/g+300 m panes
I _()_,j_Q I I
10/1 0. 90 I I
I l....l.Q.I I
I I I
Figure 10.5 Dimensions selected from the unit quantity library
It can clearly be seen from the figure that the use of computerised
techniques of this kind can be extremely cost effective where repetition
occurs. It can however also be clearly seen from the above example that
the preparation of comprehensive unit quantity libraries can be a very time
consuming business, and it is therefore most important that one only
considers the use of such methods where the extent of repetition is
sufficient to justify the initial investment involved in establishing the library.
One must also remember that, even with the best will in the world,
architects and other designers seen incapable of leaving any standard
detail unmodified for very long, and the ease with which unit quantity
libraries of this type can be amended to take account of such changes will
also be a crucial factor in their long-term cost effectiveness.
Other automated measurement techniques

An alternative approach to automated measurement is that used by the
'Calculix' system, reviewed in Hunt (1984). In this system the quantity
surveyor, rather than measuring the building in the traditional way,

'describes' the building to the computer by working round each room wall
by wall. The actual items and quantities are then extracted by the computer
from the model thus created. The technique therefore requires a different
approach to measurement but is claimed to be much more efficient than
traditional methods once the taker-off has become accustomed to using the
system. The approach again holds out the promise, in theory at least, of
direct measurement from similar models created by architects using com-
puter aided design systems.
It is also possible to use conventional spreadsheet software to assist in
the measurement process quite independently of any bill production soft-
270mm
DEPTH
d
. .
. ..• ;• ·~
.
~
..
TH ICKNESS
" ...· . ·, t
~ -· . ~ .. .. ~.. =:·· .. ·. - : ~.... ·:
WIDTH w
Figure 10.6 Simple strip foundation

ware and this can prove very cost effective in some circumstances. The
approach is illustrated using the simple strip foundation shown in figure
10.6.
If the formulae required to calculate the quantities of work required for 1
metre run of the foundation were to be stored in a spreadsheet, then given
the total trench depth, foundation concrete thickness, trench width and
total length the computer will produce a schedule of all of the quantities
involved.
Figure 10.7 shows a spreadsheet set up in this way, with the relevant
formulae set up in cells F19 - F42. Figure 10.8 shows the input dialogue
with the user after the spreadsheet borders have been removed and the
sheet 'windowed' to display only the first thirteen lines. The relevant
information is then entered into cells F1 0- F13, and the resultant schedule
of quantities is shown in figure 10.7.
I A II B II C II D II E II F II G
11
21 SPREADSHEET MEASUREMENT EXAMPLE
31 ===============================
41
51 Simple strip foundation
61 -----------------------
71
81Enter the following information:-
91
101 foundation depth (mm max 2000) 0
111 trench width (mm) 0
121 concrete thickness (mm) 0
131 total length (m) .00
141
lSISchedule of quantities
161----------------------
171
181Excavate trench maximum depth not exceeding
191 .00 m maximum depth .00 m3
201
211Level and compact bottom of excavation .00 m2
221
231Earthwork support not exceeding 2.00 m
241between opposing faces not exceeding
251 .00 m maximum depth .00 m2
261
271Concrete in foundation in trench .00 m3
281
291Half brick thick wall .00 m2

301
311Form cavity of hollow wall .00 m2
321
331Concrete filling to cavity of hollow wall .00 m3
341
351Backfill with excavated material .00 m3
361
371Backfill with hardcore .00 m3
381
391Remove surplus excavated material .00 m3
401
411Polymer damp proof course 110 mm wide
421bedded in cement mortar (1:3) .00 m
Figure 10.7 Simple spreadsheet measurement
SPREADSHEET MEASUREMENT EXAMPLE
Simple strip foundation
Enter the following information:-
foundation depth (mm max 2000) 900

trench width (mm) 750
concrete thickness (mm) 225
total length (m) 27.50
Figure 10.8 Spreadsheet measurement - input dialogue
Schedule of quantities
Excavate trench maximum depth not exceeding

1.00 m maximum depth 18.56 m3
Level and compact bottom of excavation 20.63 m2
Earthwork support not exceeding 2.00 m

between opposing faces not exceeding
1.00 m maximum depth 49.50 m2
Concrete in foundation in trench 4.64 m3
Half brick thick wall 45.38 m2
Form cavity of hollow wall 22.69 m2
Concrete filling to cavity of hollow wall 1.07 m3
Backfill with excavated material 4.46 m3
Backfill with hardcore 4.46 m3
Remove surplus excavated material 14.11 m3
Polymer damp proof course 110 mm wide

bedded in cement mortar (1:3) 55.00 m
Figure 10.9 Schedule of quantities
Data labelling
It is common for computerised Bill production systems to provide facilities

for the basic information to be resorted into a number of different forms,
for example it is common for tenders to be invited using a Bill arranged in
SMM work section order whereas it may be more convenient for contract
administration purposes if the Bill is subdivided by elements, work groups
etc. It is also necessary to allocate items to particular sub-divisions if it is
intended that an elemental cost analysis of any kind is to be produced from
the priced Bill.
In order that these additional facilities may be exploited it is necessary
for extra information to be input to the machine in addition to the description
and dimension information discussed above. This additional information,
sometimes referred to as 'data labels', is conventionally input in a coded
forrn and the majority of systems include a separate 'library' of available
types of sub-division together with the appropriate codes. A number of
coding systems have been used in the past including CI/SfB, the Building
Industry Code, the RICS Building Cost Information Service list of elements
and sub-elements and many others.
The detailed facilities offered by individual systems do differ from one
system to another, and some are more flexible than others, but typical
systems have in the past allowed sub-division of the data by, for example:
house type
block- i.e. individual building within a total scheme
sub-block -a functional sub-divison of the above element
sub-element
location
SMM work section
Any or all labels are usually allowed in any one run, often in any
combination; thus it is possible, assuming that the appropriate information
has been coded, to produce for example a Bill sorted by block, sub-block
and SMM work section for tender purposes and to subsequently resort the
information into block, sub-block, element, and SMM work section form
for use on site.
Civil engineering measurement
The measurement of civil engineering works poses its own quite distinct
problems. Civil engineering works tend to be painted on a rather wider
canvas than do building projects; the construction of a road, for example,
tends to contain less of the fiddly detail than is usual on building projects
of comparable or even much lesser value, and this is of course reflected in
the rules for measurement of civil engineering work. In addition much civil
engineering measurement is carried out by civil engineers rather than
specialist quantity surveyors, and virtually all civil engineering work is
remeasured as executed. The combination of these factors has tended to
produce much simpler rules for the measurement of civil engineering
works, culminating in the 2nd. edition of the Method of Measurement for
Civil Engineering Works (CESMM2) which embodies not only the rules for
measurement but also a full and mandatory standard library of descriptions.
In addition of course the measurement of this type of work commonly
involves much more involved calculation than does the measurement of
building works. Such a situation is, of course, tailor made for computer-
isation.
Although most (if not all) of the billing systems designed for use on
building projects could be fairly readily adapted for use on civil engineering
works, they have tended in the main to be too complex for the type of
measurement required, and neither have they provided suitable facilities to
help ease the particular mathematical problems posed by civil engineering
measurement. Consequently a number of computer systems have been
developed specifically to cater for civil engineering works and these systems
in turn have little or no application to building projects.
'Builders quantities' systems
As stated earlier, a number of the systems sometimes referred to as billing

systems are in fact aimed predominantly at the smaller builder, whose
needs are for a system enabling him to easily produce his own quantities
for plan and specification work. Since the eventual aim of such systems is
the production of an estimate for a project rather than the production of a
Bill of Quantities they therefore tend to include only rudimentary billing
facilities but quite complex estimating and pricing routines. They tend
therefore, in the main, not to be suitable for use by the quantity surveyor in
the production of bills of quantities in accordance with the Standard
Method of Measurement, but they can however be of use in other areas of
work, in particular for approximate estimating.
References
Alvey R.j. (1976) Computers in quantity surveying Macmillan

Dent C. (1964) Quantity surveying by computer Oxford University Press
Hunt G. (1984) 'Now you can measure your modules' Chartered Quantity Surveyor january
1984 p. 219
Hunt G. (1985) 'The all-singing all-dancing system?' Chartered Quantity Surveyor january
1985 p. 213
HMSO (1988) SMM7 Library of Standard Descriptions HMSO
RICS (1961) 'The use of computers for working up' Chartered Surveyor April1961 pp. 561-
563
Sweet (1978) Concepts in quantity surveying Cyri I Sweet and Partners pub I icity material
11 Computers for Post
Contract Work
Objectives

• Comment on the general use of computers in post contract work.
• Discuss the factors affecting the re-use of bill production data for post
contract work and the establishment of a comprehensive integrated
project data base.
• Discuss the advantages and disadvantages of the use of portable
computers on site
Computers in post contract work
Previous chapters have examined the various ways in which computers

may be used by the quantity surveyor in his professional role both during
the inception and feasibility stage of a project and also to aid in the
production of contract documentation, particularly the preparation of Bills
of Quantity. This chapter examines the use of the computer during the post
contract phase of a project.
In some ways this is by far the most difficult area of the quantity
surveyor's work to explore from the point of view of computer usage,
since although the use of computers during this stage is quite common the
precise use to which they may be put will depend to a large extent upon
the post contract arrangements for the particular project and the extent to
which the quantity surveyor is involved in areas such as project manage-
ment. It is therefore obvious that whilst the quantity surveyor may have
some specific and particular needs he is also in many instances likely to
make use of systems more often considered the responsibility of the
contractor or the engineer. Indeed he may well, in some cases, not only
use the software but also data created by some other member of the
construction team. The position is also rather special for the quantity
surveyor employed by a contractor. In this case the quantity surveyor is
likely to be less involved in the formal administration of the Contract, but
much more concerned with factors such as the administration of
Computers for Post Contract Work 173
nominated and domestic sub-contractors and the profitability ofthe project

from his employer's point of view. He may well therefore use computers,
indeed may be deluged with computer printout, but the systems with which
he is concerned are likely to be those to do with the management of the
building company itself. A study of systems of this type more rightly falls
within the province of building management, and is not therefore included
here. This chapter will instead concentrate on those computer systems
which might be used by the quantity surveyor acting as consultant to the
employer, specifically in the areas of contract administration and financial
management.
Contract administration and financial management
Although it is probably true that the automatic production of the Final

Account by computer is not yet feasible, mainly due to the amount of
human interaction and negotiation involved between the parties to the
contract, nonetheless knowledgeable commentators have long considered
contract administration and financial management to be among the most
fertile areas for the application of computers to quantity surveying. Gordon
Hunt for example wrote in 1981:
It is my personal view that microcomputers prove to be at their most cost
effective in a quantity surveying office when used for post contract
applications.
(Hunt 1981 a)
and Gordon Darby, well known pioneer of the use of microcomputers by
quantity surveyors is reported as saying:
Post contract operations are the most fruitful areas for micros. A large
reduction in the operating time on, say, valuations allows time for that
poor relation the final account under optimum conditions - namely
while the job is still in progress.
(Hunt 1981 b)
The contract administration and financial management functions to which
microcomputers have most often been applied include:
• interim valuations
• cost monitoring and progress reports
• calculation of fluctuations
• cash flow prediction
• prediction of 'out-turn' costs
• claims management
1 74 Computers and Quantity Surveyors
Note that use of the computer for cost analysis is discussed in chapter 9
and is therefore excluded here.
Interim valuations
The use of computers for the working up of interim valuations has been
pursued for some considerable time and a number of companies market
programs of this type, either as stand-alone programs or as modules of an
integrated suite.
There is a wide range of approaches to the problem. At one extreme
some surveyors have found it most cost effedive to use a combination of
word processing and spreadsheet software to simply calculate and print the
valuation statement complete with notifications to nominated sub-
contractors, whilst at the other end of the scale one can conceive of
systems in which the quantity surveyor actually prepares the valuation at a
screen on site, extracting items from a priced Bill held on-line, and in
which the computer automatically performs all of the calculations and
prepares the relevant documentation. It is perhaps in this area that the use
of portable computers or site-based computer terminals linked to a remote
central system is likely to expand rapidly in the future.
Much use has been made of microcomputers for so called 'tick-sheet'
valuations, where the quantity surveyor prepares his valuation using some
form of computer generated chart upon which are recorded all of the
details of work completed, variations etc. The method is particularly suited
to work with a high degree of repetition such as housing, and examples of
systems of this type are described in Hunt (1981 b) and Belcher (1983). An
example of a spreadsheet representation of the same problem is given in
Hawkins and Massey (1988).
Cost monitoring and progress reports
It is important that Employers are kept aware of changes in the estimated

cost of a projed throughout the progress of the work, and this is conven-
tionally done by providing the Employer with a series of financial statements
showing the effed of any variations at regular intervals as work progresses.
Whilst most of the integrated systems will provide for such a statement to
be produced at the touch of a button, it is also possible to use spreadsheet
software to create and maintain statements of this type.
Figure 11.1 shows a 'blank' spreadsheet of this type created using
Supercalc 2 which allows variations to be listed. In this example each
I A II B II c II D I lEI I F II G
11
21PROJECT:
31============================================================
41
51Tender value: .00
61===========================
71
BIDate:00/00/00
91=============
101
111List of variations and predicted final cost
121===========================================
131
141Arch. inst. Description Add Omit
151-------------------------------------------------------------
161
171
181Predicted final cost # .00
191
201
Figure 11. 1 Example of use of spreadsheet for cost monitoring
variation is recorded by simply inserting an extra row into the sheet after
line 16, and cell 018 contains a formula which allows the tender value
from cell CS to be adjusted by the sum of the values stored in columns
F16-F17 and G16-G17. As additional rows are inserted the spreadsheet
software itself adjusts this formula so that the amount given as 'predicted
final cost' is always the initial tender sum adjusted by the amount of total
additions and omissions from columns F and G. An example of the
spreadsheet in use, with the borders removed, is shown in figure 11.2.
Calculation of fluctuations
Most fluctuating price contracts now use one or other of the 'formula'
methods for the calculation of fluctuations. These methods rely on the use
of indices representing the average movement of prices for particular types
of work, and the most popular system for building works is the so-called
NEDO formula using indices prepared and published by central govern-
ment. The operation of this system is described in PSA (1977) and, whilst
PROJECT: Blanktown Primary School
Tender value: £ 450102.66
Date:17/09/88
List of variations and predicted final cost
Arch. inst. Description Add Omit
1 Take possession of site .00 .00

2 Steel frame P.C. 12050.63 10000.00
3 Revision to foundations 1250.50 .00
4 Heating P.C. 15210.00 17500.00
5 Omit provl.sums for rock .00 2500.00
6 Amend spec for facings 395.00 .00
7 Additional rooflights 1500.00 .00
8 Electrical P.C. 11472.36 12500.00
Predicted final cost £ 44 9481.15
Figure 11.2 Example of use of spreadsheet for cost monitoring
the method itself is quite straightforward, the lengthy calculations involved

mean that the calculation of fluctuations in this way is well suited to the
use of computers. Accordingly many organisations have developed pro-
grams specifically for this task.
The calculation has also been modelled using spreadsheets, and two
examples are given in Hawkins and Massey (1988).
Cash flow prediction
Cash flow essentially defines the pattern of expenditure on a building

project, and is therefore vitally important both to the contractor and to the
employer. Traditional methods of modelling cash flow relied upon calcula-
tions based on the contractor's construction programme, and many of the
computer systems developed for project management continue to calculate
cash flow on this basis.
It is however possible to predict cash flow patterns for a building

project independently of the contractor's programme using a method
initially developed by the Department of Health and Social Security and
first described in Hudson (1978). The method uses a series of'S' curves to
represent expenditure on a building project over time, a relationship which
had previously been recognised but not precisely analysed. The major
contribution of the DHSS work was the establishment of a precise mathe-
matical relationship between the variables involved, and the derivation of
a mathematical expression in the form of an equation and a table of
parametric values for projects of differing size. Whilst the original paper
presented by the DHSS team related purely to health buildings subsequent
unpublished empirical work appears to show that the results are equally
valid for other types of building. A spreadsheet representation of the
problem is again included in Hawkins and Massey (1988).
One further advantage of the DHSS cash flow modelling technique is
that, in addition to generating cash flow forecasts, the same algorithm may
be used, given past interim valuation amounts, to predict the contract
period. This feature is obviously very useful in cases where the contractor
appears to be falling behind programme.
Prediction of out-turn costs
Although, as stated earlier, it is common practice for clients to be provided

with financial progress reports during the post contract stage of a project,
such reports will often only consider the addition and omission of variations
from the tender value, and will make little or no attempt to forecast future
fluctuations. Whilst this approach is acceptable for contracts of short
duration in times of fairly stable prices, the same may not be true for
projects with an extended contract period or in times of high or rising
inflation. In these circumstances Employers may be much more interested
in the predicted total cost of the scheme, including fluctuations; the so-
called 'out-turn' cost.
Various statistical techniques can be employed which use the trends
revealed by a study of tables of past cost indices to attempt to predict
future cost trends, and simple BASIC programs using these methods have
been developed by a number of organisations for their own private use.
'Stand-alone' or integrated systems!
It should be obvious from the above that there are essentillly two
approaches to the use of computers for post contract work in the areas of
contract administration and financial management.
BUDGET COST
OUT-TURN
COSTS
FINAL
ACCOUNT
Figure 11.3 A cyclical view of construction cost monitoring
On the one hand lies what might be called the traditional'all-singing, all-
dancing' approach. This approach considers that since the process of
financial management of building contracts tends to be essentially cyclical
as shown in figure 11.3, centred on the use of cost data extracted from
priced Bills of Quantities, then computer systems should be linked together
to form integrated suites to facilitate this process. Such a philosophy
generally sees the central core of the system as an integrated database of
information based on the priced tender Bill. The post contract module of
such a system might typically provide facilities to handle:
• variations
• interim valuations
• cost monitoring and progress reports
• payment notifications to nominate sub-contractors
• formula fluctuations
• cash flow prediction
A diagrammatic representation of such a system is shown in figure 11.4.
Computers for Post Contract Work 1 79
VALUATION
VARIATIONS
NOTES
PAYMENTS TO
NEDO NOMINATED
INDICES SUB- CONTRACTORS
PROJECT DATA BASE
._.. BILL OF QUANTITIES

PRICED VARIATIONS
NOMINATED SUB-
CONTRACTOR PAYMENTS
INTERIM VALUATIONS
PAYMENT ADVICE
PROJECT COST TO NOMINATED
CONTROL REPORT SUB -CONTRACTORS
INTERIM VALUATION CASH FLOW

STATEMENTS REPORTS
Figure 11.4 An integrated aproach to post contract work
Suites of this type are usually menu-driven via a VDU screen and their
operation is quite straightforward. They do however tend to be rather
expensive and the cost effectiveness of such systems must be carefully
researched.
On the other hand, however, it is, as illustrated earlier, perfectly possible
to use software tools such as spreadsheets or even to write small self-
contained programs in a simple language such as BASIC to deal with
specific post contract tasks on a 'stand-alone' basis. The use of such tools
can be extremely cost effective, but one should always be conscious of the
factors discussed earlier in chapter 7.
The most important difference between the 'stand-alone' approach and
the integrated database concept described earlier is that in the latter case,
because all of the information is stored in one central set of files, then the
input of any new data via any of the routes shown in figure 11.4 will
immediately be reflected in all of the relevant outputs. Hence a change in,
for example, the estimated value of an Architect's Instruction will auto-
matically be reflected in reports s~..:ch as those for cash flow monitoring etc.
A number of firms market integrated post contract suites of the type

referred to above, but it would be quite possible for the quantity surveyor to
develop his own integrated system using data manipulation tools such as
dBase or Lotus 123. Note however that both the difficulty and time scale of
such development is usually grossly under-estimated and the comments
made in Chapter 7 regarding the development of systems in-house apply
particularly to this type of work.
The evaluation of contractual claims
As pressure from developers and building owners increases to reduce the

time available for pre-contract design and construction, so the planning
and control of construction projects becomes more critical, and the effect
of changes to the project either during the design or the construction
phases becomes more important in terms of both cost and time. Contractors
are now turning more and more towards computerised aids to project
planning in an attempt to control both time and cost within the more
stringent contractual conditions now imposed upon them. One predictable
result of this situation is an increase in claims by contractors, either in
respect of additional costs or for an extension to the contract period, and
many of these claims are now supported by reams of computer print-out
generated from the contractors management systems.
Under the JCT Standard Form of Building Contract both the architect and
the quantity surveyor have a role to play in the granting of extensions to
the contract period and in the agreement of any additional monies paid to
the contractor, and this form of contract, in common with a number of
others, now requires the contractor to produce and maintain a 'master
programme' in order that the consultants might be able more clearly to
assess the true effect of any delay or disruption which may occur. Whilst
this is a welcome first step, it is of limited value since planning of the works
remains the contractor's responsibility and the contractor is not required to
divulge the information on which his plan is based. It therefore follows
that, in the majority of cases, the consultants involved can only form very
subjective judgements about the accuracy and correctness of the plan put
forward by the contractor. In addition most consultant architects and
quantity surveyors do not clearly understand the workings of computer
based planning packages, and it is therefore very difficult, once some delay
or disruption occurs, to argue convincingly against claims submitted by the
contractor on the basis of a variance with his original plan, particularly
so when the plan has been generated from a computer based project
management and planning package.
In an attempt to ease this situation some organisations responsible for
large projects now not only require the contractor to prepare a master
programme, but also require him to plan the project using a specified
computer based project management system in which the consultants have
a degree of expertise. This method of working thus enables the consultant
to check rather more thoroughly the contractor's progress against the plan,
to evaluate much more readily the time consequences of a change to the
plan, and even to indulge in 'what-if' simulations in order to predict the
best ways in which the required changes can be implemented.
This process has become known as 'claims management', and the use of
computer based techniques such as those indicated above appears likely to
increase in the near future.
A more comprehensive discussion of the use of project planning and
control systems by the quantity surveyor is given in chapter 13.
Portable and site-based computers
Recent developments in computer hardware have made possible an

increasingly wide range of computer equipment which is rugged enough
for use on site. Site based computer equipment has been in use by some of
the larger building contractors for some time, particularly on large sites,
and may take the form either of computer terminals linked to a remote
mainframe computer, or, increasingly more likely, site based micro or
minicomputers linked to the company mainframe. Such systems can be
used for a wide variety of construction management tasks including,
among others, ordering and logging of materials, cost and bonus calcula-
tions, cash flow forecasts, cost/value reconci Iiation exercises and con-
struction planning using software such as that considered for project
management in chapter 13. Whilst such systems have always been useful
to the contractor's surveyor, it is only recently, with the increasing use of
alternative forms of procurement such as management contracts where the
consultants and the contractor work much more closely together, that the
consultant quantity surveyor has been allowed access to such systems.
Another development which has great potential for future use is the wide
range of portable computers now available. Until comparatively recently
such machines had limited memory and were largely used in industry as
'data loggers', collecting information for subsequent transmission to a
larger computer for subsequent processing. Typical examples are in the
retail industry for stock control, and in land surveying where such machines
are often used in conjunction with electronic theodolites to collect survey
data for later processing by an office based survey plotting system, but
such machines are now sufficiently powerful for surveyors to begin to use
them for actually processing data on site. The potential uses, particularly
for interim valuations and site measurement, are plain to see.
References
Belcher R. (1983) 'Speeding housing estate valuations' Chartered Quantity Surveyor june
1983 pp. 417-421
Hawkins P. and Massey R. (1988) Spreadsheet examples Department of Surveying, Trent
Polytechnic
Hudson K.W. (1978) 'DHSS expenditure forecasting method' Chartered Surveyor Building
and Quantity Surveying Quarterly Spring 1978
Hunt G. (1981a) 'Micronotes- finding the data' Chartered Quantity Surveyor February 1981
p.263
Hunt G. (1981b) 'Micronotes- micros at the post contract stage' Chartered Quantity Surveyor
March 1981 p. 285
PSA (1977) Price adjustment formulae for building contracts: Guide to the application and
procedure HMSO
12 Quantity Surveyors and
Computer Aided Design
Objectives
• Discuss the production of quantities directly from a computer aided
design system in an integrated practice.
• Comment on the possibilities for the extraction of data from the
architect's project model by other private practice quantity surveyors.
• Discuss the problems which govern the exchange of data between
different computer systems.
• Comment on the legal and economic issues involved in the use of the
architect's data by the quantity surveyor.
CAD - some general considerations

This chapter is concerned with the u~e of computers for design and
draughting within the building industry, and the possible effects that a
wider spread of such systems might have upon the work of the quantity
surveyor.
Computers have been used to aid the production of drawings for many
years, especially in the engineering disciplines, and it is therefore perhaps
a little surprising that, as recently as May 1987, Edgill and Atkin,
commenting upon a project to investigate the opportunities such systems
might hold for quantity surveyors, stated that:
... it was evident that many misconceptions (about computer aided design)
were held by quantity surveyors. Whatever these amounted to, it was
clear that CAD was not well understood.
(Edgill and Atkin 1987)
It is therefore obviously necessary to examine a little of the background of
computer aided design before attempting any kind of analysis of the effects
that the use of such systems might have on the quantity surveyor.
CAD - what is it?
One of the problems about CAD is that this particular set of initials has at
least two generally accepted meanings. CAD may, in some circumstances,
mean computer aided design, where the computer is used to actually assist
in the design process, but the same initials may equally well be used to
refer to computer aided draughting, where the computer is merely used as
a tool to produce working drawings. In an attempt to ease the confusion
some commentators refer to computer aided architectural design (CAAD),
whilst others prefer CADD (computer aided design and draughting). The
term CAD is used here to denote computer aided design.
Computers have been used to aid the engineering design process for
some 25 years, and their use in areas such the car and aeroplane manu-
facturing industries is well known. In industries such as these links have
long been formed between the computers used for design work and those
used to control the machines which actually make the parts from which
the final product is assembled; thus CAD/CAM (computer aided design/
computer aided manufacturing) is yet another acronym which the use of
computers has introduced to the vocabulary of the modern industrialist.
Almost from the very beginning engineering computer aided design
systems incorporated facilities to enable the computer to generate not only
drawings and instructions to drive machine tools but also to produce a 'bill
of materials', a list of components required to produce the final composite
assembly. The potential application of this technique to construction was
quickly recognised in theory but proved rather more difficult to achieve in
practice.
The amount of computing power required to 'model' designs for cars
and aeroplanes was huge, and more rather than less was required to model
large buildings. In addition companies involved in producing cars, aero-
planes and similar products tend to be very large, dealing in general with
sizeable production runs of a given product; construction on the other
hand tends to be a much more 'one-off' affair and most architectural
practices of the time were, with a few notable exceptions, unable to afford
the the hardware required. It therefore tended to be in the large local
authorities, largely on the back of the very large public sector building
programmes of the early 1970s, that computer aided architectural design
began to blossom. Similar developments had already taken place in civil
and structural engineering with computers being used to aid the design of
projects such as the St Lawrence Seaway and the Sydney Opera House,
and many widely used programs were developed for highway design, the
design of reinforced concrete and steel structures etc.
Probably the earliest comprehensive published account of computer
aided architectural design was Campion (1968), but by the mid- to late-
1970s a number of computer aided design systems were in regular use;
Quantity Surveyors and Computer Aided Design 185
Madden (1973) for example describes a system which had been in use in
West Sussex County Council since 1968, Carter (1973a-d), in a series of
four articles explored the application of computers in architecture, Maver
(1978) catalogues a number of contemporary CAD systems, and Daniel
(1979) and Hawkins and Davidson (1979) both describe ways in which
quantities could be extracted from such systems.
As with most other applications of computers a widespread interest in
the techniques really began with the general introduction of low cost
microcomputers during the late 1970s since which time a number of
comparatively low cost CAD systems have been developed.
Types of CAD system
There are essentially two types of system in common use. The majority of
the cheaper systems, usually based on small computers, will be described
as 2D (two-dimensional) systems, and are often little more than automated
draughting tools. They will generally permit the creation of working
drawings in a flat (i.e. two-dimensional) plane, often allowing commonly
used shapes or symbols to be defined, stored in a library and subsequently
retrieved and placed in a drawing, and will also provide functions to allow
shapes to be scaled, deformed or rotated in various ways.
In general however each drawing is a separate entity in its own right and
there is no transfer of information between the various drawings which
make up a project, hence if changes are made to one drawing they will not
automatically be reflected in other drawings for the same project.
Some two-dimensional systems do permit shapes to be given the
additional attribute of a third dimension, thus allowing some limited
facilities for the production of, for example wire-frame perspectives. Such
systems are commonly known as 2.5D systems, but the facilities offered
vary from system to system. It is therefore very difficult to define precisely
what a 2.5D system actually is.
The most sophisticated systems, in my view the only ones which can
rightly be called computer aided design systems, are those offering a true
three-dimensional modelling capability. Such systems are obviously known
as 3D systems. The most common approach for systems of this type is for
individual components to be defined to the computer in terms of plan,
front elevation and side elevation views, generally using the projection of
the object onto the top, front and side of an enclosing box. The computer
is then instructed as to the position of the component in three dimensions
within the building, and by this means a complete 'model' of the proposed
building is constructed within the computer's memory. With systems of
this type therefore individual project drawings are generated by asking the
computer to view the model from the desired direction. 'Layered' drawings
can be produced by instructing the machine only to look at particular

groups of components; thus it is possible to generate, for example, a plan
view of a building showing only the structural walls, as well as a complete
floor plan showing all ofthe features stored.
The use of 3D techniques has many advantages over 2D or 2.5D
systems, some of the most significant being:
(i) Once the original data are input views may be produced at will at
any level either in plan, elevation or section and to any scale.
(ii) Perspective, orthographic or isometric projections may be generated
on demand from any viewpoint and with hidden lines removed.
(iii) A change in the shape or position of any of the components will
instantly be reflected in any drawings subsequently produced.
It is therefore obvious that 3D systems are much more flexible than their
2D or 2.5D cousins; they are however, in general, much more expensive,
and require greater computer power to work effectively.
Most of the modern generation of 3D systems allow views to be displayed
in colour on a suitable high resolution monitor, and this can be of great
benefit to the architect in the initial design stages of a project, for
presentation to a client, or to assist with the obtaining of planning approval
for controversial sites. Techniques have been developed whereby computer
generated perspectives have been combined with photographs of the
street-scape in order to enable clients and planners to visualise the environ-
mental effects of a new building, as well as techniques using a series of
internal perspectives edited onto video tape giving an illusion of 'walking
through' the computer model. There is no doubt that this 'public relations'
aspect of CAD will assume increasing importance in the future, and some
researchers are already speculating on the development of systems using
laser-generated, three-dimensional holographic representations of the com-
puter model. Such visions may appear futuristic and fanciful, but then, ten
years ago, so did the concept of a lap-top computer and a pocket-sized
television!
CAD and the quantity surveyor
From the very beginning the use of computer aided design systems has had
a potential impact on the work of the quantity surveyor, primarily in the
areas of measurement and initial cost advice.
As regards measurement, we have already seen that CAD systems in
engineering produce bills of materials almost as a by-product, and the
construction of a building was perceived to be, in essence, no different
from the construction of an aeroplane. There seemed therefore no obvious
reason why Bills of Quantities should not be extracted from the CAD
model in the same way.
The impact upon the extent of the quantity surveyor's role in giving
initial cost advice is equally obvious since one of the primary reasons for
architects using CAD systems in the first place was to enable a wider range
of options to be examined and evaluated within the same time scale as
traditional methods.
We have also seen that the early building CAD systems were largely
developed within integrated practices; quantity surveyors were therefore
on hand and could collaborate actively in the system development.
Problems of data transfer between one system and another therefore tended
to be minimal.
Automatic measurement from CAD models
We have seen above that the majority of 3D CAD modelling systems, and
a number of 2D and 2.5D systems, operate through the use of 'components'
defined in a library held in the computer. A component might be a discrete
unit, for example a window, a door or a sanitary fitting with fixed
dimensions, or alternatively a 'stretchable' component such as an area of
floor slab or an area of brick wall. Components may also be composites
formed from combinations of other components, for example a standard
house type, and all components are generally referenced by some sort of
code.
This technique has much in common with the 'unit quantity' techniques
already in use in quantity surveyors' billing systems and described earlier
(chapter 9), and since the CAD systems already possess the ability to
'count' the number of each type of component used in a model the
integration of CAD and quantity surveyors' billing systems was, in a
number of cases, successfully exploited. Considerable success was achieved
in the 1970s particularly in the local authority sector, in West Sussex,
Derbyshire and Clwyd County Councils among others, and a typical
example is described in Daniel (1979). More recent work using the same
techniques is described by Wager and Wilson (1986), whilst a rather
different perspective on the same problem envisaging the use of data
extracted directly from a CAD model and subsequently converted to SMM
form via a spreadsheet is briefly described in Curran (1987).
Cost advice
The other major area of influence on the quantity surveyor's role lies in the
provision of initial cost advice for a number of different options on a given
scheme, and this problem was historically solved by including costs with
the graphical description of the component stored in the CAD system

library. Since, as we have already seen, the CAD system itself was able to
total the numbers of each component used in any particular model, the
generation of project cost became almost a trivial matter which simply
required the quantity surveyor to update the computer forecast for time,
location and local difficulty. The evaluation of alternatives during the
development of the design thus became much more straightforward.
Some system developers exploited this technique in other directions,
adding additional data to the component description such as heat loss co-
efficients or cost in use information. The computer was then able to
generate total heat losses, cost in use reports etc. An account of such a
system is given in Jelly (1977).
Transfer of data between computer systems
It has already been noted (chapter 2) that grave problems arise when
attempting to transfer data from one computer system to another, and this
is particularly true where CAD systems are concerned. Whilst it has
already been noted that data were transferred between CAD and quantity
surveyors' billing systems, this transfer in general took place among
machines of the same type, and generally involved the output of specially
processed data from one machine onto some physical medium, often
punched cards, which could subsequently be read back into another
machine. This method of transfer is extremely expensive and inefficient in
that it requires a special program to convert data from the host machine
into a suitable form for processing by the target system. Add to this the fact
that, on any particular project, data might also need to be exchanged
between the architect and quantity surveyor and the services and structural
engineers and the number of permutations, and hence the number of
special translation programs required becomes unmanageable.
Wix and Mclelland (1986) conducted a study of the problem, and report
on the development of a data exchange standard for graphics systems
called the Initial Graphics Exchange Standard (IGES). The central concept
underlying IGES is the establishment of a standard format to which all
graphics data should be converted. Hence if an archited with one CAD
system wished to transfer some data to an engineer with another system
then the architect would simply instruct his computer to output the data in
IGES format and the engineer would instruct his system to expect the input
in the same form. Each separate type of CAD system thus only needs to
provide one output processor and one input processor to be able to
communicate with all other IGES systems. IGES or systems like it could
therefore, in the future, provide a common interchange format for data
transfer not only among CAD systems but also between CAD systems and
quantity surveying packages, project management systems and the like.
Other attempts at standardisation of format and equipment are described
in RIB E. V. (1982) and CICA (1982).
Legal and economic issues
It has been shown above that a considerable amount of information useful

to the quantity surveyor can already be extracted from CAD systems, in
particular measurement and cost information, and there have been those
who suggest that this might serve to bring the two professions closer
together.
Whilst this may be very welcome within the context of an integrated
pradice, problems of responsibility for the accuracy and completeness of
the information produced may occur where more traditional consultancy
relationships apply. It is for example very easy to imagine a scenario in
which a quantity surveying pradice prepares an early estimate of the cost
of a project using quantity or even cost data extracted from an architect's
CAD system, but what would be the consequences if the estimate proved
to be wildly inaccurate because the original information was incorrect?
The answer is surely that the quantity surveyor, since he is responsible for
the estimate and is presumably being paid a fee for producing it, must
carry the responsibility for the error, and this leads one to the view that the
quantity surveyor would need to check very carefully any information
provided by other consultants. Such checking could well, of course, take
longer and be more costly than preparing the same information manually
in the first place, but it is a well recognised phenomenon that people are
likely to accept computer produced information at face value, and it is
interesting to speculate as to how many quantity surveyors would be
prepared to take the risk.
There are also economic fadors to consider. Again the problem is less
acute within an integrated pradice, but where a number of independent
consultants are concerned one can easily imagine architects' reluctance to
provide the quantity surveyor with quantity or cost information free of
charge.
Some have speculated that one way in which the above problems might
be overcome would be for the quantity surveyors themselves to invest in
CAD systems, but this appears at present to be unlikely at least on a large
scale. Whatever the outcome it is clear that there are influences here
which might well act against the wider dissemination of information from
CAD systems in the immediate future.
Benefits and disadvantages to the quantity surveyor
The use of CAD systems has proved both good and bad from the quantity
surveyor's point of view. In terms of benefits the automatic generation of
quantities and cost information has undoubtedly in many cases proved
most helpful, saving both time and cost and providing the client with better
value for his money.
On the other hand however there have been some disadvantages:
(i) The creation of the unit quantity libraries, described in chapter 9,

which are necessary to take advantage of the automatic measure-
ment facilities discussed above is very time consuming and very
costly. The quantity surveyor needs to be assured of a number of
uses if the initial investment is to be worthwhile. It is not therefore,
in general, worth doing for a 'one-off' scheme, but only where the
same components or constructional detai Is are to be re-used as
'standards' on several projects. This method tends therefore to be
much more attractive in the context of an integrated practice than
in the case of a consultant quantity surveyor who may only work
occasionally with a particular architect. It can however work well
in the case of a programme of similar projects and a good example
is the redevelopment of petrol filling stations by Mobil Oil described
in Wager and Wilson (1987).
(ii) Once the components have been defined and stored in the CAD
system library the architect must be prepared to leave them
unchanged as far as possible. This seems self-evident, and indeed
one might consider that the ability to recall tried and tested
constructional details might be one very good reason for having a
CAD system in the first place. It appears however that this is not
the case. A major survey of communications in the construction
industry conducted in 1985/86 found that:
One surprising finding which emerged was how little CAD systems
seem to be used for producing standard details in architects'
offices... In practice we found that this hardly ever happened
although in the case of one architectural practice there had been a
considerable investment creating such a library. However when it
came to to using the library it was found that architects preferred to
work out their own details for a job rather than resorting to those
already available.
(Day, Faulkner and Happold 1986:p. 20)
Whilst changes made to a component may be minor in graphical

terms, perhaps a change to a dimension, the effect upon the quantity
surveyor's pre-measured quantities can be catastrophic!
(iii) Even worse than (ii) above, although in a way related to it, is the
tendency shown by some architects to manually alter drawings
produced by a CAD system. Again the architect will often see such
changes as trivial but the possible effect upon pre-measured
quantities is again obvious. Manual amendment of computer
generated quantities to take account of such changes is usually
very difficult and the modification and adjustment of such quantities
may take longer than measuring from scratch by manual methods.
(iv) The information flow from the architect may become disjointed by
the use of CAD. Day, Faulkner and Happold (1986) found that:
Because a whole set of drawings can be plotted relatively quickly,
there is a tendency to leave the plotting until immediately before
the due date for the information in order to ensure that all amend-
ments are incorporated.
and later:
... it appears that where there is the potential to issue a 'perfect'
drawing there is much greater reluctance to issue any drawing
until that perfect drawing is available.
The future impact on the quantity surveyor
The probable future impact of an increase in the use of CAD techniques on

the role of the quantity surveyor is very bard to predict. On the one hand
Bruce Graham, senior architect with the London office of Skidmore,
Owings, Merrill is reported as saying that:
The quantity surveyor is not necessary with the computer. The computer
measures the quantities much more accurately than a quantity surveyor.
whereas on the other hand Edgill and Atkin (1987) assert that:
It (CAD) will only present problems if quantity surveyors fail to recognise
the impact it will have on present methods of working. CAD is only
likely to be a threat if it is perceived to be one, with little or no attempt
being made to investigate its positive aspects.
The diversity of these two views shows neatly the difference of opinion
which presently exists. Whilst it is tempting to disregard Graham's view as
the opinion of one who apparently has only the sketchiest of notions about
what a quantity surveyor does, there is really little comfort to be gained
from Edgill and Atkin's view that CAD is only a threat if it is perceived to
be one.
Clearly the spread of powerful low cost CAD systems will influence
dramatically the role of the quantity surveyor, and it is therefore, by some,

perceived to be a threat. Brandon (1983) for example says:
For the first time in the history of mankind, technology has produced a
machine, the computer, which will not only enhance human abilities in
certain areas but also replace them in a more efficient manner. Measure-
ment as a technique (not only quantity surveying measurement) is
threatened because, by its very nature, it is usually an unambiguous
process which can be wholly described in a logical manner and which
relies heavily on another process, design, planning or creation.
It is, I believe, inevitable that some changes will take place in the quantity
surveyor's role; the question is no longer whether changes will occur but
when will the changes occur and what will the scope of them be.
The period since 1983 has already seen evidence of fundamental
changes; quantity surveyors are concentrating less on their historical role
as measurers and valuers and are instead promoting much more aggressively
the image of the quantity surveyor as manager and facilitator of the
bui Iding procurement process. Such a movement is obviously totally consist-
ent with the pressure which a spread of CAD activity would impose, and
those quantity surveyors who are prepared to make the required role shift
will, I believe, be far more likely to survive than those with more traditional
views.
What is much more difficult to predict is when the changes will
occur. Some commentators seem to imply that the change is imminent and
will be dramatic, but the pace of change depends on many factors. In
order for such a scenario to happen a large majority of architects and
designers would need to become active users of 3D modelling systems for
the majority of their work. Whilst the most recent published survey (Peat,
Marwick, Mclintock 1987) shows that 62 per cent of architectural practices
surveyed presently use CADD software with a further 16 per cent forecasting
use in the future, indications are that the majority of these employ low cost
2D or 2.SD systems simply to cope with the draughting work. Such
systems are in general quite unsuited to the generation of quantity data in a
form necessary to invite competitive tenders. In addition 3D CAD systems
are still very expensive, and both Wager and Wilson (1986) and Atkin
(1987) report that even these systems cannot themselves generate quantity
data in an acceptable form at the present time without the links to quantity
surveying software reviewed earlier.
Whilst it is no doubt inevitable that the price of 3D CAD systems will
fall, probably dramatically as new computing technologies such as parallel
processing become widely available, and the problems posed by the
conversion of graphical data into a form required for tendering will eventu-
ally be resolved, experience seems to show that the latent inertia of the
construction industry will ensure that the widespread adoption of these
techniques by the rank and file is likely to take a lengthy period of time.
What is perhaps likely to happen in the long term (say within the next
twenty years or so) as computer systems become more complex and the
problems of data exchange between systems are resolved is a fundamental
change in the tendering process itself to a stage where bills of quantities
are no longer necessary. Nationally agreed methods of measurement are,
after all, a compromise between what the contractor actually wants and
what the quantity surveyor is prepared to provide; the bill gives tendering
contractors little information about 'buildability' and yet the contractor's
assessment of this factor is arguably the one thing upon which he needs to
exercise his skill if his tender is to be successful and he is to gain an
adequate return on his investment.
It is, for example, easy to imagine a scenario in which tendering
contractors are each presented with a copy of the architect's CAD model in
computer readable form. From such a model contractors could not only
extract for themselves details of the quantities of work and materials
required for the scheme but could also, in conjunction with their own
management software, perform 'what-if' simulations in order to determine
the most efficient and economical method of construction bearing in mind
their own particular resource constraints.
A further factor which will influence the speed with which such changes
come about is the economic state of the construction industry. As we have
already seen, many of the earlier developments in CAD can be traced back
to a period of intense constructional activity in which standardisation and
computerisation became almost hallowed words. A significant increase in
demand could well accelerate the process of change, and this factor is
itself likely to be affected critically both by the possible influx of European
companies following the removal of trade barriers in 1992 and also the
increasing activity of foreign firms such as the major Japanese construction
companies within the United Kingdom.
References
Atkin B. (1987) CAD techniques: opportunities for chartered quantity surveyors Royal
Institution of Chartered Surveyors
Brandon P. (1983) 'The writing on the wall' Chartered Quantity Surveyor May 1983 pp. 378-
379
Campion D. (1968) Computers in architectural design Elsevier
Carter j. (1 973a-d) 'Computers and the architect' The Architect's journal 3/10/73 pp. 81 5-
817,10/10/73 pp. 865-870,24/10/73 pp. 1003-1011 and 31/10/73 pp.1053-1060
CICA (1982) The specifiCation of a building industry computer workstation Construction
Industry Computing Association
Curran j. (1987) 'The quantity surveyor and the CAD system' Chartered Quantity Surveyor
October 1987 p. 11
Daniel P. (1979) 'Bills of quantity in a multi-disciplined design system' Chartered Quantity
Surveyor May 1979 pp. 141-142
Day A., Faulkner A. and Happold E. (1986) Communications and computers in the building
industry Construction Industry Computing Association
Edgill B. and Atkin B. (1987) 'CAD opportunities for the quantity surveyor' Chartered Quantity
Surveyor May 1987 p. 9
Hawkins M. and Davidson j. (1979) 'Computer aids Bills of Quantities' Chartered Quantity
SurveyorApril1979 pp. 99-105
jelly P. (1977) 'Project estimating by computer - a system in use' The Quantity Surveyor Apri I
1977 pp. 161-165
Madden L.W. (1973) 'Computer programs for We'll. Sussex' Building4/5f73 pp. 103-107
Maver T. (1978) 'The benefits of using computer aided design' RIBA journal March 1978 pp.
101-104
Peat, Marwick, Mclintock (1987) Building on IT Peat, Marwick, Mclintock Publications
RIB E.V. (1982) Feasibility study of common 1/0 conventions for the building industry
Rechen-und Entwicklungsinstitut Fuer Edv im Bauswesen
Wager D. and Wilson R. (1986) CAD systems and the quantity surveyor Construction Industry
Computing Association
Wix j. and Mclelland C. (1986) Data exchange between computer systems in the construction
industry Building Research and Information Association
13 Computers, Quantity
Surveyors and Project
Management
Objectives
• Identify those parts of the project management process to which
computers are most likely to be applied.
• Comment on the early history of computer based project management
systems.
• Identify and comment on the features which a good project manage-
ment system should possess.
• Discuss the advantages and disadvantages of modern microcomputer
based project management software.
• Comment on the use of computer systems in the selection of an
appropriate.procurement methodology.
• Comment upon the possible uses of computers in quality control.
Introduction
Preceding chapters have considered the use of computers in 'traditional'
quantity surveying activities, and it is perhaps appropriate that the book
should end with a brief examination of the use of computers in project
management since it is, in a sense, through the management function that
all of the individual facets are united.
Quantity surveyors and project management

(Project management) is not for the faint hearted. The service attracts a
high level of liability, needs an aggressive approach and will require
quantity surveyors to open their doors to those who have not trained as
Chartered Surveyors. With this service there is a tendency to become
more commercial and less professional, especially if one moves towards
the contracting end of the spectrum. Finally to achieve the rewards this
service can offer, one must be entrepreneurial, flexible and be prepared to

take reasonable business risks.
(R. Morris, Hanscombe International Inc., Chicago; RICS Annual Confer-
ence 1986)
The increasing interest shown towards project management is no doubt

based upon the perceived opportunities of a new market in which to
apply quantity surveying skills. While this may be true in some cases,
the offering of a complete management service involves more than just
the repackaging of traditional quantity surveying. Managing complete
projects requires specific skills which some quantity surveyors may not,
as yet, possess.
(Brian Atkin, Lecturer in Construction Management, University of Reading;
Chartered Quantity Surveyor September 1986.)
Despite comments such as those recorded above, increasing client dis-

satisfaction with traditional building procurement methods in the United
Kingdom has led a number of quantity surveying practices to move away
from the traditional role of measurer and valuer, and instead to assume the
role of manager of the procurement process.
There have been notable successes; perhaps the best known is sti II the
construction of the National Exhibition Centre in Birmingham, although
others are reported from time to time.
Notwithstanding the fact that this form of procurement is no longer
unusual, there still appears to be some confusion of terminology, with
terms like 'project management', 'construction management' and 'project
co-ordination' apparently being used almost synonymously. Project manage-
ment is considered here to include the provision of all of the skills required
to oversee the development process from inception to the hand-over of the
completed building, whereas construction management is considered as
that part of the total process concerned with the construction phase.
Construction management can therefore be seen as one part of project
management in its widest sense. Project management to most quantity
surveyors implies the high level management and co-ordination of a
project rather than the detailed management of the complexities of construc-
tion, and construction management is therefore likely to be done by a
specialist under the control of the project manager.
The areas in which computer systems are likely to be most useful are
therefore:
(i) co-ordination of the design phase

(ii) co-ordination and control of the construction phase
(iii) choice of a suitable contractual method
(iv) quality control and checking procedures.
Computers, Quantity Surveyors and Project Management 197
It must always be remembered however that although the computer may

be used to aid the procurement process, effective management of any kind
is not achievable by purely mechanistic means. No computer system can,
for example, tell the project manager what to do about the consultant who
persistently fails to meet his agreed programme dates or the contractor
who, at every site meeting, promises that tomorrow he'll have more labour
on site, and despite the fact that he's lost another three weeks in the last
four week period he'll pull it all back eventually and finish on time. The
computer can only help the project manager to evaluate the alternatives, to
show what the consequences of a particular course of action are likely to
be in terms of time and cost. Good project management software is no
substitute for good management.
Project management techniques
Formalised project management techniques had their infancy in the 1950s,

and were extensively developed in the United States during the Polaris
programme of the late 50s and early 60s. The best known techniques were
Critical Path Method (CPM) and the Programme Evaluation and Review
Technique (PERT).
By 1960 these techniques were implemented on mainframe computers
on both sides of the Atlantic, and their principal use was in the project
based industries such as aerospace and construction. The techniques were
intended to aid the qmtrol of time, cost and resources, and as is often the
case were perceived by many in the construction industry to be a panacea
for all ills. There was therefore considerable initial enthusiasm since the
techniques apparently offered sophisticated methods of analysis applicable
to many typical project management problems, for example the problem of
time/cost trade-off.
As with so many early applications of computers, however, the early
promise was not fulfilled and there was a disappointingly slow spread of
the technique. The main reasons for this seem to have been:
(i) Small and medium sized firms were unable to afford to purchase a
computer of their own and hence used bureaux services. These
were cumbersome, expensive and time consuming.
(ii) The gradual realisation that computer software, no matter how
clever, was no substitute for good management.
(iii) The technique required substantial volumes of data to be input.
Most firms did not have sufficiently detailed information about the
behaviour of costs and their relationship with work methods.
(iv) The systems were not user friendly, some required specialised staff
and they tended to be remote from the managerial work-face.
Some, usually large, organisations who could afford to install in-house

computers and who had sufficient information persevered with the tech-
nique, but the majority of companies either abandoned its use after a short
time or never implemented the technique in the first place.
It was not, as with so many applications of computers, until low cost,
powerful microcomputers became readily available at the beginning of the
1980s that a dramatic new groundswell of interest in project mamagement
systems occurred, and this renewed interest coincided with the drive by
clients for shorter development periods requiring much tighter planning
and control than had been the case in the past.
Features of a good project management system
The characteristsics of a good project management system may be summar-

ised as follows:
Simplicity
There seems to be a belief among academics and some management

consultants that for any technique to be successful it must be complicated
and only operable by highly qualified staff with deep specialist knowledge.
Not only is this plainly untrue in project management it is arguably the
most common cause of failure.
For any plan to work successfully it must be understood by everyone in
the team, it must be communicated clearly and its authority must be
unassailable.
Systematic
For any plan to succeed each activity must be considered systematically

and its effect upon the project evaluated. The results of actions must then
be measured against the plan to ensure that the plan is proceeding smoothly
towards its objectives.
Flexibility
Many project plans fail because they are too rigid. Any plan must be
dynamic and capable of rapid change to accommodate changes in circum-
stances. It is said that the only thing which is certain about any plan is that
it will go wrong, so when it does management must be able to evaluate the
effect of the problem and initiate the necessary corrective action.
Responsibility
Every member of the team must know precisely what his responsibilities
are for the cumulative success of the total plan, and must be aware of the
probable effects of failure to deliver.
In order that these characteristics may be effective the plan must include
the following:
Objectives
A clear statement of objectives is needed before any project is begun. In
most cases these will be the conditions expected to apply once the project
is complete and will generally be expressed in terms of time, cost and
quality.
A plan of action and a programme

The plan is simply a step-by-step statement of the action which has to be
taken, by whom, when, at what cost and to what specification. A pro-
gramme is simply a plan on a timescale.
A means of measuring performance

A plan is of no value unless it is capable of. being monitored to ensure that
it is moving smoothly towards the objective. To do this the plan must be
designed so that 'milestones' are incorporated in order that reviews may
take place at critical points. Criteria for measurement must be established
at each milestone, again normally in terms of time, cost and quality.
The result of such measurement is normally a variance, and in this event
the plan must be capable of being recast in order to place the project back
on track.
A means of comparing achievement with initial objective

When the project has been completed it is necessary to compare what has
been achieved with what was originally intended.
Software for co-ordination and planning
Operations research has provided several tools useful to project manage-

ment, and those most often implemented in software of this kind are
spreadsheets to show resource allocation, bar or Gantt charts and network
analysis. All of these will normally be combined in a modern package.
At the heart of most systems is an algorithm for the calculation of the

route through the network which links all of the most critical activities.
This will normally show the quickest route to completion, and any overrun
on a critical activity will cause an extension to the projected total project
time. In practice of course this may not be strictly true since subsequent
critical activities may be completed ahead of schedule and the lost time
may therefore be regained.
In an effort to mitigate this problem some techniques, for example PERT,
allow up to three durations to be set against each activity, namely an
optimistic time, the most likely time and a pessimistic time. Planners can
therefore explore probabilities, assess the effects of changes in more detail
and make contingency plans in advance of problems actually occurring.
Atkin (1986) contends that few commercial packages handle probabilities,
but this is probably because few construction industry planners are
sufficiently au fait with probability techniques to properly understand their
use. Software suppliers assert that if such facilities were to be requested
they could readily be made available.
Resources can also be assigned to activities. Often the duration of an
activity implies a minimum resource usage derived from the work method.
In other situations the duration is assigned regardless of work method. The
resource pool available for a project can be stated and the resource
forecast is calculated by aggregating the requirements and availability for
each project time period. The result is often displayed as a histogram.
Where the timescale requires a resource level higher than that available an
overload occurs, and these will need to be eliminated by adjusting the
plan, a process known as resource levelling. Many computer systems claim
to offer resource levelling, but in reality only a few systems can achieve
this to a satisfactory degree.
Once a project schedule is determined in terms of time and resources it
can be stored and used to generate cost forecasts, resource requirement
reports etc. for the use of the project team. This 'baseline' can then be used
for comparison purposes as the work proceeds.
There is a very wide variety of packages available. Most will produce
networks and Gantt charts; however some packages are network based and
can handle extensive manipulations of networks, while some packages are
Gantt based and can handle extensive manipulation of programme charts.
The range of facilities offered with some of the newer medium priced
packages may be greater, but for speed and capacity the larger packages
are usually better. Some packages require 'bolt-on' graphics extensions, or
may require their file contents to be exported to external databases or
spreadsheets before display. The smaller packages are usually limited in
facilities, although the facilities available do vary widely from one package
to another. Many small packages are not aimed specifically at the
construction industry and it will be necessary to take this into account

when evaluating them in terms of value for money.
Some packages are now available which provide ready made 'templates'
for the design stage of building projects, including standard networks and
checklists covering concept, design development and tender documentation
stages according to different contractual routes. Hunt (1987) describes
such a system.
Costs may range from a few hundred pounds to several thousand, and
generally the price band tends to reflect the facilities incorporated in the
package. The low cost systems may have limited logic linkages, simple
resource aggregation, elementary costing and limited activity handling. The
more expensive packages are likely to offer several different types of logic
linkage, resource levelling, multiple report formats and more sophisticated
tracking, budgeting and costing features. A guide to the evaluation and
selection of project management software is given in Barrett (1987).
In terms of hardware some of the smaller packages will run successfully
on an IBM PC with one disk drive and a minimum of 512K RAM, whereas
the more sophisticated packages are likely to require at least 640K RAM
and 20MB hard disk.
A more detailed exposition of the use of computers in construction
planning and control can be found in Jackson (1986).
Computer software designed to advise on procurement method
The choice of procurement method for any particular construction project

will normally be made taking into account a number of factors concerning
the client, the time available for the scheme, cost and the type of project.
An incorrect choice is likely to lead to dissatisfaction on the part of the
client either during the progress of the scheme or upon completion, and
the problems will generally concern time, cost, quality or the degree of
accountability, either individually or in combination. It is therefore import-
ant that the correct choice is made at the inception of the project, since if
the choice of procurement method is grossly inappropriate the scheme is
almost certainly doomed from the start no matter how good the subsequent
project management.
Morledge (1987) develops an algorithm to guide the choice of an
appropriate procurement methodology, implemented using a simple spread-
sheet package, and an alternative approach to the same problem is proposed
by Brandon, Basden, Hamilton and Stockley (1988).
Both systems attempt to attach value weightings to a wide range of
factors under the general headings given above, and thus attempt to rank
the most suitable procurement strategies.
Whilst in the majority of cases the most appropriate procurement
strategy often appears fairly obvious, programs such as these can assist in
clearly identifying those factors which are important in any particular
scheme, particularly perhaps in the case of an inexperienced client whose
true needs and priorities may not actually be those articulated in the
project brief.
Software for quality control and checking
Whilst at present the construction industry makes little use of computer

technology for this purpose, systems for quality control are in use in other
industries.
A good example is the quality control system implemented by a leading
car manufacturer in the final vehicle inspection area. This system consists
of a small computer equipped with a voice synthesiser which prompts the
vehicle inspector for each acceptance check required on a vehicle before it
is allowed to leave the production line. The inspector is required to
confirm to the computer that each check has been carried out and whether
the item was satisfactory or not. Any unsatisfactory items are logged for
subsequent attention.
One obvious use for such a system is in the final acceptance of building
services installations or even complete buildings. At the moment many
acceptance checks tend to be carried out on a fairly cursory basis, and the
use of more structured methods might lead to both an improvement in
building quality and a reduction in the number of dissatisfied clients.
References
Atkin B. (1986) 'Project planning and control' Chartered Quantity Surveyor September 1986
pp.11-13
Barrett D. (1987) 'Head scratching stuff' Computer Weekly 24/9/87 p. 38
Brandon P.S., Basden A, Hamilton 1., and Stockley j. (1988) Expert systems: The strategic
Hunt G. (1987) 'The Tuxedo package' Chartered Quantity Surveyor june 1987 p. 13
jackson M.j. (1986) Computers in construction planning and control Allen and Unwin
Morledge R. (1987) Procurement management - Evaluation and development of choice
methodology M.Sc. Thesis, Heriot-Watt University
Ind ex
Index
Abacus unit, University of BCIS On-line 13, 95, 117, 139,

Strathclyde 137 140
abstracting and billing of data 145, benchmark tests 28
148, 149 bills of quantity (production of) 5-
accommodation (for computer 10,12, 13,95, 135,138,140,
systems) 60-61 144-171
acoustic coupler 98 Blackbox 159
activity network 142 bootstrap loader 18
Alvey programme 13, 119, 128, BRIX 95
141 budget estimates 135
Apple 11 builders' quantities 145, 171
approval ware 97 Building Industry Code 169
approximate estimating 140, 142 building procurement method 129,
approximate quantities 136-137 192, 195, 201
articulated phrase library of bulletin board 49, 93
descriptions 152, 154 byte 28
artificial intelligence 117-119
automated measurement
techniques 149, 157, 159-169
automatic measurement (from CAD Calculix 165
systems) 187 cash flow prediction 173, 176,
178, 181
Ceefax 95
Babbage 3 central processing unit (CPU) 20,
back-up copies 53 27
backing storage 20, 28, 34-37 CESMM2 151 I 170
bar charts 199 chip technology 17
bar codes 156 CI/Sfb 169
BASIC 177, 179 civil engineering measurement
basic bill production system 145, 145, 170
157 claims management 173, 180-181
basic price data (use of) 139, 140 CLASP 5, 10
batch processing 23 Clwyd C.C. 187
baud rate 99 COBOL80
206 Index
Colossus 3 cost and bonus calculation 181

coding of data 7, 146-157 cost checking (of design) 136
commercial data processing 3, 23 cost data 135, 136, 140, 178
communications 97-100 cost effectiveness (of billing
communications protocol 40, 99- systems) 148-149, 156
100 cost estimating 136, 139-140,
communications software 100 187-188
computer aided architectural cost forecasting 137, 139, 200
design (CAAD) 184 cost in use information 188
computer aided design (CAD) 12, cost modelling 12, 135, 136-137,
95, 135, 140, 142, 166, 183- 141-142
193 cost monitoring 173, 174, 178
computer aided design (benefits cost planning 135-143
and disadvantages to the Q.S.) cost prediction 135
190-193 cost profile (of construction
computer aided design/computer projects) 141
aided manufacture (CAD/CAM) cost/value reconciliation 181
184 costing of alternative design
computer aided design and concepts 137, 188
draughting (CADD) 184 Coventry City Council 5
computer aided draughting 184 Critical Path Method (CPM) 197
computer generations 18 cursor 156
computer security 55-58 cut and shuffle 144, 148, 149
computer useage (extent of) 13
computerised measurement
systems 144 daisy-wheel printer 32, 83
construction management 141, database 13, 25, 139, 140, 172,
181,196 178,200
construction planning 181 database management 90
construction time forecasting 129 database management software
Construction Industry Computing 80,88
Association 12 dataentry 146, 148
Context 95 data exchange (between computer
contract administration 173 systems) 37-38,183,188-189
contract management 138, 148 Data General DG/1 19
contract planning and control 180 data labelling 138, 169-170
contracts for supply of computer data processing 3, 23
systems 72-74 Data Protection Act (1984) 52, 70-
co-ordinated project information 6, 72
7 data transmission 37-38
copyright 66 dBase 180
cost analysis 135,138,139,141, decision support systems 115-117
148, 169, 174 DENDRAL 119
Index 207
Department of Health and Social estimating systems 135

Security 177 expert systems 13, 115-131, 141,
Department of the Environment 142
154 expert system shells 128
Derby City Counci I 5
Derbyshire C. C. 10, 156, 187
desk top computing 25, 80 facets (of description) 152-155
desk top publishing 33, 91 facsimile transmission (fax) 94
development cost appraisal 129 Federation Against Software Theft
Development Group of Chartered (FAST) 67
Quantity Surveyors 6 Ferranti Mk.1 4
digitised input 140, 149, 157 fifth generation computing 13, 119
digitisers 13, 31,149,157 file encryption systems 58
direct access 35 · final account 148, 173
direct entry (of dimensions) 156 financial budgeting 129
disaster recovery planning 54-55 financial management 1 73
disk drive 17, 20, 27, 35-37 financial statements 174
diskette 37 fixed head disk units 35
distributed processing 25 flat bed plotters 34
Doomsday project 37 Fletcher and Moore 151
drum plotters 34 floppy disk 17, 3 7
fluctuations 173, 175
formula fluctuations 175, 178
early computer systems 3 Fortran 80
economic aspects of data transfer fourth generation languages 89-90
189 fraud 56-57
EbiCON 94-95
editing (of Bills) 151
EDSAC4 Gantt charts 199, 200
electro-mechanical punched card general purpose software 80-91
equipment 3, 5 giga-byte (GB) 28, 35
electronic mail 94 graphics tablet 31
electronic measurement systems GUIDE (Graphical User
149, 157-159 Interrogation of Design
element unit quantity 138 Economics) 141
element unit rate 138
elemental cost checking 136
Elliot 8038 10 hacking 56-58
Elsie 129 hand-held data collection devices
Enviro 10, 154-155 32
ENIAC 18 hardware 17-40, 181
EPROM chip 18 hardware acquisition 50-51
208 Index
Hayes compatibility 100 LAMSAC 10

health and safety 62-63 land surveying 181
heat loss co-efficients 188 lap-top computer 22
Hertfordshire C.C. 51 10 laser printer 32-33 1 83 1 91
historical cost data 135 1 1381 139 1 leasing 51
14111421151 legal aspects of computing 65-791
Hollerith 3 1831 189
human considerations 61-63 1 84 LE04
Hurd C. 4 levels (of description) 152-155
library of descriptions 5 1 1401 144 1
150-155
IBM 4 1 23 1 201 light pen 156
IBM 370/3031 10 local area network (LAN) 22 1 27 1
IBM Selective Sequence Electronic 38-40
Calculator 4 logging of materials 181
illness among VDU operators 62- look-up tables 161-164
63 Lotus 123 180
lnfoline 95 Lyons J. & Co. 4
in house systems development 49 1
101-102
initial cost advice (from CAD MADM4
systems) 187 magnetic disk 181 35-37
initial cost estimating 129 magnetic ink character recognition
Initial Graphics Exchange Standard (MICR) 30
(IGES) 188 magnetic tape 171 181 34
ink-jet printer 33 main memory 28
input devices 20 1 29-32 mainframe computer 21 1 32 1 351
integrated software packages 90 1 1381 181
177-180 maintenance (of hardware) 51-52
intelligent terminals 27 management (of computer systems)
intelligent management 42-63
information and decision support management (of systems
systems 130 development) 112-114
interactive processing 25 management contracts 181
interim valuations 173 1 1741 1781 management information systems
181 1151 117
matrix printer 32-33 1 83
measures (of size and performance)
Kermit 100 27
keyboard 17 I 20 1 29 mega-byte (MB) 28 1 35
keyboard dynamics 58 mega-hertz (MHz) 28
keywords 157 mflops 29
kilo-byte (KB) 28 microcomputer 11 122 1 27 1 32 1 37 1
knowledge engineering 127 60-6111811 185
Index 209
mini-computer 21 1 25 1 32 1 35 1 181 post contract systems 1351 1481

Ministry of Public Buildings and 172-181
Works 154 portable computers 1721 1741 181
mips 29 prediction of future cost trends 139
modem 22 1 40 1 97-98 pressure pads 31
mouse 156 Prestel 94
moving head disk units 35 printer 171 27 1 32-33 1 83-84
MYCIN 130 Programme Evaluation and Review
Technique (PERT) 1971 200
progress reports 173 1 1741 178
NALGO 62 project co-ordination 196
National Consultative Committee project management 1351 1421
for Building and Civil 1721 1761 1811 195-202
Engineering 9 project management software 198-
NEDO formula 12 1 175 201
network analysis 199-200 PROM chip 18
Newham Borough Council 62 public domain software 96-97
North East London Polytechnic 62 Public Domain Software Interest
Northamptonshire C.C. 1 0 Group (PDSIG) 96
Nottinghamshire C.C. 5 pub Iic information systems 93-1 00
punched cards 23
punched paper tape 23
operational procedures 52-55
operations manager 23
optical character recognition
quality control 1951 202
(OCR) 30-31
quantity surveying applications 87 1
optical disk units 37
135
optical mark recognition (OMR) 30
quantity surveying expert systems
Oracle 95
128-130
out-turn costs 173 1 1 77
quantity surveying packages 49
output devices 20 1 32-34
parameters 163 re-use of data by contractors 6

Pascal 3 Reading University 12
password protection 57-58 relational data base 88
patents 66 resource levelling 200
payment notifications 1741 178 resource requirement reports 200
peripherals 17 RICS 81 1281 1291 1391 141 1 1501
PET 11 151
PHIBS (Phraseology Independent RICS Building Cost Information
Billing System) 156-157 Service (BCIS) 131 95 1 1171 1391
plotters 33-34 1401169
210 Index
RICS Computer Applications systems analyst 44-46, 103

Steering Committee 8
RICS Standard Form of Cost
Analysis 1 38 Techsonix 31, 157
Ripac 148 tender preparation 140
robot 118 Textline 95
ROM chip 18 Telecom Gold 94
rogue items 155 tender price indexing 139
tender prices 139, 148
tick-sheet valuations 1 74
'S' curves 177 time slicing 25
Salford University 13 three-dimensional spreadsheet 141
screen 17, 20, 30 TP-DOS Online 96
serial access 34 trade secrets 66, 68
shareware 97 training 59-60, 73
simulation of construction 142 transaction processing system 1 1 7
site based computers 174, 181 Trent Polytechnic 96
site measurement 181
size of computer systems 27
skeleton library of descriptions 152 UNIVAC 1 4
SMM7 151,152 unit quantities 149, 159-164, 190
SMM7 Library of Standard University of London 139
Descriptions 151, 154-155, 161
software 17, 47-50
software development 48-50, 101- valuations 148, 173, 178
114 variations 174, 1 78
software errors 77-78 video display uni-t/terminal (VDU)
software piracy 67 23,29-30,62,155-158,179
software purchase 48, 75-78
sorting (of data) 149-150
spreadsheet 12, 49, 80, 84-87, War Office, Quantity Surveying
117,141,166-169,174-175, Division 5
179, 18~ 199,200,201 warranties 74
squaring of data 148 Watson T. Snr. 4
standard library of descriptions 5, West Sussex C.C. 185, 187
140, 144, 150-155 Whetstone test 29
standard price schedule 139 'what-if' simulation 87, 137, 141,
Stanford Research Institute 1 19 181
statistical methods (of cost Winchester technology 35
planning) 137, 139,142 word processing 49, 80-84, 174
Supercalc 2 174 WORM technology 37
Sweet, Cyri I & Partners 152
system Iife cycle 104-1 12
systems 103 XMODEM 100

Computers and Quantity Surveyors (PDFDrive) PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Computers and Quantity Surveyors (PDFDrive) PDF

Uploaded by

Copyright:

Available Formats

Computers and Quantity Surveyors

Macmillan Building and Surveying Series

Series Standing Order

Customer Services Department, Macmillan Distribution Ltd

All rl.ghts reserved. No reproduction, copy or transmission

No paragraph of this publication may be reproduced, copied

Any person who does a.ny unauthorised a.ct in relation to

First published 1989

British Library Cataloguing in Publication Data

ISBN 978-0-333-49161-4 ISBN 978-1-349-10857-2 (eBook)

Cover photograph of a dlgitiser kindly supplled by Techsonlx (U.K.) Ltd.

2 COMPUTERS APPLIED TO QUANTITY SURVEYING 133

I am indebted to Brian George, Senior Lecturer in the Department of

At the end of this chapter you should be able to:

Early developments in computing and data processing

The first consideration when discussing the development of early computer

The quantity surveyor and data processing

The origins ofthe quantity surveyor's love-hate relationship with computing

There are unlikely to be any insuperable problems to the use of

Standard descriptions are inevitable ... Standardisation of descrip-

Quantity surveyors find that the above factors cause working up to

1960-1980- Triumph or disaster?

Following the publication of the RICS working party reports in 1961, a

Extracted from RICS (1967b):appendix C

Revolution! The introduction of the microcomputer

Although the principles underlying the development of the modern micro-

Quantity surveyors, as any word processor salesman will tell you in an

profession actually using computers effectively and my current guestimate

Alvey R.j. (1976) Computers in quantity surveying Macmillan

Daniel P. (1979) 'Bills of quantity in a multi-disciplined design system' Chartered Quantity

RICS (1967b) Computer Techniques RICS

At the end of this chapter you should be able to:

Definition of some basic terms

Computer hardware is the term used to describe the actual computer

Brandon and Moore (1983:p5) classify modern microcomputers as sixth

Table 2. 1 Computer generations

Generation Technology Comments

1st (1940-52) Vacuum tubes (valves) large, expensive,

Table 2.2 Comparison of EN lAC with DG/1 lap-top portable

Features EN lAC Data General

No. of components > 100,000 valves, <100 chips

Functional components of a computer system

Having defined in very broad terms what is meant by computer hardware,

CENTRAL PROCESSING UNIT

HARO COPY OUTPUT

Figure 2. 1 Functional components of a simple computer system

Types of computer system

The range of computer systems currently available may be broken down

Minicomputers will usually perform rather more slowly than mainframe

Microcomputers are, in physical terms at least, the smallest end of the

Methods of processing data

Figure 2.2 Batch processing

Batch processing is clearly suited to extensive data processing operations

In an attempt to overcome the difficulties posed by the above systems, and

Figure 2.3 Interactive processing

microcomputers used as intelligent terminals to the central machine. Such

USER < USER 3

Some commonly quoted measures of size and performance

Microcomputer manufacturers usually describe the 'size' of a system by

example of a set developed to compare the relative performance of micro-

Video display terminals

Screen displays are available in a range of colours, green characters on a

Optical mark recognition (OMR), magnetic ink character recognition