sustainability

Article
Cost Calculation of Construction Projects Including
Sustainability Factors Using the Case Based
Reasoning (CBR) Method
Agnieszka Leśniak ID
and Krzysztof Zima * ID

Institute of Management in Construction, Faculty of Civil Engineering, Cracow University of Technology,

31-155 Krakow, Poland; [email protected]
* Correspondence: [email protected]; Tel.: +48-628-23-54




Received: 23 April 2018; Accepted: 11 May 2018; Published: 17 May 2018


Abstract: The idea of sustainable development and the resulting environmentally friendly attitudes
are increasingly used in construction projects. Designing in accordance with the principles of
sustainable development has an impact on the costs of construction works. The authors of this
paper proposed an approach to estimate the costs of sports field construction using the Case Based
Reasoning method. In their analysis, they distinguished 16 factors that affect the cost of a construction
project and are possible to already be described at an early stage of its preparation. The original
elements of the work include: consideration of such environmental factors as the environmental
impact of the building, materials used, the impact of the facility on the surroundings affecting the
amount of implementation costs and development of own database containing 143 construction
projects that are related to sports fields. In order to calculate the similarity of cases, different
calculation formulas were applied depending on the type of data (quantitative, qualitative, uncertain,
no data). The obtained results confirmed that the CBR method based on historical data and using
criteria related to sustainable development may be useful in cost estimation in the initial phase
of a construction project. Its application to the calculation of the costs that are related to the
implementation of sports fields generates an error of 14%, which is a very good result for initial
calculations. In the short run, such factors as the impact of the object and the type of materials that
are used from the perspective of their influence on the environment may be decisive as far as the
costs determined in the life cycle of the building are concerned, as well as the lowest costs of the
building construction ensuring the appropriate quality and respect for the environment.

Keywords: Case Based Reasoning; construction; cost estimation; sports field; sustainability

1. Introduction
The implementation processes of sustainable development in the area of construction are of great
importance. Construction processes have an important role in creation of built environment and their
impacts have to be measured as construction contributes to air pollution, land use and contamination,
usage of resources, water and materials depletion, water pollution, impacts on human health, and
climate change [1]. The erection of a construction object is usually associated with the development
of the biologically active surface, and subsequently, with putting a burden on the environment.
The severity of these kinds of impact on the environment varies depending on the stage of the life cycle
of the building. The analysis of the object vs. natural environment relation allows for distinguishing
four basic stages of impact that are linked to the following processes: extraction of raw materials,
production of materials, construction of the facility, operation of the facility, and its demolition [2].
As [3] proved, the results of developing sustainable architecture are based on changing the function of a

Sustainability 2018, 10, 1608; doi:10.3390/su10051608 www.mdpi.com/journal/sustainability

Sustainability 2018, 10, 1608 2 of 14

building from a linear approach to a closed circulation plan. The linear approach treats the building as
the “place of processing natural resources into waste” (for instance, water is transformed into sewage,
energy into heat losses, building materials into waste). The other approach, namely in the closed
circulation plan, a building can change from a consumer of energy and other resources into a virtually
self-sufficient unit (through energy recovery or the re-use of water or waste). Designed buildings are
characterized by diversified energy demand, which depends on many factors; for example, the material
and construction solutions applied, the type of object, heating system, and its efficiency. According
to the Energy Performance of Buildings Directive [4], all new buildings must be nearly zero-energy
buildings by the end of 2020 and all new public buildings by 2018. In general, in terms of sustainable
development and sustainable construction, refurbishment of buildings is preferred to new construction,
because this helps to save energy and building materials in construction phase, also reduces generation
of waste and other emissions [5]. A building object should be designed, constructed, operated,
and demolished in accordance with the requirements of sustainable development. The construction
materials used to build objects greatly influence the impact of the building on the natural environment,
which was noticed in many works, as in [6–10]. The choice of building materials with the appropriate
sustainability criteria is not straightforward. What is relatively easy to determine are such objective
factors as cost constraints and design considerations, yet other, often subjective, factors have an impact
on the selection, which influences the achievement of sustainability goals. In [11], one can find an
optimization model for sustainable materials selection, while in [12], a model using a multi-criteria
analysis enabling the selection of sustainable materials. Researchers [13] noticed that due to possible
harmful effects of construction equipment on the environment, their evaluation with sustainability
considerations can be considered as a helpful activity to move toward the sustainability in construction.
The paper [14] presents a review of the literature on the sustainable built environment, which was
made on the basis of the articles that were published between 1998 and 2015. The authors believe that
the welfare of the whole society depends on the sustainability of the built environment [14].

2. Literature Review
The construction objects that are built today should show adequate durability, affect the
environment in a harmless way, be economical in the consumption of materials and energy, and
take into account the consequences of failure from the point of view of human life and health [15].
Choosing the right construction and material solutions, as well as the technologies that are applied
in the first stage of the investment process influences the costs of its implementation and subsequent
operation. All of the construction projects are risky, and different techniques and tools are proposed
for assessment of risk in Construction project [16,17]. In [18], authors noticed that the fluctuation of
material prices is one of the risk factor leading to cost overruns problems. In construction, reliable
estimation of costs is important for both the investor who finances the investment and the contractor
who has to estimate the costs and achieve a satisfactory level of profit.
To determine the price of construction works, the direct costs that are connected with realization
of the works, overhead costs and profit must be consider. The contractors can calculate costs of
works using unit price. This method requires a lot of experience from the contractor. The unit price
must include all of the mentioned cost’s elements. Traditional method of cost estimating is based on
calculation of costs elements of separately: the direct cost (labour, materials, and equipment), indirect
cost, and profit [19]. This traditional approach is accurate but time consuming, and therefore new
methods are still being sought by means of new mathematical tools that can support the effectiveness
of the calculation. Studies that are worth considering include ones that are employing artificial
neural networks [20–22], linear regression [23], fuzzy sets [19], and support vector machines [24].
Researchers [25] proposed a hybrid model where multivariate regression method and the artificial
neural network (ANN) method have been combined to provide a cost estimate model.
One of the methods that is proposed in the literature, which can be used in estimating costs in
construction, is CBR (Case Based Reasoning) [26]. Acquiring knowledge as a result of researching the
Sustainability 2018, 10, 1608 3 of 14

correctness of data, assimilating or formulating new concepts that are based on examples from the
past, (CBR—Case Based Reasoning) in contrast to relying on individual experiences, can accelerate the
process of estimating costs in construction. CBR can be defined as systems that solve new problems by
adapting the results that were used to solve old issues [27]. In [28], the authors compare the system of
inference from cases to the black box. The input data describe the problem; the output data form a
solution to the problem, while the memory of past cases and the box contain a reasoning mechanism.
In the case-based inferencing, the basic source of knowledge is a database that is containing not rules
but a set of cases from the problems encountered and resolved. New problems are solved by searching
for the most similar cases and their possible adaptation. The CBR models used for cost estimates
may be based on both quantitative and qualitative data [29]. In [30,31], it was noticed that, although
some information is not specified, CBR models for long periods of use maintain quality and ability to
solve problems and work better than other models. The use of CBR to estimate construction costs is
based on searching for similar investments already completed. This provides a simple way to measure
construction costs, given that, according to most studies, there are non-linear relationships between
cost and factors that affect it [32–35]. An interesting example of the use of CBR in the cost estimation
process can be a model using the AHP method to determine the weights of criteria, proposed by [26]
or the CBR model using genetic algorithms to estimate the construction costs [36] or unit cost of
residential construction projects [37]. Models are also being created that predict both construction time
and cost at an early stage of a construction project [38]. Ryu et al. [39] proposed the CONPLA-CBR
tool that generates master schedules at the preconstruction stage.
The paper presents the concept of supporting the estimation of construction costs in the initial
investment phase based on Case Based Reasoning (the CBR method). The process of information
management and the use of historical data differ from the models that have been proposed so far.
Typically, global assessment of construction projects or facilities based on a few or several criteria
describing the construction project in a general way, such as the usable area, cubic volume of the
building, height, complexity, or location. Such criteria often do not take into account differences in
the type of materials used and the details of solutions or environmental impact. In cost calculations,
fuzzy logic is rarely used when there is uncertain or imprecise information for the problem of cost
calculation for the implementation of construction works.
In the proposed algorithm, among the proposed 16 explanatory variables, it is suggested to
take into account the features related to the environmental impact of the building, its influence on
the surroundings and the parameters of construction materials in relation to the idea of sustainable
development. Four different calculation formulas are also proposed for quantitative and qualitative
data, for cases without data or ones with uncertain or inaccurate data. It is worth noting that there are
no proposals for methods and tools supporting the cost estimation of such sports facilities as football
pitches, treadmills, or skate parks.

3. Cost Calculations Method Based on Case Based Reasoning with Sustainability Criterion
The database CSDB (Cost Solution DataBase) that was developed to support index cost estimation
with the CBR method includes cases containing four groups of information: information about
geometry of building object—GEO, solution of problem (unit cost of object)—SoP, description of
the construction—DoC, data for adaptation process—Ad. The database CSDB was defined by the
following formula:
CSDB = Uni=1 Casei {GEOi , SoPi , DoCi , Ad} (1)

where:

Casei —i-th case from the database,

GEOi —graphical representation of the construction elements for the i-th case,
SoPi —solution of problem,
DoCi —description of construction,
Sustainability 2018, 10, 1608 4 of 14

Ad—data for adaptation process such date, localization of construction, and

n—number of old cases in the database.
Sustainability 2018, 10, x FOR PEER REVIEW 4 of 14
GEO is information about the graphic appearance of the construction object or element in the
form of quantitative data. For
GEO is information example,
about it canappearance
the graphic be informationof the about the volume
construction object of
or an object,
element in surface,
the
or dimensions, such asdata.
form of quantitative length, width, and
For example, it canheight. Solutionabout
be information of problem
the volume SoPofcontains
an object,information
surface,
about a unit price such
or dimensions, of construction
as length, width,object andreferenced to, forofexample,
height. Solution problem SoP surface areainformation
contains or volumeabout of object.
Description of construction
a unit price of construction DoC contains
object referencedall the necessary
to, for example, information
surface areaabout factors
or volume of (including
object.
Descriptionfactors)
sustainability of construction
influencing DoCtocontains
the cost all the project
of the necessary information
in the about factors
form of qualitative (including
data.
sustainability
The presented factors)
CSBD influencing
databasetowas the created
cost of the inproject
order into the form of
support qualitative
the process of data.
estimating costs
Thephase
in the early presented CSBD
of the database was
investment, createdquantitative
including in order to support the process
variables (defining of estimating
the bill of costs in
quantities)
the early phase of the investment, including quantitative variables (defining
and qualitative variables. The novel element of supporting the process of cost estimation is one that is the bill of quantities)
and qualitative variables. The novel element of supporting the process of cost estimation is one that
involving the variables affecting sustainable development. The authors of the paper note the increasing
is involving the variables affecting sustainable development. The authors of the paper note the
influence of factors that are related to sustainable development and their impact on the investment
increasing influence of factors that are related to sustainable development and their impact on the
price. The assessment of sustainable development consists of three basic elements: impact on the
investment price. The assessment of sustainable development consists of three basic elements: impact
environment (consumption
on the environment of natural of
(consumption resources, consumption
natural resources, of energyofresources,
consumption and emissions
energy resources, and of
pollutants
emissionsinto
of the environment),
pollutants comfort, and
into the environment), qualityand
comfort, of quality
life, as of
well
life,as
as cost (the
well as costamount of funds
(the amount
allocated to achieve the expected value in use and reduction of
of funds allocated to achieve the expected value in use and reduction of the impact on thethe impact on the environment).
In terms of costs,Inthis
environment). termsmeans constructing
of costs, a building at
this means constructing the lowest
a building possible
at the cost while
lowest possible cost obtaining
while
obtaining
adequate adequate
quality quality and
and respect respect
for the for the environment,
environment, as well asas well
the low ascosts
the low of costs of its operation.
its operation.
Figure
Figure 1 presents
1 presents the algorithm
the algorithm for theforcost
the estimation
cost estimation
method method
basedbasedon theon the Based
Case Case reasoning
Based
reasoning method, taking into account
method, taking into account the above assumptions. the above assumptions.

Figure1.1.Algorithm
Figure Algorithm used in
in the
themethod
methodproposed.
proposed.

The first step in the method that is proposed is to define a preliminary set of variables, divided
into GEO geometry variables, and variables describing the DoC structure. Using the case database
Sustainability 2018, 10, 1608 5 of 14

The first step in the method that is proposed is to define a preliminary set of variables, divided
into GEO geometry variables, and variables describing the DoC structure. Using the case database
collected, one can perform a correlation analysis in step 2 to eliminate the variables that have no impact
or have a small impact on the unit price of construction works. The result of the correlation analysis
is to create the final set of variables (Final set). The third step of the analysis is the assessment of the
validity of variables, which can be performed by means of expert assessment; alternatively, it may
result from the determined strength of correlation between the individual variables and the price.
The fourth step of the method begins the proper analysis based on the inference from cases.
The new case providing a problem to analyse is compared in pairs with all the old cases from the
database. The purpose of the comparison is to find the most similar cases, and, by analogy, to determine
the unit price of the construction works for the new case.
The similarity of cases can be calculated using different formulas depending on the type of
explanatory variables. The calculation formulas applied in the algorithm are presented below:
• For quantitative explanatory variables (as per [40]):
 

 wN − wj 
sim wN , wj =1 − (2)
wmax − wmin

where:

wN —value of the explanatory variable for the new case,

wj —value of the explanatory variable for the j-th old case, and
wmax , wmin —minimum and maximum values for all the old cases included in the database.

• For qualitative explanatory variables:

|n(w N) − n(w j )|
sim wN , wj = 1− (3)
M − 1

where:

n(wN ), n(wj )—place in an ordered array of values n(w) = 1, 2, . . . , n,

M—number of values.

• For uncertain or inaccurate variables:

∑4i=1 |ai − bi |
sim(ANC , BCi ) = 1 − (4)
4
where:

sim(ANC , BCi )—similarity between fuzzy numbers,

ANC —fuzzy number for the new case,
BCi —fuzzy number for the old case taken from the database, and
ai , bi —characteristic points for fuzzy numbers ANC = (a1 , a2 , a3 , a4 ) and BCi = (b1 , b2 , b3 , b4 ).

The formula for uncertain or inaccurate variables is given as an example and concerns a situation
in which the shape of the membership function is trapezoidal. It is defined for a situation in which
two trapezoidal membership functions assume values ANC = (a1 , a2 , a3 , a4 ) and BCi = (b1 , b2 , b3 , b4 ).
The characteristic points allow for describing the limit values for the shape of the membership function
that accepts the values 0 and 1. This allows for describing the fuzzy number using the four real
numbers, which allows for the quick execution of actions using only these characteristic point values.
When adopting other forms of membership functions, other calculation formulas should be used.
Sustainability 2018, 10, 1608 6 of 14

• In the absence of data for an explanatory variable for a new case, an old case or both:

sim(wN , wS ) = 0 (5)

where:

wN —an explanatory variable for the new case and

wS —an explanatory variable for the old case.

The formulas given above are applicable when comparing local similarity, defined as a pair
comparison of similarities for subsequent explanatory variables.
The next, fifth calculation step is to calculate the global similarity, which is the weighted sum of
the local similarities of all the explanatory variables that are used in analysis and collected in the final
set of variables. In order to calculate global similarity, the following formula was used:
n
∑ ωi (sim i (V Ni , VSji ))


SIM VN , VSj = (6)
i=1

where:

ωi —weight of the i-th explanatory variable,

SIM(VN ,VSj )—global similarity between the old Vj and the new case VN , and
simi (VNi ,VSji )—local similarity for the i-th explanatory variable between the old Vj and the new case VN .

It is necessary to determine a few basic assumptions in the course of the analysis that uses the
CBR method:

1. old cases with the highest global similarity SIM(VN , Vj ) are selected. The minimum number of
cases entering the selected set of cases is 3. This assumption is to limit the possibility of choosing
an accidental solution—if three solutions are chosen, then the possible extreme solution is rejected
to limit the possibility of overestimating or underestimating the price of works;
2. the minimum value of global similarity for cases included in the set of solutions must be greater
than 70%;
3. the resulting value of similarities is given as a percentage and is a natural number; and,
4. cases from the set of selected cases are rejected as extreme when the difference between the
selected cases is greater than 50%.

The next step is to adapt the calculated unit price for the new case. The value correction concerns
a situation in which there is a difference between the duration of the construction works and the
location between the new case and the selected old cases. To do the adaptation, two coefficients were
applied: regional coefficient determined in the case of a difference in the location of a construction
project, (introduces price correction due to the difference in the local prices of services and materials)
and inflation rate (introduces a price correction due to changes in prices of services and materials over
time). Corrections are made individually for each of the unit prices that were selected in the course of
the analysis, resulting from the selection of the old case as the most similar and meeting the conditions
that are given in the assumptions.
The unit price of the new case, resulting from several finally selected old cases after adaptation,
is calculated as the weighted average of the weights of individual old cases (where the basis for the
calculation of the weight is the degree of the similarity to the new case) and the unit prices of old
cases. The proposed unit price is accepted as the final price of the new case and in the next step it is
multiplied by the amount of works from the bill of quantities. After the settlement of the construction
works, the price result obtained is verified by the actual costs. The verified result is saved in the
database for its subsequent use.
Sustainability 2018, 10, 1608 7 of 14

4. An Example of Supporting Cost Calculation with the Use of the CBR Method, Taking into
Account the Factors of Sustainable Construction
The calculation example concerns the estimation of the costs of the implementation of sports
fields in the early phase of the investment, and is thus based on the investment concept. The cost
analysis is performed in accordance with the algorithm provided above, which is based on the CBR
method. The purpose of the exemplary cost analysis is to determine the price for the implementation
of a new sports facility, which is a sports field that is based on historical data from the investments
already completed.
In the first step, a preliminary set of criteria was compiled. The set of criteria was created on
the basis of literature studies and the analysis of advertisements that were placed in Poland on the
public procurement website. The analysis included the selection of information describing the subject
of the contract.
On the basis of literature studies [22,41–43] and the analysis of announcements in public
procurement, 16 variables explaining sports fields were distinguished:

1. Quantitative variables (type of information—GEO)

intended use of the field (five types of fields)
surface area of the field (variables range: 275–8714 m2 )
surface area of the access paths and routes (variables range: 0–1753 m2 )
green surface area (variables range: 0–6017 m2 )
surface area of the ball containment netting (variables range: 0–2212 m2 )
fence length (variables range: 0–602.5 m)
2. Qualitative variables (type of information—DoC)
type of the material for sports surface (six types of materials)
type of the material for access routes (five types of materials)
type of the fence (five types of fences)
type of sports equipment—handball (yes or no)
type of sports equipment—volleyball (yes or no)
type of sports equipment—basketball (yes or no)
type of sports equipment—football (yes or no)
type of sports equipment—tennis (yes or no)
impact of the construction on the environment (rating 1–5)
impact on the surroundings (rating 1–5)

The presented example includes the analysis of the costs of sports fields, both single- and
multifunctional. The information contained in the database comes from the investments executed in
2014–2016 and includes 143 building projects concerning the construction of sports fields. The data
comes from advertisements and building cost estimates prepared for tender proceedings for the
implementation of sports fields in Poland. The data from cost estimates and the public procurement
description were placed in the database that was created. The data includes all of the variables
highlighted in the course of the analysis of the initial set of criteria: GEO and DoC, and the information
about the SoP solution and data allowing for the later adaptation of the Ad solution (tender date and
location of a construction project). The method of recording information in subsequent records in the
database is described below.
Casen {GEO = (intended use of the field, surface area of the field, surface area of the access paths
and routes, green surface area, surface area of the ball containment netting, fence length); DoC = (type
of the material for sports surface, type of the material for access paths, type of the fence, type of sports
equipment—handball, type of sports equipment—volleyball, type of sports equipment—basketball,
type of sports equipment—football, type of sports equipment—tennis, impact of the construction
on the environment, impact on the surroundings); SoP = (unit price of the field surface area); and,
Ad = (location, date of the bid)}
Sustainability 2018, 10, 1608 8 of 14

The GEO information contains basic parameters concerning the size of individual construction
works. On the other hand, the DoC information contain a description of the scope of works and
the types of solutions used in the implementation of individual elements of the sports field and its
surroundings, and the characteristics of the undertaking related to the idea of sustainable development.
The features that are related to the impact of a building object on the environment (sustainability
factors) included in the initial set of variables involve in point 2. Qualitative variables (type of
information—DoC) such variables:

• impact of the construction on the environment (for instance, energy demand, use of renewable
energy, efficiency of energy systems);
• materials (for instance, materials with low environmental impact, materials with low risk of health
hazard, recyclable materials): type of the material for sports surface and type of the material for
access routes; and,
• impact on the surroundings (such as air pollution, noise, vibration, wind effects and shading of
the area, the effect of the thermal island)

The determination of the correlation strength was made on the basis of the Guilford scale:

r = 0 no correlation,
0 < r < 0.1 barely perceptible correlation,
0.1 < r < 0.3 poor correlation,
0.3 < r < 0.5 average correlation,
0.5 < r < 0.7 high correlation,
0.7 < r < 0.9 very high correlation,
0.9 < r < 1 almost full correlation, and
r = 1 full correlation.

Table 1 depicts the results of the Pearson and Spearman correlation analysis. For quantitative
variables, the analysis involved Pearson’s correlation; while for quality variables, it used Spearman’s
correlation. In order to create the final set of explanatory variables, variables that showed no correlation
or very low correlation were rejected in the course of the analysis. The final set of criteria with
designated weights is shown in Table 2. The weights were determined based on the correlation
strength of individual variables with the unit price of a construction work.

Table 1. The correlation coefficients of explanatory variables from initial set.

Variables Correlation Coefficient:

Surface area of the fields −0.192
Surface area of the access paths 0.232
Green surface area 0.079
Fence length 0.249
Surface area of the ball containment netting 0.056
Intended use −0.472
Material for sport surface −0.299
Material for access paths −0.283
Type of sports equipment—Handball 0.244
Type of sports equipment—Basketball 0.477
Type of sports equipment—Volleyball 0.363
Type of sports equipment—Football −0.289
Type of sports equipment—Tennis −0.071
Fence type −0.045
Impact of the construction on the environment 0.095
Impact on the surroundings −0.640
Sustainability 2018, 10, 1608 9 of 14

Table 2. The final set of explanatory variables with their weights.

Variables Weights ωi
Surface area of the fields 6.2%
Surface area of the access paths 7.5%
Fence length 8.0%
Intended use 15.2%
Material for sport surface 9.7%
Material for access paths 9.1%
Type of sports equipment—Handball 7.9%
Type of sports equipment—Basketball 15.4%
Type of sports equipment—Volleyball 11.7%
Type of sports equipment—Football 9.3%

The final set of variables consists of two explanatory variables that are related to the idea of
sustainable development and included in the Materials (materials for sports surfaces and materials for
access). The total validity of these variables is 18.7%, thus they are quite a significant element affecting
the unit price of a construction project, although not being decisive. Due to the low correlation with
the unit price, the other variables that are related to sustainable construction were rejected, that is, the
impact of the building on the environment and the impact on the surroundings, which is two of the
four factors that were originally considered.
After creating the final set of variables and determining their validity, local similarities were
calculated for the subsequent explanatory variables between the New Case and all the Old Cases
from the database. Local Formulas (2)–(5) were used to calculate local similarities. An example of
calculating a local similarity for a quantitative variable of an access surface, between the New Case
and the Old Case 7 using the Formula (2) has been shown below:

|458 − 295|
= 1 − 0.029 = 0.971 ∼


sim wN , wj = 1 − = 97% (7)
5569 − 0

where:

wN equals 458 and determines the size of the access surface for the New Case—458 m2 ,
wj equals 295, which means the access surface for Old Case 7 equal 295 m2 ,
wmax equals 5569, which means the maximum access area for the whole set of cases is equal 5569 m2 , and
wmin is equal 0, which means the minimum access space for the whole set of cases is equal 0 m2 (some
orders for the implementation of sports fields did not include the implementation of access routes).

On the other hand, the local similarity for the quality variable: material for sports surfaces,
between the New Case and the Old Case 7 using the formula (3) is shown below:

|5 − 5|
sim wN , wj = 1 − = 1 − 0 = 1 = 100% (8)
6 − 1

where:

n(wN ) is equal 5 and determines the value given to the material which is the polyurethane from which
the New Case surface is planned to be made,
n(wj ) is equal 5, which means the value given to the material which is the polyurethane from which
the Old Case 7 surface was made, and
M = 6, which results from the specification of six types of materials used for sports surfaces found in
the database (natural grass; surface from natural dried wood chips—technologically softened along
fibers with 5–50 mm fraction; artificial grass; brick flour; polyurethane surface; asphalt surface).
Sustainability 2018, 10, 1608 10 of 14

A list of calculated local similarities and global similarity computed in accordance with the
formula (6) taking into account the weights given in Table 2 between the New Case and the selected
Old Cases is shown in Table 3.

Table 3. The values of the local similarities and the global similarities for the chosen cases.

Local Similarities
Variables
Case 7 Case 8 Case 9
Surface area of the fields 94% 88% 94%
Surface area of the access paths 97% 87% 92%
Fence length 86% 79% 79%
Intended use 86% 0% 100%
Material for sport surface 100% 0% 100%
Material for access paths 100% 100% 0%
Type of sports equipment—Handball 100% 100% 100%
Type of sports equipment—Basketball 100% 0% 100%
Type of sports equipment—Volleyball 100% 0% 100%
Type of sports equipment—Football 100% 100% 100%
Global similarities 98% 45% 88%

The highest similarity of 98% occurred for three cases: Case 7—SIM(VTest case 1 , VCase 7 ); Case
23—SIM(VTest case 1 , VCase 23 ); and, Case 83—SIM(VTest case 1 , VCase 83 ).
For the selected cases in the analysis of three cases with the highest similarity rate, the adaptation
of unit prices of these old cases was made due to the time difference in the calculation and the difference
in the prices of services and materials in different regions of Poland. The inflation factor considering
the time difference was calculated on the basis of the price index “The Sekocenbud forecasting and
indexation bulletin” [44]. The regional coefficient, in turn, was calculated based on the newsletter
“The Sekocenbud regional price bulletin” [45]. The adjusted unit prices are shown in Table 4.

Table 4. Adjusted unit prices of selected cases.

Case Case 7 Case 23 Case 83

Unit price 80.25 € 84.23 € 79.34 €
Regional factor 1.028 0.960 1.028
Indexation factor 100.1% 101.3% 101.3%
Adjusted price 82.58 € 81.91 € 82.62 €

For example, the price after the adaptation of Case 7 has been calculated as 80.5 € × 1.028 × 1.001,
where the regional factor is—1.028 and the indexation factor is equal to (1 + 0.01%) = 100.1%. The final
price for test case 1 is an arithmetic mean of the unit prices of case 7, 23, 83, and equals:

(C Case
jA
7
+ CCase
jA
23 Case 83
CjA (82.58 + 81.91 + 82.62)
= = 82.37 €/m2 (9)
3 3

After calculating the unit price after adaptation for the New Case equal 82.37 €/m2 it is necessary
to specify the value of the works. For this purpose, the unit price was multiplied by the number of the
works representing the construction works consisting in the performance of the sports field, namely
the area of the pitch.
The value of the works for the New Case is thus finally: 82.37 [€/m2 ] × 1200 [m2 ] = 98 844.00 €.

5. Discussion
An analysis of the error that was generated by the forecasts for the costs of sports field construction
using the presented method was made. Testing was performed on selected 15 cases for which the costs
Sustainability 2018, 10, 1608 11 of 14

of building a construction project were known. Mean Absolute Estimate Error (MAEE) for 10 test cases
was 14%. The error range for individual cases is 2–34%, except that only for three test cases it exceeded
20%. In five cases, the error was smaller than 20%. Figure 2 shows the distribution of the MAEE error
in the following ranges: up to 10%, 10–20%, and above 20%.
Sustainability 2018, 10, x FOR PEER REVIEW 11 of 14

Figure 2. The distribution of the Mean Absolute Estimate Error (MAEE) error with the cost calculation
Figure 2. The distribution of the Mean Absolute Estimate Error (MAEE) error with the cost calculation
broken down into percentage ranges.
broken down into percentage ranges.

In accordance with the PMI 2008 [46] guidelines, the error of 14% meets the requirements of an
In accordance
acceptable with theerror
cost calculation PMImade
2008 in
[46]
theguidelines,
early stagestheoferror of 14% meets
construction theinrequirements
projects of an
the range of −30%
+50%. cost calculation error made in the early stages of construction projects in the range of −30%
acceptable
to
to +50%.
Also, the requirements of the American Association of Cost Engineers (AACE)—error size of
−10%Also, the requirements
to +15%—and of the American
the requirements Association of
of the Construction Cost Engineers
Industry (AACE)—error
Institute—error size of −30%size to
of
−10%(for:
+50% to +15%—and
[47]) are met. the requirements of the Construction Industry Institute—error size of −30% to
+50%It(for: [47]) are met.
is worth noting that, in similar studies, only the use of artificial neural networks in supporting
costsItinissuch
worth annoting that, in similar
early investment studies,
phase givesonly the use
similar good of results
artificial neuralHowever,
[48,49]. networks evenin supporting
in these
cases, the maximum calculation error exceeds 50%, and in the presented method it was 34%. in these
costs in such an early investment phase gives similar good results [48,49]. However, even
cases,It the maximum
seems calculation
that both the use error exceeds
of the 50%, and
CBR method ininsupported
the presented cost method it was
calculation and 34%.
the use of
It seems that both the use of the CBR method in supported cost calculation
variables based on the idea of sustainable development can bring good results. According and the use of variables
to the
based on the idea of sustainable development can bring good results. According
authors, along with the development of the idea of sustainable development of such variables to the authors, along
affecting price, there will be increasingly more research on the impact of these factors on costs. there
with the development of the idea of sustainable development of such variables affecting price,
will be
Forincreasingly
the CBR method, more research on the impact
further research of thesethe
must address factors
impact on costs.
of changing calculation formulas
For the CBR method, further research must address the
that estimate local and global similarities on the MAEE error. The search impact of changing calculation
for formulas formulas
reflecting the
that estimate local and global similarities on the MAEE error. The search for formulas
specificity of data used during calculations may bring even better results. In-depth research into the reflecting the
specificity of
adaptation data used
process duringsolutions
of selected calculations mayalso
should bring
beeven better results. In-depth research into the
performed.
adaptation process of selected solutions should also be performed.
6. Conclusions
6. Conclusions
Supporting cost calculations at an early stage of a construction project is a vital problem. Despite
Supporting cost calculations at an early stage of a construction project is a vital problem.
the fact that cost-support models that are based on numerous mathematical methods have already
Despite the fact that cost-support models that are based on numerous mathematical methods have
been presented in the world literature many times, only a few generate a calculation error below 20%.
already been presented in the world literature many times, only a few generate a calculation error
It seems that only these models that are based on historical data, artificial neural networks and case
below 20%. It seems that only these models that are based on historical data, artificial neural networks
based reasoning methods reach an acceptable level of error. The vast majority of methods do not take
and case based reasoning methods reach an acceptable level of error. The vast majority of methods do
into account the variables that are based on the idea of sustainable development.
not take into account the variables that are based on the idea of sustainable development.
The authors believe that the proposed method using Case Based Reasoning has the potential
The authors believe that the proposed method using Case Based Reasoning has the potential and
and can be useful for practice. It generated an Mean Absolute Estimate Error of 14%. It should be
can be useful for practice. It generated an Mean Absolute Estimate Error of 14%. It should be noticed
noticed that only for 3 test cases the MAE Error exceeded 20%. To present performing of the model
that only for 3 test cases the MAE Error exceeded 20%. To present performing of the model an example
an example of calculation the unit price for the New Case was successfully conducted. The value of
the works for the New Case finally equalled 98 844.00 €.
The original element of the of cost estimation process using the proposed model is involving the
factors affecting sustainable development. The authors note the increasing influence of factors that
are related to sustainable development on the investment price. Environmental impact, comfort, and
quality of life, in the near future can be decisive in terms of costs determined in the lifecycle of a
Sustainability 2018, 10, 1608 12 of 14

of calculation the unit price for the New Case was successfully conducted. The value of the works for
the New Case finally equalled 98 844.00 €.
The original element of the of cost estimation process using the proposed model is involving
the factors affecting sustainable development. The authors note the increasing influence of factors
that are related to sustainable development on the investment price. Environmental impact, comfort,
and quality of life, in the near future can be decisive in terms of costs determined in the lifecycle of a
building, taking into account the construction of a building object as the cheapest cost when obtaining
adequate quality and respect for the environment, while obtaining low operating costs.

Author Contributions: The individual contribution and responsibilities of the authors were as follows: K.Z.
designed the research main idea and collected the data. Authors together analyzed the data and the obtained
results. A.L. provided extensive advice throughout the study results and methodology. K.Z. and A.L. wrote the
paper. All the authors have read and approved the final manuscript.
Funding: This research received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References and Notes

1. Zavadskas, E.K.; Vilutienė, T.; Tamošaitienė, J. Harmonization of cyclical construction processes: A systematic
review. Procedia Eng. 2017, 208, 190–202. [CrossRef]
2. Golański, M. Wybór materiałów budowlanych w kontekście efektywności energetycznej i wpływu
środowiskowego. Budownictwo i Inżynieria Środowiska 2012, 3, 39–53.
3. Bonenberg, W.; Kapliński, O. The Arcithect and the Paradigms of Sustainable Development: A Review of
Dilemmas. Sustainability 2018, 10, 100. [CrossRef]
4. European Union (EU). Directive 2010/31/EU of 19 May 2010 on the Energy Performance of Buildings; European
Union: Brussels, Belgium, 2010.
5. Zavadskas, E.K.; Antucheviciene, J.; Kalibatas, D.; Kalibatiene, D. Achieving Nearly Zero-Energy Buildings
by applying multi-attribute assessment. Energy Build. 2017, 143, 162–172. [CrossRef]
6. Calkins, M. Materials for Sustainable Sites: A Complete Guide to the Evaluation, Selection, and Use of Sustainable
Construction Materials; John Wiley & Sons: New York, NY, USA, 2008.
7. Thormark, C. The effect of material choice on the total energy need and recycling potential of a building.
Build. Environ. 2006, 41, 1019–1026. [CrossRef]
8. Leśniak, A.; Zima, K. Comparison of traditional and ecological wall systems using the AHP method.
In International Multidisciplinary Scientific GeoConference Surveying Geology and Mining Ecology Management;
SGEM: Albena, Bulgaria, 2015; Volume 3, pp. 157–164.
9. Drozd, W.; Leśniak, A.; Zaworski, S. Construction Time of Three Wall Types Made of Locally Sourced
Materials: A Comparative Study. Adv. Mater. Sci. Eng. 2018, 1–8. [CrossRef]
10. Švajlenka, J.; Kozlovská, M. Houses Based on Wood as an Ecological and Sustainable Housing
Alternative—Case Study. Sustainability 2018, 10, 1502. [CrossRef]
11. Florez, L.; Castro-Lacouture, D. Optimization model for sustainable materials selection using objective and
subjective factors. Mater. Des. 2013, 46, 310–321. [CrossRef]
12. Akadiri, P.O.; Olomolaiye, P.O.; Chinyio, E.A. Multi-criteria evaluation model for the selection of sustainable
materials for building projects. Autom. Constr. 2013, 30, 113–125. [CrossRef]
13. Ghorabaee, M.K.; Amiri, M.; Zavadskas, E.K.; Antucheviciene, J. A new hybrid fuzzy MCDM approach for
evaluation of construction equipment with sustainability considerations. Arch. Civ. Mech. Eng. 2018, 18,
32–49. [CrossRef]
14. Ubarte, I.; Kaplinski, O. Review of the sustainable built environment in 1998–2015. Eng. Struct. Technol. 2016,
8, 41–51. [CrossRef]
15. Runkiewicz, L. Realizacja obiektów budowlanych zgodnie z zasadami zrównoważonego rozwoju.
Przeglad
˛ Budowlany 2010, 81, 17–23.
16. Burtonshaw-Gunn, S.A. Risk and Financial Management in Construction; Routledge: Abington-on-Thames, UK, 2017.
17. Chatterjee, K.; Zavadskas, E.K.; Tamošaitienė, J.; Adhikary, K.; Kar, S. A hybrid MCDM technique for risk
management in construction projects. Symmetry 2018, 10, 46. [CrossRef]
Sustainability 2018, 10, 1608 13 of 14

18. Odimabo, O.O.; Oduoza, C.; Suresh, S. Methodology for Project Risk Assessment of Building Construction
Projects Using Bayesian Belief Networks. Int. J. Constr. Eng. Manag. 2017, 6, 221–234. [CrossRef]
19. Plebankiewicz, E.; Leśniak, A. Overhead costs and profit calculation by Polish contractors. Technol. Econ.
Dev. Econ. 2013, 19, 141–161. [CrossRef]
20. Juszczyk, M. Application of PCA-based data compression in the ANN-supported conceptual cost estimation
of residential buildings. AIP Conf. Proc. 2016, 1738, 200007. [CrossRef]
21. Leśniak, A.; Juszczyk, M. Prediction of site overhead costs with the use of artificial neural network based
model. Arch. Civ. Mech. Eng. 2018, 18, 973–982. [CrossRef]
22. Juszczyk, M.; Leśniak, A.; Zima, K. ANN Based Approach for Estimation of Construction Costs of Sports
Fields. Complexity 2018, 7952434. [CrossRef]
23. Sonmez, R. Parametric range estimating of building costs using regression models and bootstrap. J. Constr.
Eng. Manag. 2008, 134, 1011–1016. [CrossRef]
24. Kim, G.H.; Shin, J.M.; Kim, S.; Shin, Y. Comparison of school building construction costs estimation methods
using regression analysis, neural network, and support vector machine. J. Build. Constr. Plan. Res. 2013, 1,
1–7. [CrossRef]
25. Yazdani-Chamzini, A.; Zavadskas, E.K.; Antucheviciene, J.; Bausys, R. A Model for Shovel Capital Cost
Estimation, Using a Hybrid Model of Multivariate Regression and Neural Networks. Symmetry 2017, 9, 298.
[CrossRef]
26. An, S.-H.; Kim, G.; Kang, K. A Case-based reasoning cost estimating model using experience by analytic
hierarchy process. Build. Environ. 2007, 42, 2573–2579. [CrossRef]
27. Riesbeck, C.K.; Schank, R.C. Inside Case-Based Reasoning; Lawrence Erlbaum Associates: Hillsdale, NJ, USA, 1989.
28. Pal Sankar, K.; Shiu Simon, C.K. Foundations of Soft Case-Based Reasoning; John Wiley & Sons, Inc.: Hoboken,
NJ, USA, 2004.
29. Mendes, E.; Mosley, N.; Counsell, S. The application of casebased reasoning to early web project cost
estimation. In Proceedings—IEEE Computer Society’s International Computer Software and Applications Conference;
Institute of Electrical and Electronics Engineers Computer Society: Oxford, UK, 2002.
30. Kim, G.-H.; An, S.-H.; Kang, K.-I. Comparison of construction cost estimating models based on regression
analysis, neural networks, and case-based reasoning. Build. Environ. 2004, 39, 1235–1242. [CrossRef]
31. Duverlie, P.; Castelain, J.M. Cost estimation during design step: Parametric method versus case based
reasoning method. Int. J. Adv. Manuf. Technol. 1999, 15, 895–906. [CrossRef]
32. Chou, J.-S.; Peng, M.; Persad, K.; O’Connor, J. Quantity-based approach to preliminary cost estimates for
highway projects. Transp. Res. Rec. 2006, 1946, 22–30. [CrossRef]
33. Emsley, M.W.; Lowe, D.J.; Duff, A.R.; Harding, A.; Hickson, A. Data modelling and the application of a
neural network approach to the prediction of total construction costs. Constr. Manag. Econ. 2002, 20, 465.
[CrossRef]
34. Lowe, D.J.; Emsley, M.W.; Harding, A. Predicting construction cost using multiple regression techniques.
J. Constr. Eng. Manag. 2006, 132, 750–758. [CrossRef]
35. Marir, F.; Wang, F.; Ouazzane, K. A case-based expert system for estimating the cost of refurbishing
construction buildings. In Proceedings of the 4th International Conference on Enterprise Information
Systems, Ciudad Real, Spain, 3–6 April 2002; Volume 1, pp. 391–398.
36. Ji, S.-H.; Park, M.; Lee, H.-S. Cost estimation model for building projects using case-based reasoning. Can. J.
Civ. Eng. 2011, 38, 570–581. [CrossRef]
37. Dogan, S.Z.; Arditi, D.; Günaydın, H.M. Determining attribute weights in a CBR model for early cost
prediction of structural system. J. Constr. Eng. Manag. 2006, 132, 1092–1098. [CrossRef]
38. Koo, C.; Hong, T.; Hyun, C.; Koo, K.A. CBR-based hybrid model for predicting a construction duration and
cost based on project characteristics in multi-family housing projects. Can. J. Civ. Eng. 2010, 37, 739–752.
[CrossRef]
39. Ryu, H.G.; Lee, H.S.; Park, M. Construction planning method using case based reasoning (CONPLA-CBR).
J. Comput. Civ. Eng. 2007, 21, 410–422. [CrossRef]
40. Traczyk, W. Inżynieria Wiedzy/Knowledge Engineering; Akademicka Oficyna Wydawnicza EXIT: Warszawa,
Poland, 2010.
Sustainability 2018, 10, 1608 14 of 14

41. Zima, K. Cost Estimating of Football Pitches Construction in the Concept Phase Using Case Based Reasoning.
In Proceedings of the International Conference on Economics and Management Engineering (ICEME 2017),
Wuhan, China, 24–26 March 2017; DEStech Transactions on Economics, Business and Management; 2017;
pp. 24–26.
42. Polish Green Building Council Homepage. Available online: https://plgbc.org.pl/od-czego-zaczac/
(accessed on 3 March 2018).
43. Public Procurement Control Department Homepage. Available online: https://www.uzp.gov.pl/ (accessed
on 6 March 2018).
44. Sekocenbud. Forecasting and Indexation Bulletin—ZWW, 2014–2017; PROMOCJA Sp. z o.o.: Warszawa, Poland.
45. Sekocenbud. Regional Price Bulletin—BCR, 2014–2017; PROMOCJA Sp. z o.o.: Warszawa, Poland.
46. PMI. A Guide to the Project Management Body of Knowledge; Project Management Institute: Newtown Square,
PA, USA, 2008.
47. Kim, S.; Shim, J.H. Combining case-based reasoning with genetic algorithm optimization for preliminary
cost estimation in construction industry. Can. J. Civ. Eng. 2014, 41, 65–73. [CrossRef]
48. Gunaydin, H.; Dogan, S. A neural network approach for early cost estimation of structural systems of
buildings. Int. J. Proj. Manag. 2004, 22, 595–602. [CrossRef]
49. Juszczyk, M. Application of committees of neural networks for conceptual cost estimation of residential
buildings. AIP Conf. Proc. 2015, 1648, 600008. [CrossRef]

© 2018 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).