REVERSE LOGISTIC OPTIMIZATION:

Application to the collect and the reuse of reusable containers

ELMERINI RAJAE BENSLIMANE MOHAMED ZOUADI TARIK

Abdelmalek Essaidi University, Faculty Abdelmalek Essaidi University, Faculty BEAR Lab, Rabat Business School
of Science, of Sciences International University of Rabat (UIR)
Tetouan, Morroco Tetouan, Morroco Rabat, Morocco
[email protected] [email protected] [email protected]

ABSTRACT In the same line of thought, there is a growing interest for the
1Increasing concerns for reusable containers from manufacturers reusable packaging such as wooden pallets and plastic crates, see
and environmentalist is warranted with their less environmental (Accorsi and al., [2]). This concern is warranted by reason of
impact and higher economic benefits. The present study aims to economic benefits related to decrease requirements of containers
propose a model that helps companies to optimize the reusable items. Furthermore the price of reusable containers is amortized
containers pickup and delivery costs within a supply chain and over their life cycle and more the container is reused more the
under a carbon emission constraint. We propose a Mixed Integer cost per shipment reduced, as denoted by (Chandoul and al., [8]).
programming model to solve the problem. The MIP model was In our study we assume that the containers items concerned are
tested in Cplex software optimizer. choosing by reason of economic and environmental significance
of their use, consult (J. Bhattacharjya and P. Kleine-Moellhoff,
[6]).
KEYWORDS In other side, the reuse activity of containers requires a reverse
Reverse logistics, reusable containers, multi-level supply chain, supply chain with particular processes starting with the pick-up of
reuse activity, empty containers from customers; passing by the treatments
process of their quality state as the cleaning and maintaining
ACM Reference format:
operations, the exploitation of containers in a new production line
R. Elmerini, M. Benslimane, T. Zouadi. SIG Proceedings Paper in word in addition to the redistribution of full containers over again,
Format. In Proceedings of ACM LOPAL conference, Rabat, Morocco,
through an organized network, see (Ech-Charrat and al., [10]).
May 2018 (LOPAL’18), 6 pages.
In the present paper, we investigate the reuse activity of
reusable packaging in a reverse supply chain, over a multi-level
1 INTRODUCTION network and under a carbon emission constraint. This model is a
In the industrial world, many companies are working to meet basic formulation of the case with set of supplier sites, set of
economics and environments requirements, by regarding the warehouse sites and set of customer sites by taking into account
trade-offs between costs, environmental impacts and social several simplifying assumptions. A mathematical modeling was
aspects of their activities. Studies such as (Hervani and al., [12]) developed in purpose to optimize the total cost resulting from the
show the need to corporate the sustainability measures within the reverse supply chain management for reusable containers, and
supply chain management and more specifically, care for the solved thereafter with Cplex software optimizer.
reverse logistics, in order to satisfy both customers and
environmentalists alike. Indeed, Reverse logistics management is 2 LITERATURE REVIEW
a planned process of movement of items in inverse direction from
The growing interest in reverse logistics is clear from the
customers back to manufacturers. However, it is more than just
literature of the last years related to different application fields.
returns management, reverse logistics is all activities related of
Agrawal and al., [3] review the current reverse logistics problems
either repair, re-manufacture, reuse, recycle or dispose of items,
as implementation, outsourcing and management of product
according to (Agrawal and al., [3]).
returns.

Permission to make digital or hard copies of all or part of this work for personal or
classroom use is granted without fee provided that copies are not made or distributed
for profit or commercial advantage and that copies bear this notice and the full
citation on the first page. Copyrights for components of this work owned by others
than ACM must be honored. Abstracting with credit is permitted. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prior specific
permission and/or a fee. Request permissions from [email protected]
LOPAL '18, May 2–5, 2018, Rabat, Morocco
© 2018 Association for Computing Machinery.
ACM ISBN 978-1-4503-5304-5/18/05…$15.00
https://doi.org/10.1145/3230905.3230966
 LOPAL’18, May 2018, Rabat, Morocco R. ELMERINI et al.

In reverse logistic, reusing is one of the widely used recovery operations. Santos and al., [15] address also the Pickup and
activity within direct recovery operation alternatives. Chandoul Delivery Problem with Cross-Docking, wherein an Integer
and al., [8] deal with the operational side of the reusable Programming formulation and a Branch-and-price algorithm for
containers management in a network flow of three kinds of site: the problem are proposed. Kheirkhah and Rezaei, [18] consider
customer sites, supplier sites and depots, and propose a model that the problem of a multi-echelon reverse logistics network with
optimizes the total cost resulting from transportation between cross-docking centers, and propose a mathematical model to
sites, purchasing and storage, in a recent paper. Chandoul and al., minimize the total costs across the network. Also Ahmadizar and
[9] investigate the case of returnable container management with al., [4] develop a genetic algorithm hybridized with a local search
purchasing policy, which consist to purchase a quantity of new procedure that deals with the cross-docking problem in a three-
reusable containers at the right period in order to meet system echelon supply chain of suppliers, cross-docks and retailers, in
requirements for product transportation. In the same sense, Ech- order to minimize the purchasing, transportation and the holding
Charrat and al., [10] present a mixed integer programming model, costs. However Küçükoğlu and Öztürk, [17] present the
to solve the dynamic assignment problem of reusable containers transportation problem of cross-docking network design, a Mixed
considering a carbon emission constraint in a finite planning Integer Programming model was developed and to solved with a
horizon. Unlike, Goudenege and al., [11] develop a generic model simulated annealing meta-heuristic algorithm.
for reusable containers in reverse logistics management, focused
on industrial application at a luxury goods company. The case
study shows that the management of reusable containers is not 3 PROBLEM DEFINITION
always less expensive than the traditional cardboard. We address in this study the problem of the reverse logistics
Another stream of research highlights the environmental of reusable containers over a multi-level distribution network.
impacts and sustainability within the supply chain management. The model adapted is a flow distribution through two directions,
J. Bhattacharjya and P. Kleine-Moellhoff, [6] outline key between three kinds of sites the suppliers, warehouses and the
practices and environmental concerns in the management of the clients/stores, as shown in the following figure:
pallet life cycle over a supply chain. Accorsi and al., [2] provide
an assessment of economic and environmental impacts of a multi-
use plastic containers compare to an ordinary single-use
packaging, in a real case study of food manufacturing of a
distribution supply chain for fresh fruits and vegetables.
The cost of carbon emissions is mostly ignored in sourcing
decisions. However, it becomes an important decision variable in
supply chains, Aljazzar and al., [5] investigate a strategy to reduce
additional carbon emissions from transportation in a two- echelon
coordinated supply chain. While Absi and al., [1] mainly focus on
the carbon emission constraints, in multi-sourcing lot-sizing
problems by presenting four types of carbon emission constraints
(Periodic, cumulative, global and rolling). Further, Bouchery and
al., [7] present the multi-objective optimization of costs and Figure 1: Generic model scheme
carbon emission within the non-coordinated two-echelon supply
chain and the centralized chain cases.  A Supplier: is a facility management that product and reuse
Some deterministic optimization approaches have been the empty containers received from Warehouses in order to
proposed for the pickup and delivery problems, Nalepa J. and load and send them back to Warehouses.
Blocho M., [13] discuss the pickup and delivery problem with  A Warehouse: is a facility for storage that keeps the full
time windows (PDPTW), and develop an island-model parallel containers delivered from Suppliers before they distributed to
memetic algorithm to solve the problem. Considering a new city their final destination. Also it used to store the empty
logistics problem arising in the last mile distribution of e- containers received from stores/client to achieve the treatment
commerce between Depots, Satellites, pickup Facilities and process before get back to Suppliers.
Customers Zhou and al., [19] introduce a hybrid multi-population  A Client: is a facility that receives the full containers from
genetic algorithm to minimize the total distribution cost. Warehouses, and return them after the consumption.
Nikolopoulou and al., [14] this study proposes a local-search For a period t, in order to satisfy the Clients demand. The full
algorithm, to tackle the optimization framework of the direct- containers of goods produced by suppliers are delivered and
shipping and cross-docking strategies, in order to highlights the stored in warehouses before they distributed to the clients, in the
characteristics of the two compared strategies. Similar work has inverse direction the empty containers consumed by the
been developed by Vincent and al., [16] which investigate the clients/stores are recuperated and stored in the warehouses for
cross-docking concept, and present a simulated annealing treatment process and then recovered by the suppliers.
heuristic algorithm to minimize the transportation cost and to The aim of this study is to decide on each period, the deliveries
respect the time horizon, for the whole pickup and delivery of each Supplier/Warehouse respectively for each Warehouse
2
 LOPAL’18, May 2018, Rabat, Morocco R. ELMERINI et al.

/Client, and the collected containers quantities from each S0Ei : Initial inventory of Empty containers in a supplier i.
Client/Warehouse and their destinations, in a way to optimize the SFj,t : Inventory stock level of Full containers in a warehouse j at
vehicle routing, so as to minimize the carbon emission. period t.
SEj,t : Inventory stock level of Empty containers in a warehouse
j at period t.
1.1 Assumptions: S0Fj : Initial inventory of Full containers in a warehouse j.
It is important to note that: S0Ej : Initial inventory of Empty containers in a warehouse j.
 The production capacity is different from a supplier to SFy,t : Inventory stock level of Full containers in a client y at
another. period t.
 The storage capacity is different from a site to another, site: SEy,t : Inventory stock level of Empty containers in a client y at
supplier, warehouse and client. period t.
 Each supplier serves many warehouses in the same period, S0Fy : Initial inventory of Full containers in the client y.
S0Ey : Initial inventory of Empty containers in the client y.
and each warehouse delivers many clients.
QRi,t : Quantity of Full containers reproduced by a supplier i at
 Each warehouses has its own returnable containers
period t.
treatment cost. QMj,t : Quantity of Empty containers maintained by a warehouse
 Each supplier, warehouse or client has its own holding cost j at period t.
for full and empty containers.
 The transportation costs are fixed per unit and per client. 1.1.3 Costs data:
 Every vehicle tour is consecrated to the distribution of full CTi,j : Unit Transportation cost of a container between supplier i
containers and the collect of the empty ones, under a limited and warehouse j per distance.
vehicle capacity. CTj,y : Unit Transportation cost of a container between
 The ordering and recovering costs are different for each warehouse j and client y per distance.
CIFi : Inventory cost of full container of supplier i.
suppliers and warehouses.
CIEi : Inventory cost of empty containers of supplier i.
 The delay of the production, treatment and the consumption CIFj : Inventory cost of full container of warehouse j.
is fixed and determined respectively by the suppliers, CIEj : Inventory cost of empty containers of warehouse j.
warehouses and clients. CIFy : Inventory cost of full containers of clients y.
CIEy : Inventory cost of empty containers of client y.
1.2 Model description: COi,j : Ordering cost of warehouse j from supplier i.
Considering a supply chain management of three-level COj,y : Ordering cost of client y from warehouse j.
distribution network, the definition of a various parameters used CRi,j : Recover cost from warehouse j to supplier i.
in the model adapted in this theses is shown in the following: CRj,y : Recover cost from client y to warehouse j.
CPi : Production cost of full containers in the supplier i.
1.1.1 Model data: CMj : Maintaining cost of empty containers in the warehouse j.
N : Number of suppliers.
M : Number of warehouses. 1.1.4 Additional decision data:
P : Number of clients. CO2 : Carbon emissions per kilometer.
T : Number of periods. Emax : maximum average emission per unit.
Sci : storage capacity of supplier i. T: carbon tax
Scj : storage capacity of warehouse j. Z : an arbitrarily large number.
Scy : storage capacity of client y. Fi,j,t : Binary variable representing expedition of full containers
dj,i : Distance between a supplier i and a warehouse j. to warehouse j by supplier i at period t.
dy,j : Distance between a warehouse j and a client y. Ei,j,t : Binary variable representing recuperation of empty
containers by supplier i from a warehouse j at period t.
1.1.2 Decision variables: Fj,y,t : Binary variable representing expedition of full containers
demj,t : Demand of warehouse j at period t. to client y by warehouse j at period t.
demy,t : Demand of client y at period t. Ej,y,t : Binary variable representing recuperation of empty
DFi,j,t : Delivered quantity of Full containers from a supplier i to containers by warehouse j from a client at period t.
a warehouse j at period t.
DEi,j,t : Delivered quantity of Empty containers from a 1.3 Modeling:
warehouse j to a supplier i at period t.
DFj,y,t : Delivered quantity of Full containers from a warehouse According to the management objective of the three-level
j to a client y at period t. distribution network supply chain, and the model data mentioned
DEj,y,t : Delivered quantity of Empty containers from a client y above, for a set of N Suppliers, M Warehouses and P Clients
to a warehouse j at period t. throughout the supply chain, the general investment functions
SFi,t : Inventory stock level of Full containers in a supplier i at over a planning period t can be presented as:
period t. TC: Total amount of Delivery and collect transport, under the two
SEi,t : Inventory stock level of Empty containers in a supplier i at levels of supply chain.
period t.
S0Fi : Initial inventory of Full containers in a supplier i.

3
 LOPAL’18, May 2018, Rabat, Morocco R. ELMERINI et al.

𝑻𝑪 = ∑N M T
i=1 ∑j=1 ∑t=1[(Fijt × COij ) + (Eijt × CR ij ) + ((DFijt + 𝑁
DEijt ) × CTij × dij )] + ∑M P T
j=1 ∑y=1 ∑t=1[(Fjyt × COjy ) + 𝑆𝐸𝑗,𝑡 = 𝑆𝐸𝑗,𝑡−1 − ∑ 𝐷𝐸𝑖𝑗𝑡 + 𝑄𝑀𝑗,𝑡−𝑇𝑚
𝑖=1
(Ejyt × CR jy ) + ((DFjyt + DEjyt ) × CTjy × djy )] 𝑃
+∑ 𝐷𝐸𝑗𝑦𝑡 (9)
𝑦=1
SC: Total amount of Storage includes the initial stock investment.
∀ j ( ∀ Tm < 1, if not QM = 0)
𝑺𝑪 = ∑N T
i=1 ∑t=1[(SFit × CIFi ) + (SEit × CIEi )] + ×∑N
i=1[(S0 Fi
CIFi ) + (S0 Ei × CIEi )] + ∑M T
j=1 ∑t=1[(SFjt × CIFj ) + (SEjt × 𝑆𝐹𝑦,1 = 𝑆0 𝐹𝑦 − 𝑑𝑒𝑚𝑦,1 + ∑𝑀
𝑗=1 𝐷𝐹𝑗𝑦1 ∀y (10)
CIEj )] + ∑Mj=1[(S0 Fj × CIFj ) + (S0 Ej × CIEj )] +
∑Py=1 ∑Tt=1[(SFyt × CIFy ) + (SEyt × CIEy )] + ∑Py=1[(S0 Fy ×
CIFy ) + (S0 Ey × CIEy )] 𝑆𝐹𝑦,𝑡 = 𝑆𝐹𝑦,𝑡−1 − 𝑑𝑒𝑚𝑦,𝑡 + ∑𝑀
𝑗=1 𝐷𝐹𝑗𝑦𝑡 ∀y (11)

PC: Total amount Production represents the reproduction and

maintaining cost of the empty containers. 𝑆𝐸𝑦,1 = 𝑆0 𝐸𝑦 + 𝑑𝑒𝑚𝑦,1 − ∑𝑀
𝑗=1 𝐷𝐸𝑗𝑦1 ∀y (12)
N T M T

𝑷𝑪 = ∑ ∑[QR it × CPi ] + ∑ ∑[QMjt × CMj ]

i=1 t=1 j=1 t=1 𝑆𝐸𝑦,𝑡 = 𝑆𝐸𝑦,𝑡−1 + 𝑑𝑒𝑚𝑦,𝑡 − ∑𝑀
𝑗=1 𝐷𝐸𝑗𝑦𝑡 ∀𝑦 (13)

P: Penalty imposed when the carbon emission exceeds the

𝑁 𝑀 𝑇
maximum threshold.
∑ ∑ ∑[(𝐷𝐹𝑖𝑗𝑡 + 𝐷𝐸𝑖𝑗𝑡 ) × 𝐶𝑂2 × 𝑑𝑖𝑗 ]
𝑖=1 𝑗=1 𝑡=1
𝑷 = 𝐶𝑂2 × 𝑇 𝑀 𝑇 (14)
≤ 𝐸𝑚𝑎𝑥 × ∑ ∑ 𝑑𝑒𝑚𝑗𝑡
The objective function is given by:
𝑗=1 𝑡=1

𝑀𝑖𝑛(𝑃𝐶 + 𝑆𝐶 + 𝑇𝐶 + 𝑃) (1) M P T

∑ ∑ ∑[(DFjyt + DEjyt ) × CO2 × djy ]

Subject to: i=1 j=1 t=1
P T (15)
𝑀
𝑆𝐹𝑖,1 = 𝑆0 𝐹𝑖 + 𝑄𝑅𝑖,1−𝑑𝑒𝑙𝑎𝑦 − ∑ 𝐷𝐹𝑖𝑗1 ≤ Emax × ∑ ∑ demyt
𝑗=1 (2) y=1 t=1
∀ i ( ∀ delay < 1, if not QR = 0)
𝑀
DFijt ≤ Z × Fijt ∀i ∀j ∀t (16)
𝑆𝐹𝑖,𝑡 = 𝑆𝐹𝑖,𝑡−1 + 𝑄𝑅𝑖,𝑡−𝑑𝑒𝑙𝑎𝑦 − ∑ 𝐷𝐹𝑖𝑗𝑡
𝑗=1 (3)
∀ i ( ∀ delay < 1, if not QR = 0) DEijt ≤ Z × Eijt ∀i ∀j ∀t (17)

𝑆𝐸𝑖,1 = 𝑆0 𝐸𝑖 − 𝑄𝑅𝑖,1 + ∑𝑀
𝑗=1 𝐷𝐸𝑖𝑗1 ∀i (4)
DFjyt ≤ Z × Fyjt ∀j ∀y ∀t (18)

𝑆𝐸𝑖,𝑡 = 𝑆𝐸𝑖,𝑡−1 − 𝑄𝑅𝑖,𝑡 + ∑𝑀

𝑗=1 𝐷𝐸𝑖𝑗𝑡 ∀i (5)
DEjyt ≤ Z × Ejyt ∀j ∀y ∀t (19)

𝑆𝐹𝑗,1 = 𝑆0 𝐹𝑗 + ∑𝑁 𝑃
𝑖=1 𝐷𝐹𝑖𝑗1 − ∑𝑦=1 𝐷𝐹𝑗𝑦1 ∀j (6)
SFit + SEit ≤ 𝑆𝑐i ∀i ∀t (20)

𝑆𝐹𝑗,𝑡 = 𝑆𝐹𝑗,𝑡−1 + ∑𝑁 𝑃
𝑖=1 𝐷𝐹𝑖𝑗𝑡 − ∑𝑦=1 𝐷𝐹𝑗𝑦𝑡 ∀j (7)
SFjt + SEjt ≤ 𝑆𝑐j ∀j ∀t (21)

𝑁
𝑆𝐸𝑗,1 = 𝑆0 𝐸𝑗 − ∑ 𝐷𝐸𝑖𝑗1 + 𝑄𝑀𝑗,1−𝑇𝑚 SFyt + SEyt ≤ 𝑆𝑐y ∀y ∀t (22)
𝑖=1
𝑃
+∑ 𝐷𝐸𝑗𝑦1 (8)
𝑦=1

∀ j ( ∀ Tm < 1, if not QM = 0)

4
 LOPAL’18, May 2018, Rabat, Morocco R. ELMERINI et al.

[11] Goudenege, G., Chu, C., & Jemai, Z. (2013, January). Reusable
Containers Management: From a Generic Model to an Industrial Case
𝑆𝐹𝑖,𝑡 ; 𝑆𝐸𝑖,𝑡 ; 𝑆𝐹𝑗,𝑡 ; 𝑆𝐸𝑗,𝑡 ; 𝑆𝐹𝑦,𝑡 ; 𝑆𝐸𝑦,𝑡 ; 𝑆0 𝐹𝑖 ; Study. In Supply Chain Forum: An International Journal (Vol. 14, No. 2,
pp. 26-38). Taylor & Francis.
𝑆0 𝐸𝑖 ; 𝑆0 𝐹𝑗 ; 𝑆0 𝐸𝑗 ; 𝑆0 𝐹𝑦 ; 𝑆0 𝐸𝑦 ; 𝑄𝑅𝑖,𝑡 ; 𝑄𝑀𝑗,𝑡 ; 𝐷𝐹𝑖𝑗𝑡 ; (23) [12] Hervani, A. A., Sarkis, J., & Helms, M. M. (2017). Environmental goods
𝐷𝐸𝑖𝑗𝑡 ; 𝐷𝐹𝑗𝑦𝑡 ; 𝐷𝐸𝑗𝑦𝑡 ; 𝑑𝑒𝑚𝑖𝑡 ; 𝑑𝑒𝑚𝑗𝑡 ; ≥ 0 valuations for social sustainability: A conceptual framework.
Technological Forecasting and Social Change, 125, 137-153.
[13] Nalepa, J., & Blocho, M. (2017, March). A Parallel Memetic Algorithm
The objective function (1) minimizes the transportation, for the Pickup and Delivery Problem with Time Windows. In Parallel,
Distributed and Network-based Processing (PDP), 2017 25th Euromicro
treatment, filling, capping and holding cost of reusable containers International Conference on (pp. 1-8). IEEE.
among Suppliers, warehouses and the Clients. Constraints (2&3), [14] Nikolopoulou, A. I., Repoussis, P. P., Tarantilis, C. D., & Zachariadis, E.
(4&5), (6&7), (8&9), (10&11) and (12&13) respectively ensure E. (2017). Moving products between location pairs: cross-docking versus
direct-shipping. European Journal of Operational Research, 256(3), 803-
the flow conservation of the storage for full and empty containers 819.
in the Suppliers, warehouses and Clients. Constraints (14) & (15) [15] Santos, F. A., Mateus, G. R., & Da Cunha, A. S. (2013). The pickup and
respectively guaranteed that the carbon emission does not exceed delivery problem with cross-docking. Computers & Operations Research,
the maximum unitary environmental impact allowed, during 40(4), 1085-1093.
[16] Vincent, F. Y., Jewpanya, P., & Redi, A. P. (2016). Open vehicle routing
transport to warehouses and Clients. Constraints (16&17) and problem with cross-docking. Computers & Industrial Engineering, 94, 6-
(18&19) presents respectively the cancellation of full and empty 17.
containers deliveries when no delivery is programmed in [17] Küçükoğlu, İ., & Öztürk, N. (2014). Simulated annealing approach for
transportation problem of cross-docking network design. Procedia-Social
Suppliers or Warehouses. Constraints (20), (21) and (22) show and Behavioral Sciences, 109, 1180-1184.
respectively the respect of the inventory capacity of the Suppliers, [18] Kheirkhah, A., & Rezaei, S. (2016). Using cross-docking operations in a
Warehouses and Clients. Constraints (23) note that all variables reverse logistics network design: a new approach. Production
are positive. Engineering, 10(2), 175-184.
[19] Zhou, L., Baldacci, R., Vigo, D., & Wang, X. (2018). A Multi-Depot
Two-Echelon Vehicle Routing Problem with Delivery Options Arising in
4 Perspectives & Conclusion: the Last Mile. Distribution. European Journal of Operational Research,
265(2), 765-778.
In this work, we tackle the problem of the reusable containers in
a reverse supply chain, over a multi-level network and under a
carbon emission constraint. A mathematical model was developed
in order to optimize the total costs of network, and tested with the
Cplex software optimizer. Along the same lines, we are currently
developing two approaches based on meta-heuristics algorithm to
obtain a good solutions quality in a moderate computational time.

5 References:
[1] Absi, N., Dauzère-Pérès, S., Kedad-Sidhoum, S., Penz, B., & Rapine, C.
(2013). Lot sizing with carbon emission constraints. European Journal of
Operational Research, 227(1), 55-61.
[2] Accorsi, R., Cascini, A., Cholette, S., Manzini, R., & Mora, C. (2014).
Economic and environmental assessment of reusable plastic containers: A
food catering supply chain case study. International Journal of Production
Economics, 152, 88-101.
[3] Agrawal, S., Singh, R. K., & Murtaza, Q. (2015). A literature review and
perspectives in reverse logistics. Resources, Conservation and Recycling,
97, 76-92.
[4] Ahmadizar, F., Zeynivand, M., & Arkat, J. (2015). Two-level vehicle
routing with cross-docking in a three-echelon supply chain: A genetic
algorithm approach. Applied Mathematical Modelling, 39(22), 7065-
7081.
[5] Aljazzar, S. M., Gurtu, A., & Jaber, M. Y. (2018). Delay-in-payments-A
strategy to reduce carbon emissions from supply chains. Journal of
Cleaner Production, 170, 636-644.
[6] Bhattacharjya, J., & Kleine-Moellhoff, P. (2013, September).
Environmental concerns in the design and management of pallets. In
Working Conference on Virtual Enterprises (pp. 569-576). Springer,
Berlin, Heidelberg.
[7] Bouchery, Y., Ghaffari, A., Jemai, Z., & Tan, T. (2017). Impact of
coordination on costs and carbon emissions for a two-echelon serial
economic order quantity problem. European Journal of Operational
Research, 260(2), 520-533.
[8] Chandoul, A., Cung, V. D., & Mangione, F. (2007). Reusable containers
within reverse logistic context
[9] Chandoul, A., Cung, V. D., & Mangione, F. (2009, December). Optimal
repositioning and purchasing policies in returnable container
management. In Industrial Engineering and Engineering Management,
2009. IEEM 2009. IEEE International Conference on (pp. 1439-1443).
IEEE.
[10] Ech-Charrat, M. R., AMECHNOUE, K., & Zouadi, T. Hybrid resolution
approaches for dynamic assignment problem of reusable containers in
close loop supply chain.