A Multi-Agent based Decision Framework for Sustainable Supplier
Selection, Order Allocation and Routing Problem
Maria Drakaki
1
, Hacer Güner Gören
2
and Panagiotis Tzionas
1
1
Department of Automation Engineering, Alexander Technological Educational Institute of Thessaloniki,
P.O. Box 141, GR-574 00, Thessaloniki, Hellas, Greece
2
Department of Industrial Engineering, Pamukkale University, Kinikli Campus, Denizli, Turkey
Keywords: Supplier Selection, Order Allocation, Vehicle Routing, Sustainability, Multi-Agent System, Web Services.
Abstract: Supply chain decisions should aim for sustainability, in order to meet the global market needs, as well as the
Industry 4.0 requirements, therefore they should consider beyond economic and environmental, societal
dimensions as well. The complexity in decision making increases, moreover, supply network relationships
become important, including inter-relationships and those developed with the suppliers. Agent technology is
compatible with Industry 4.0, whereas multi-agent systems (MAS) can provide decision support for supply
chain management and model the relatationships and interactions between entities in the supply chain
environment. Therefore, in this paper, a MAS-based framework is proposed to address sustainability focused
decision making in supplier selection, order allocation and routing. Fuzzy Multi Criteria Decision Making
(MCDM) approaches and multi-objective programming are used by the agents in the MAS in order to adress
sustainability requirements. Futrhermore, developed agent services for the supply chain business processes
are integrated with web services, in order to facilitate business process execution as web services.
1 INTRODUCTION
From the perspective of a network, supply chain
entities must collaborate in order to supply, produce,
deliver and recover products, therefore, relationships
including coordination and collaboration, between
supply chain partners, as well as with suppliers and
customers affect the network performance.
Furthermore, globalisation has shifted the focus of
supply chain performance from pure economic
profitability or even economic and environmental
aspects to sustainability. The performance of supply
chains with respect to sustainability is measured in
terms of operations that meet the needs of current
population which do not compromise future needs
(Krysiak, 2009). Sustainability dimensions, labelled
as Triple Bottom Line (TBL) dimensions (Elkington,
1997), include economic, environmental and social
ones. Economic sustainability refers to fiscal
performance, whereas environmental sustainability
relates to green supply chain management and the
management of scarce environmental resources.
Social sustainability refers to fair practices at work,
occupational health and safety, as well as social
welfare (Aktin and Gergin, 2016). Globalisation is a
key driver for integration of sustainability in supply
chain management. Global supply chains face
increased risks, whereas sustainability integration
could address these risks (Giannakis and
Papadopoulos, 2016).
Global supply chains consist of distributed and
autonomous business entities which collaborate with
each other, whereas they communicate with the
Internet. Agent technology is increasingly being used
in supply chain business processes, due to its
distributed artificial intelligence origin and the
capability of enabling interactions between the
different autonomous, distributed software agents
(Woolridge and Jennings, 1995), connected in a
network. Agents can represent various supply chain
entities, business processes, machines, vehicles, as
well as information and material elements. Business
entities in global supply chains use negotiation,
coordination and cooperation mechanisms in order to
jointly deliver supply chain services and products,
whereas these interaction features are inherent in
agent technology (Swaminathan, 1998; Long, 2011).
In the context of the supply chain, multi-agent
systems (MAS) enable decision support by using
individual agents, each one with local knowledge,
Drakaki, M., GÃ˝uren, H. and Tzionas, P.
A Multi-Agent based Decision Framework for Sustainable Supplier Selection, Order Allocation and Routing Problem.
DOI: 10.5220/0007833306210628
In Proceedings of the 5th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2019), pages 621-628
ISBN: 978-989-758-374-2
Copyright
c
2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
621
capable of interacting to achieve the global supply
system goal. Collaborative manufacturing decision
support including supplier selection (such as Jiao et
al., 2006, Trappey et al., 2007; Yu et and Wong, 2015;
Ghadimi et al., 2018), and vehicle routing as well as
intelligent transportation has been provided by MAS
and agent technology (such as Davidsson et al, 2005;
Martin et al., 2016). Collaboration in sustainable
supply chain management aims to meet the
sustainability goals. However, the impact of logistics
in sustainability has not been well explored.
Logistics, including transportation, contribute to the
total carbon footprint, therefore, efficient planning
and execution of logistics and transportation network
can positively affect the carbon footprint of supply
chains as well the long-term sustainability goals
(Reefke and Sundaram, 2017). Hofmann and Rusch
(2017) explore the potential of Industry 4.0 in
logistics management. They argue that Industry 4.0
will result in the deployment of autonomous
knowledge-based, self-regulating production
systems, as well as the emergence of new services and
business models. They suggest a physical supply
chain model which includes autonomous, self-
controlled logistics sub systems, such as transport
units, or order processing units, interacting with each
other. The digital supply chain model includes
different types of data, transferred via a connectivity
layer, such as in the cloud, in order to be processed,
delivering value-added business services. Just-in-
Time systems, which focus on buyer-supplier
relationships, will benefit, since suppliers will receive
real-time production order information at buyer sites
via cloud-based ERP systems, thereby, triggering
their production.
Digital integration as well as servitisation add
value in smart supply chains. Supplier selection is a
critical component of supply chain performance
(Ghadimi and Heavy, 2014). Ho et al. (2009)
emphasize that supply chain management goals
include long term partnerships with suppliers,
therefore a few but reliable suppliers are prefered.
Govindan et al. (2015) further consider suppliers as
the fifth echelon in the sustainable order allocation
and supply chain network design. Sustainable
supplier selection and subsequent order allocation is
crucial in supply chain management, therefore
organisations must cooperate with suppliers on
sustainable practices. Besides, in the sustainable
supplier selection process, long-term relationship-
continuity has been identified as a sustainability
criterion (Gören, 2018). The critical decisions in
supplier selection include the types of products,
identification of suitable suppliers, order quantities
and time periods for order allocation (Songhori et al.,
2010). Order allocation refers to the decisions
regarding the order quantities to order from each
selected supplier. Traditional supplier selection has
focused on economic criteria, such as cost, quality,
delivery times and has been summarised in reports
(such as Ho, 2009). Sustainable supplier selection
considers the tripple-bottom line dimensions (Gören,
2018). Research on sustainable supplier selection
considering economic, environmental, as well as the
social dimension (such as Kuo et al., 2010; Govindan
et al., 2013) is growing. However, sustainable
supplier selection with order allocation has been the
focus of a limited number of studies (such as Αktin
and Gergin, 2016; Gören, 2018; Govindan et al.,
2015; Ghadimi, 2018). Transportation and
distribution decisions affect logistics performance.
Vehicle Routing Problem (VRP) optimises routes for
vehicles from a set of depots to a set of destinations
(Laporte, 1995). VRP is a supply chain optimisation
method including optimisation of routes between
suppliers and customers (Wang et al., 2018).
Songhori (2010) consider supplier selection with
order allocation and optimal selection of
transportation alternatives. Supplier selection with
order allocation and vehicle routing has been studied
in the literature only recently (Govindan et al., 2017;
Nasiri et al., 2018). Nasiri et al. (2018) address
supplier selection and order allocation and vehicle
routing for the multi-cross-dock problem with mixed
integer linear programming. Govindan et al. (2017)
present a closed loop supply chain network design
which integrates decisions on supplier selection,
order allocation, vehicle selection and routing.
However, the authors have not addressed
sustainability in their method. Ghadimi et al. (2018)
develop a MAS method for sustainable supplier
selection and order allocation. The authors argue that
the proposed MAS enhanced structured
communication and information exchange in the
partnership, and therefore, enhanced the long-term
relationships between buyer and suppliers as well as
their partnership.
In this paper, an intelligent MAS is presented to
assist in the integrated decision making in sustainable
supplier selection with order allocation and routing.
MAS agents represent supply chain entities such as
project managers, information elements such as
knowledge manager, business processes, such as
supplier selection, order allocation and vehicle
routing suppliers, as well as vehicles. Agent types are
categorized to execution, information, outsourcing
partner and mobile agents. Individual agents use
different methods for local decision making include
LogiTrans 4.0 2019 - Special Session on Logistics and Transport in the Industry 4.0
622
fuzzy MCDM and optimization. The MAS global
goal is to make decisions in supplier selection, order
allocation and vehicle routing taking into account
logistics oriented Industry 4.0 concepts including
sustainability. The proposed approach facilitates
cooperation and communication between different
supply chain members and enhances supply chain IT
performance, since it can integrate with web services.
In the following, the proposed method is
presented next, followed by a description of the
proposed method. Finally, conclusions are drawn.
2 THE PROPOSED METHOD
2.1 The MAS
The MAS is developed in order to achieve the
following tasks:
a) to evaluate and select appropriate supplier(s)
taking into account sustainability dimensions,
b) to allocate orders to the selected suppliers, and
c) to configure product pickup vehicle routing
starting from the depot and visiting suppliers in
order to collect purchased products.
The MAS has been developed with Java Agent
Development Framework (JADE) (Bellifemine et al.,
2007). Agents interact with the FIPA-ACL
Interaction Protocol (IP). The MAS agents are
classified as execution, information, outsourcing
partner and mobile agents. Agents located in different
sites and JADE platforms communicate with the
HTTP Message Transport Protocol. Lim and Zhang
(2004) have classified agents such as information and
execution. Wang and Lin (2009) classified agents as
soft agents, namely information and execution ones
as well as mobile ones. In this paper, execution agents
are responsible for carrying out procedures and
making decisions. Information agents are responsible
for giving information including data to other agents
upon request. Ousourcing partner agents represent
suppliers, whereas mobile agents represent vehicles,
Outsourcing partner agents represent outsourcing
partners, they can make decisions as well as provide
data upon request and include the supplier agents.
Mobile agents can move according to a scheduling
and routing plan, provide information upon request
and include vehicle agents. However, mobile agents
could represent different mobile elements such as
data or products. The MAS consists of a project
manager agent (PMA), a coordinator agent (CA), a
supplier selection agent (SSA), a knowledge manager
Figure 1: The MAS agent interaction diagram for the
supplier selection process.
Table 1: MAS agents and their respective goals.
Agent Goal
Project Release Agent (PRA) Releases sequentially the supplier selection, order allocation and
vehicle routing tasks to be executed by the MAS. Communicates with
the CA.
Coordinator Agent (CA) Coordinates task executions. Communicates with the SSA, OAA and
VRA.
Knowledge Agent (KMA) Retrieves and stores data and knowledge to the databases.
Communicates with the SSA, OAA and VRA.
Supplier Selection Agent (SSA) Executes the supplier selection task, evaluates and ranks potential
suppliers. Communicates with the SAs, KMA and CA.
Supplier Agent (SA) Represents potential suppliers. Each SA provides supplier data
necessary for the supplier selection task. Communicates with the SSA.
Order Allocator Agent (OAA) Executes the order allocation task. Communicates with the CA, KMA
and OA.
Optimisation Agent (OA) Communicates with the OAA and KMA.
Vehicle Routing Agent (VRA) Executes the vehicle routing task. Communicates with the CA, KMA,
VAs and OA
Vehicle Agent (VA) Represents vehicles to be used for routing. Each VA provides vehicle
data necessary for the vehicle routing task. Communicates with the
VRA.
A Multi-Agent based Decision Framework for Sustainable Supplier Selection, Order Allocation and Routing Problem
623
agent (KMA), supplier agent(s) (SA), vehicle
agent(s) (VA), an order allocation agent (OAA), an
optimisation agent (OA) and a vehicle routing agent
(VRA). Execution agents include the project manager
agent (PMA), the coordinator agent (CA), the
supplier selection agent, the order allocation agent,
the optimisation agent and the vehicle routing agent.
Information agents include the knowledge manager
agent, supplier agent and product agents. Table 1
shows the agents of the MAS. Figure 1 shows the
MAS agent interaction diagram for the supplier
selection process and Figure 2 shows the agent IP for
the presented MAS focusing on supplier selection and
order allocation processes.
Figure 2: The agent interaction protocol for the presented
MAS for the supplier selection and order allocation
processes.
2.2 Elements of the MAS
Project Release Agent (PRA)
PRA decides on outsourcing tasks based on
information received from project managers and
releases project tasks to the coordinator agent. Tasks
include supplier selection, order allocation and
vehicle routing, taking into account sustainability
(TBL) dimensions. PRA requests from CA to initiate
supplier selection process. Upon receiving
notification from CA that the task has been
completed, he requests from CA order allocation
execution. Upon receiving notification from CA that
the task has been completed, he requests from CA
vehicle routing execution. PRA communicates with
the FIPA-Request IP.
Coordinator Agent (CA)
CA
coordinates
the
execution
of
the
tasks
issued
by
PRA. He receives requests for task execution from
PRA. When he receives request from PRA for
supplier selection, the agent requests from SSA to
evaluate and rank suppliers, taking into account
sustainability. After SSA informs CA that he finishes
the task, CA informs PRA of the results. When he
receives request from PRA for order allocation, he
requests from OAA to allocate orders to the suppliers.
After CA receives results from OAA, he informs PRA
of the results. When CA receives request from PRA
for vehicle routing execution, he requests VRA to do
the vehicle routing. After he receives the vehicle
routing results from VRA, the agent informs PRA of
the results. CA communicates with the FIPA-Request
IP.
Knowledge Manager Agent (KMA)
KMA receives requests for information regarding the
list of potential suppliers and sustainability criteria
from the SSA and informs SSA on the requested
information based on information, he retrieves from
the databases. The agent has access to supplier
database. KMA receives requests for information
regarding ranking of suppliers from OAA and
informs OAA about the results. He receives order
allocation results from OAA and informs the supplier
and manufacturer databases. KMA receives requests
for order allocation information from VRA and
informs VRA about the results. He receives vehicle
routing results from VRA and informs the supplier
and manufacturer databases. KMA communicates
with the FIPA-Request IP.
Supplier Selection Agent (SSA)
SSA receives request for sustainable supplier
selection from CA. SSA communicates with KMA to
receive list of suppliers and list of sustainability
criteria for supplier evaluation. Next, he
communicates with the potential supplier agents
(SAs) in order to obtain the necessary data for
supplier evaluation. He may use Contract Net
Protocol (CNP) for communicating with the
suppliers. A supplier agent (SA) may refuse to enter
into negotiation with SSA. Alternatively, he may
communicate with potential suppliers with FIPA
Request IP. He applies different fuzzy MCDM
approaches such as AHP, TOPSIS etc. to evaluate and
rank suppliers. SSA sends evaluation results to CA
and KMA. SSA communicates with the FIPA-
Request IP and FIPA-CNP.
Order Allocation Agent (OAA)
OAA receives request for order allocation from CA.
OAA requests from KMA supplier supplier ranking
results and product data. OAA executes a bi-objective
LogiTrans 4.0 2019 - Special Session on Logistics and Transport in the Industry 4.0
624
model to allocate orders to potential suppliers. When
the agent receives optimisation results from OA, he
informs both CA and KMA about the vehicle routing
results. OAA informs order allocation results to the
CA and KMA. OAA communicates with the FIPA-
Request IP or FIPA-CNP.
The Bi-objective Model for Order Allocation
Multiple products and multiple periods are assumed
in the model which is adapted from Gören (2018).
Indices:
i: index of a supplier
j: index of a product
t: index of a period
Parameters:
S: number of suppliers
P: number of products
T: number of periods
C
i
: capacity of supplier i
P
ij
: Purchasing price of product j from supplier i
W
i
: Supplier rating value
O
ij
: Ordering cost of product j from supplier i
H
j
: holding cost for product j
q
ij
: Average defect percent of supplier i for product j
Q
j
: Maximum acceptable defect ratio for product j
p
ij
: production time of supplier j for product j
Decision Variables:
X
ijt
: Quantity of product j delivered by supplier i in
period t.
Y
ijt
: Binary variable equal to 1 if an order is placed to
supplier i for product j in period, else 0.
I
jt
: Available inventory of product j at the end of
period t.
Objective Functions and Constraints
The first objective function aims to minimise the total
cost of purchasing (TCP) from the suppliers.




∗


∗





∗
(1)
The second objective function aims to maximise the
total sustainability value (TSV) of the suppliers.



∗




(2)
Constraints:
Demand constraint:





∀
,
(3)
Quality constraint:


∗



∀
,
(4)
Supplier capacity constraint:


∗



∀,
(5)
Non-negativity constraint:

0 ,
,
(6)
Each objective function subject to the constraints is
solved separately. Results of each objective function
are sent to the optimisation agent (OA). The bi-
objective model is finally optimised and solved by
OA.
Optimisation Agent (OA)
OA receives requests for optimisation from OAA and
VRA. He informs the results of optimisation to OAA
and VRA. The objective of OA is to find a set of
optimal solutions that satisfy multiple objectives
which could be conflicting, subject to a set of
constraints.
OA follows the procedure adopted in (Hamdan
and Cheaitou, 2017). OA receives from either OAA
or VRA the solution to each objective function solved
separately. A weighted approach is followed by OA
in order to merge the two objective functions in a
single objective function, f, which is shown below in
response to OAA requests
 


(7)





(8)


(9)
where α
1
and α
2
are relative weights. Their sum is
equal to 1. Equation (9) gives the optimized solution
for the order allocation. OA informs OAA with the
results.
Vehicle Routing Agent (VRA)
VRA receives request for vehicle routing from CA.
VRA requests from KMA order allocation results as
well as product data and from vehicle agents data
regarding vehicle parameters. VRA executes a bi-
objective model for vehicle routing. Each objective
A Multi-Agent based Decision Framework for Sustainable Supplier Selection, Order Allocation and Routing Problem
625
function subject to the constraints is solved
separately. Results of each objective function are sent
to the optimisation agent (OA). The bi-objective
model is finally optimised and solved by OA. When
the agent receives optimisation results from OA, he
informs both CA and KMA about the vehicle routing
results. VRA communicates with the FIPA-Request
IP or FIPA-CNP. Heterogeneous vehicles with
different capacities, costs and carbon emission rates
are represented by the vehicle agents.
The Bi-objective Model for Vehicle Routing
The problem is formulated such that each route starts
and ends at the depot. The load of each vehicle should
not exceed its capacity. Each supplier is visited once
by one vehicle during each period.
Indices:
i: index of a supplier
j: index of a product
t: index of a period
v: index of a vehicle
Parameters:
S: number of suppliers
P: number of products
T: number of periods
V: number of vehicles
W
j
: weight of product j
CP
v
: capacity of vehicle v
FC
v
: Fixed cost for vehicle v
VC
v
: Variable cost for vehicle v
CE
v
: Carbon emission of vehicle v per km
dn
i
n
j
: Distance between suppliers (nodes) i and j
Decision Variables:
X
ijvt
: Quantity of product j of supplier i in period t
delivered by vehicle v.
y
ninjtv
: : Binary variable equal to 1 if arc (n
i
, nj) is part
of the route of vehicle v in period t.
c
0ijtv
: Binary variable equal to 1 if vehicle v starts from
depot and visits immediately after supplier i in period
t.
Objective Functions and Constraints
The first objective function aims to minimise the total
cost of transportation activities (TCT).
The second objective function aims to minimise
total carbon emissions from the vehicles used to
pickup and deliver the purchased products (TCE).


∑∑∑


∗





∑∑

∗






(10)







∗


(11)
Constraints:
Vehicle load capacity constraint:


∗



∀,
(12)
Equation (13) states that each vehicle visits at most
one node at the beginning of the pickup:


1∀,




(13)
Equations (14)-(16) state the degree constraints and
route continuity:


1 ,
1,2.,

(14)


1 ,


1,2….,
(15)




∀,




(16)
The bi-objective model is solved following with the
procedure adopted in (Hamdan and Cheaitou, 2017)
by OA. VRA requests from OA optimisation,
providing to the agent TCT, TCT
min
, TCE and TCE
min
results.
The Java API of CPLEX is used by the MAS for
the multi-objective programming development.
2.3 Integration of MAS with Web
Services
Agent services can be published as web services by
using the Web Service Integration Gateway (WSIG)
in JADE (
Bellifemine et al., 2007). Figure 3 shows the
integrated MAS with WSIG architecture. WSIG
consists of two basic components, namely the WSIG
Servlet and the WSIG Agent. The WSIG Servlet is
the front-end to the internet. Its tasks include serving
incoming HTTP/SOAP (Simple Object Access
Protocol) requests, determining the requested agent
action and informing the WSIG agent, as well as
sending the HTTP/SOAP response to the client. The
WSIG Agent is the gateway between the internet and
LogiTrans 4.0 2019 - Special Session on Logistics and Transport in the Industry 4.0
626
the agent worlds. The tasks include forwarding
requested agent actions to the agents in order to
perform them and receiving the responses from the
agents, as well as creating the Web Service
Description Language (WSDL) corresponding to
each agent registered service and publish it in a
Universal Description, Discovery and Integration
(UDDI) repository.
Figure 3: The integrated MAS with WSIG architecture.
3 CONCLUSIONS
Digital servitisation in smart supply chains requires
synchronisation of business processes and real time
information exchange between supply chain
members. Challenges created by sustainability
requirements lead to the emergence of new methods
in order to address business processes. Therefore, in
this paper, a MAS-based framework is proposed to
address sustainability requirements and integrated
decision making in supplier selection, order
allocation and vehicle routing. Fuzzy MCDM
approaches and multi-objective programming are
used by the agents in the MAS. Agents represent
different supply chain entities, business processes,
information elements as well as vehicles. They are
categorized as execution, information, outsourcing
partner and mobile agents. Furthermore, developed
agent services for the supply chain business processes
are integrated with web services, in order to facilitate
business process execution as web services. The
proposed method will be tested on a real case study in
the future studies.
REFERENCES
Bellifemine, F. L., Caire, G. & Greenwood, D. Developing
multi-agent systems with JADE (Vol. 7). John Wiley &
Sons (2007).
Wooldridge, M. & Jennings, N.R. (1995). Intelligent
agents: Theory and practice. The Knowledge
Engineering Review. 10, 115–152.
Gören Güner, H. (2018). A decision framework for
sustainable supplier selection and order allocation with
lost sales. Journal of Cleaner Production, 183, 1156-
1169.
Ho, W., Xu, X., Dey, P.K., (2009). Multi-criteria Decision
making approaches for supplier evaluation and
selection: a literature Review. European Journal of
Operational Research, 202, 16-24.
Swaminathan JM, Smith SF and Sadeh NM. (1998).
Modeling supply chain dynamics: a multi-agent
approach. Decision Sciences, 29 (3), 607–632.
Krysiak, F.C. (2009). Risk management as a tool for
sustainability. Journal of Business Ethics, 85 (3), 483-
492.
Long Q, Lin J, and Sun Z. (2011). Modeling and distributed
simulation of supply chain with a multi-agent platform.
International Journal of Advanced Manufacturing
Technology, 55, 1241–1252.
Elkington, J. (1998). Cannibals with forks: The triple
bottom line of the 21st century. Stoney Creek/CT: New
Society.
Aktin, Tülin, Gergin, Zeynep, (2016). Mathematical
modelling of sustainable procurement strategies: three
case studies. Journal of Cleaner Production, 113, 767-
780.
Giannakis, M., Papadopoulos, T. (2016). Supply chain
sustainability: a risk management approach.
International Journal Production Economics, 171, 455-
470.
Wooldridge, M. & Jennings, N.R. (1995) Intelligent agents:
Theory and practice. Knowledge Engineering Review,
10, 115–152.
Jiao J., You X. & Kumar A. (2006). An agent-based
framework for collaborative negotiation in the global
manufacturing supply chain network, Robotics
Computer-integrated Manufacturing, 22, 239–55.
Trappey. J. C., T.-H. Lu & L.-D. Fu (2007). Development
of an intelligent agent system for collaborative mold
production with RFID technology, Journal of
Computer Integrated Manufacturing, 20, 5, 423–435.
Yu, C., & Wong, T.N. (2015). A multi-agent architecture
for multi-product supplier selection in consideration of
the synergy between products. International Journal of
Production Research, 53 (20), 6059-6082
Ghadimi, P., Toosi, F. G. & Heavey, C. (2018). A multi-
agent systems approach for sustainable supplier
selection and order allocation in a partnership supply
chain. European Journal of Operational Research, 269,
286–301.
Davidsson, P., Henesey, L., Ramstedt, L. (2005). Tornquist,
J., Fredrik Wernstedt, F., An analysis of agent-based
approaches to transport logistics. Transportation
Research Part C, 13, 255-271.
Martin, S., Ouelhadj, D., Beullens, P., Ozcan, E., Juan, A.
A., & Burke, E. K. (2016). A multi-agent based
cooperative approach to scheduling and routing.
A Multi-Agent based Decision Framework for Sustainable Supplier Selection, Order Allocation and Routing Problem
627
European Journal of Operational Research, 254(1),
169–178.
Reefke, H. & Sundaram, D. (2017). Key themes and
research opportunities in sustainable supply chain
management—Identification and evaluation. Omega,
66, 195–211.
Hofmann, E. & Rüsch, M. (2017). 4.0 and the current status
as well as future prospects on logistics. Computers in
Industry, 89, 23–34.
Govindan, K., Jafarian, A., & Nourbakhsh, V. (2015). Bi-
objective integrating sustainable order allocation and
sustainable supply chain network strategic design with
stochastic demand using a novel robust hybrid multi-
objective metaheuristic. Computers and Operations
Research, 62, 112-130.
Govindan, K, Darbari, J. D., Agarwal, V., Jha, P. C. (2017).
Fuzzy multi-objective approach for optimal selection of
suppliers and transportation decisions in an eco-
efficient closed loop supply chain network. Journal of
Cleaner Production, 165, 1598-1619.
Songhori, J.M., Tavana, M., Azadeh, A., & Khakbaz, M.H.
(2011). A supplier selection and order allocation model
with multiple transportation alternatives. International
Journal of Advanced Manufacturing Technology, 52,
365-376.
Kuo, R.J., Lee, L.Y., & Hu, T.-L. (2010). Developing a
supplier selection system through integrating fuzzy
AHP and fuzzy DEA: a case study on an auto lighting
system company in Taiwan. Production Planning and
Control, 21 (5), 468-484.
Govindan, K., Khodaverdi, R., & Jafarian, A. (2013). A
fuzzy multi criteria approach for measuring
sustainability performance of a supplier based on triple
bottom line approach. Journal of Cleaner Production,
47, 345e354
Laporte, G., & Osman, I. H. (1995). Routing problems: A
bibliography. Annals of Operations Research, 61, 227–
262.
Yong, W., Shuanglu, Z., Kevin, A., Jianxin, F., Maozeng,
X., & Yinhai, W. (2018). Economic and environmental
evaluations in the two-echelon collaborative multiple
centers vehicle routing optimization, Journal of
Cleaner Production, 197 (1), 443-461.
Nasiri, M. M., A. Rahbari, F. Werner, & R. Karimi. (2018).
Incorporating Supplier Selection and Order Allocation
into the Vehicle Routing and Multi-cross-dock
Scheduling Problem. International Journal of
Production Research, DOI: 10.1080/00207543.2018.
1471241
Lim, M. K., & Zhang, D. Z. (2004). An integrated agent-
based approach for responsive control of manufacturing
resources. Computers & Industrial Engineering, 46,
221–232.
L.C. Wang & S.K. Lin. (2009). A multi-agent based agile
manufacturing planning and control system. Computers
& Industrial Engineering, 57, 620-640.
Chen C., Lin C., Huang S. (2006). A fuzzy approach for
supplier evaluation and selection, International Journal
of Production Economics, 102, 289–301.
Hamdan, Sadeque & Cheaitou, A. (2017). Supplier
selection and order allocation with green criteria: an
MCDM and multi-objective optimization approach.
Computers and Operations Research, 81, 282-304.
LogiTrans 4.0 2019 - Special Session on Logistics and Transport in the Industry 4.0
628