A multi-cycle and multi-echelon location-routing problem for integrated reverse logistics
| DOI | https://doi.org/10.1108/IMDS-01-2022-0015 |
| Published date | 21 June 2022 |
| Date | 21 June 2022 |
| Pages | 2237-2260 |
| Subject Matter | Information & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems |
| Author | Xiaofeng Xu,Wenzhi Liu,Mingyue Jiang,Ziru Lin |
A multi-cycle and multi-echelon
location-routing problem for
integrated reverse logistics
Xiaofeng Xu and Wenzhi Liu
School of Economics and Management, China University of Petroleum,
Qingdao, China
Mingyue Jiang
Shandong Provincial Science and Technology Department,
Innovation Development Institute, Jinan, China, and
Ziru Lin
School of Economics and Management, China University of Petroleum,
Qingdao, China
Abstract
Purpose–The rapid development of smart cities and green logistics has stimulated a lot of research on reverse
logistics, and the diversified data also provide the possibility of innovative research on location-routing
problem (LRP) under reverse logistics. The purpose of this paper is to use panel data to assist in the study of
multi-cycle and multi-echelon LRP in reverse logistics network (MCME-LRP-RLN), and thus reduce the cost of
enterprise facility location.
Design/methodology/approach –First, a negative utility objective function is generated based on panel
data and incorporated into a multi-cycle and multi-echelon location-routing model integrating reverse logistics.
After that, an improved algorithm named particle swarm optimization-multi-objective immune genetic
algorithm (PSO-MOIGA) is proposed to solve the model.
Findings –There is a paradox between the total cost of theenterprise and the negative social utility, which
meansthat it costs a certain amountof money to reduce the negativesocial utility. Firmscan first design an open-
loop logisticssystem to reduce cost, and at the same time,reduce negative social utilityby leasing facilities.
Practical implications –This study provides firms with more flexible location-routing options by dividing
them into multiple cycles, so they can choose the right option according to their development goals.
Originality/value –This research is a pioneering study of MCME-LRP-RLN problem and incorporates data
analysis techniques into operations research modeling. Later, the PSO algorithm was incorporated into the
crossover of MOIGA in order to solve the multi-objective large-scale problems, which improved the
convergence speed and performance of the algorithm. Finally, the results of the study provide some valuable
management recommendations for logistics planning.
Keywords Reverse logistics, PSO-MOIGA, LRP, Panel data
Paper type Research paper
1. Introduction
From the International Electrotechnical Commission (IEC) report, 66% of the world’s
population will live in urban areas by 2050 (https://www.iec.ch). As urban construction is
driven by the Internet of things (IoT), cloud computing and mobile Internet, which are the
typical representatives of new generation information technology, smart city has come into
being. The aims of smart city involve the improvement of spending and lifestyle of residents,
and the extension of applied range of modern information and communication technologies
(Ullah et al., 2020). During the formation of smart city, the terminal of logistics industry such
Location-
routing
problem
2237
Funding: This research was supported by the National Natural Science Foundation of China (Grant No.
71871222) and the China University of Petroleum Funds for “Philosophy and Social Sciences Young
Scholars Support Project”(Grant No. 20CX05002B).
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/0263-5577.htm
Received 10 January 2022
Revised 31 March 2022
Accepted 19 April 2022
Industrial Management & Data
Systems
Vol. 122 No. 10, 2022
pp. 2237-2260
© Emerald Publishing Limited
0263-5577
DOI 10.1108/IMDS-01-2022-0015
as online car, takeaway delivery and freight moving has gradually become an important part
of resident life. However, with the popularity of O2O business model and same-city services,
there are new features appeared in the logistics terminal distribution, such as point-to-point,
multi-batch and small-scale, which bring new challenges on response speed, flexibility and
timeliness (Wang and Zhao, 2021). In reality, the logistics terminal distribution is called the
“last mile”, which is the last link between manufactures and consumers affecting customer
satisfaction and operational efficiency of logistics firms, and the customer benefit and
operational cost of this link are not only related to the distribution service, but also limited by
the distribution network. Therefore, the location of distribution network facilities and service
planning become the core of terminal distribution optimization.
With the policy support for logistics industry and the prosperity of e-commerce, a
synergistic development situation of competition and cooperation but resource sharing
(infrastructure, storage facility, human resource, etc.) is gradually formed among firms, and
the logistics facility location is shifting from the traditional self-built and self-used mode to
the shared leasing mode. The traditional mode requires the owner to spend a long time and
high cost to build and operate facilities, versus while the shared mode requires flexible,
cyclical solutions for facility location to response fast-paced social demands. Compared to the
high cost and fixed nature of self-built facilities, a diverse and inexpensive means of leasing
facilities is more beneficial to the long-term development of the business. Most previous
studies on facility location problem (FLP) have gathered in sub-regional consideration of
transportation costs between a single distribution center and a single tier of customers, but in
fact FLP should consider the overall regional economic and social benefits of many-to-many,
forming a synergistic environment in where multiple distribution centers can radiate
throughout the whole distribution region. At the same time, the relationship between multi-
echelon facilities and social benefits should be simultaneously considered in order to make
decisions that best meet the current social development and enhance business benefits.
Especially after the outbreak of the Corona Virus Disease 2019, a host of reverse logistics
research has developed rapidly, focusing on medical waste recycling, emergency material
recovery and so on. Take the current medical waste recycling as an example, which is a multi-
cycle and multi-echelon open-loop reverse logistics network. The network involving hospital-
staging room and staging room-disposal center two levels, can divided into classification and
collection, transfer, temporary storage and disposal four processes. In addition, medical
waste during an epidemic is divided into two types of medical waste, ordinary medical waste
and epidemic medical waste, which forces the recycling work to be divided into multiple
cycles. There are problems in this logistics network such as high change costs, uneven
capacity of disposal centers, and high and polluting waste volume. The above mentioned
problems drive us to design efficient route planning and location selection.
Therefore, multi-cycle and multi-echelon location-routing problem in reverse logistics network
(MCME-LRP-RLN) has important research value in the field of logistics and supply chain
management. At the same time, reverse logistics is highly responsive to the green transformation
and intensive construction of logistics. The optimization objectives such as negative social utility
and social responsibility often exist simultaneouslyalongwithreverse logistics, and it is
necessary to consider the optimization cost objectives and negative utility objectives when
solving LRP. Some of the current research problems faced by MCME-LRP-RLN are as follows:
RQ1. How to use case-related data to generate scientific and objective negative utility
objective functions to replace the artificially set objective functions in the big data
environment?
RQ2. How can we transform the firm LRP from a single-cycle decision problem to a multi-
cycle sequential decision problem on a multi-echelon basis in response to the
emergence of shared resource approaches such as facility leasing?
IMDS
122,10
2238
To continue reading
Request your trial