Integrated optimization model for distribution network design: a case study of the clothing industry

• Author: Xifen Xu,Ding Zhang,Ming Liu
• DOI: http://doi.org/10.1111/itor.12628
• Date: 01 July 2019
• Published date: 01 July 2019

Intl. Trans. in Op. Res. 26 (2019) 1269–1292

DOI: 10.1111/itor.12628

INTERNATIONAL

TRANSACTIONS

IN OPERATIONAL

RESEARCH

Integrated optimization model for distribution network design:

a case study of the clothing industry

Ming Liua,XifenXu

aand Ding Zhangb

aSchool of Economics and Management, Nanjing University of Science and Technology, Nanjing 210094, P.R. China

bSchool of Business, State University of New York, Oswego, NY 13126, USA

E-mail: liuming@njust.edu.cn [Liu]; pelovely@126.com [Xu]; ding.zhang@oswego.edu [Zhang]

Received 26 January 2018; received in revised form 21 December 2018; accepted 2 January 2019

Abstract

Distribution network design affects a firm’s operating costs and its customer service level. In this paper,

we consider distribution network design for improving the closed-loop logistics in a clothing company. We

formulate the problemas a mixed-integer nonlinear programming model with an objective of minimizing the

annual operating costs. Our model simultaneously determines the optimal number of regional distribution

centers (RDCs), identifies location and relative size for each RDC, allocates each city distribution center to a

specific RDC, decides on supply ratio for each contracted plant, and specifies the annual operating costs and

service level for the best scenario, as well as other scenarios. Test results show that this marketing initiative

at the studied company can effectively reduce its annual operating costs. Although this paper is a specific

case study, it provides several managerial insights and modeling references for similar facility location and

allocation problems.

Keywords:distribution network design; facility location–allocation; mixed-integer nonlinear programming;optimization;

clothing industry

1. Introduction

As location and convenience have emerged as major factors in consumer decisions for products

and services, retailers have attempted to locate their stores everywhere to attract more consumers.

Since there are more and more retail stores, manufacturers should respond by redesigning their

distribution network or at least relocating their distribution centers to minimize the supply cost and

improve the service level. In practice, distribution network redesign aims to shape the structure of

logistics networks, determine the number of echelons and location of facilities where the product

is temporarily stored on its way to the end customers (Ambrosino and Scutella, 2005). A good

location can provide the manufacturer with strategic advantages that other competitors may find

difficult to overcome.

C

2019 The Authors.

International Transactionsin Operational Research C

2019 International Federation ofOperational Research Societies

Published by John Wiley & Sons Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main St, Malden, MA02148,

USA.

1270 M. Liu et al. / Intl. Trans. in Op. Res. 26 (2019) 1269–1292

Early scholars conducted studies on the relationship between the management of inventory,

location of facilities, and the determination of transportation policy simultaneously in a distribu-

tion network design (Mourits and Evers, 1995; Jayaraman, 1998; Chu and Chu, 2000). At that

time, location–allocation problems were formulated in practice and the corresponding models were

solved using different algorithms (Cunha and Sousa, 1999; Eusuff and Lansey, 2003; Farmani

et al., 2005). Recently, distribution network design has been studied together with hot keywords,

including energy policy, low carbon, and big data. For example, Yang et al. (2016) designed a

low-carbon city distribution network with resource deployment. Wang et al. (2018) considered how

to utilize big data to identify the right number of distribution centers and the right assignment of

customers to the opened distribution centers with the objective of minimizing the total operating

costs. Literature review on this topic can be found in Farahani et al. (2014) and Mangiaracina

et al. (2015). Although many scholars have conducted research on distribution network design, few

focus on how manufacturers redesign their distribution network in response to retailer location

networks.

According to the literature on closed-loop network design with location and allocation, early

scholars preferred to integrate the forward and reverse logistics networks to avoid the suboptimal

designs with respect to costs, service levels, and responsiveness (Lee and Dong, 2008; Pishvaee

et al., 2009, 2010). For example, Chen et al. (2015) presented an integrated closed-loop supplychain

model with location–allocation problem and product recycling decisions. In their study, cartridge

recycling in Hong Kong was used as the research background. Yuchi et al. (2018) examined a

location inventory routing problem in a closed-loop supply chain thatconsidered random demands,

as well as random returns, from customers. They developed a mixed-integer nonlinear model to

determine the locations of distribution centers and remanufacturing centers, the inventory level

of the system, the delivery routes from distribution centers to customers and from customers to

remanufacturing centers.Although these scholars have proved thatthe consideration of closed-loop

network is more effective for distribution network design, the optimization models they used are

deterministic and discrete, but not continuous.

In recent years, the problem of facility location–allocation on a continuous space has attracted

more attention due to the rapid economic growth. For instance, Pacheco et al. (2015) considered an

ambulance location problem in Tijuana-Mexico by using a continuous location model. Gokbayrak

and Kocaman (2017) proposed a mixed-integer quadratically constrained programming model.

The model determines the number and locations of the facilities to be opened, and assigns the

demand points to these facilities. Beheshtifar and Alimoahmmadi (2014) proposed a multiobjec-

tive optimization approach for location–allocation of clinics. Meier (2017) presented an improved

mixed-integer program for hub location problems with a stepwise cost function. Koo and Moon

(2018) developed a mixed-integer programming model to determine the timing of unit relocation

for continuous resupply, safe locations for support units, and delivery amounts. The objective is

to minimize total risk in wartime logistics that was represented by unsatisfied demand, hazard

at the support site, and the number of relocations. In particular, Shang et al. (2009) redesigned

distribution network for GlaxoSmithKline Company. The core problem was to address facility

location–allocation problem on a continuous space with an unknown number of potential sites.

However, they overlooked the reverse logistics.

Table 1 summarizes the relevant literature in recent years. As shown in this table, integer

programming, mixed-integer linear programming and mixed-integer nonlinear programming

C

2019 The Authors.

International Transactionsin Operational Research C

2019 International Federation ofOperational Research Societies