Collaborative logistics pickup and delivery problem with eco-packages based on time–space network

Highlights

• Collaborative logistics pickup and delivery problem with eco-packages is studied.

• A multi-objective mixed-integer programming model is developed based on TS network.

• A hybrid method combining multidimensional K-means, DP and RP-NSGA-II is devised.

• The collaborative profit and alliance stability are tested under different methods.

• Limited truck storage space sometimes undermines logistics operation efficiency.

Abstract

Collaboration among logistics companies offers a simple and effective way of increasing logistics operation efficiency. This study designs an optimal collaboration strategy by solving the collaborative logistics pickup and delivery problem with eco-packages (CLPDPE). This problem seeks to minimize the total operational costs by forming collaborative alliances and allocating trucking resources based on time–space (TS) network properties. The synchronization of two-echelon logistics networks is improved by solving this problem. Moreover, this study considers the stability of collaboration (i.e., the willingness of logistics companies to join and remain in collaborative alliances) by comparing different profit allocation strategies in the CLPDPE solving process. A novel methodology that combines multi-objective mixed integer programming, multidimensional K-means clustering, reference point based non-dominated sorting genetic algorithm-II (RP-NSGA-II), forward dynamic programming and improved Shapley value method is developed to formulate and solve CLPDPE. Our results show that the proposed algorithm outperforms most other algorithms in minimizing the total cost, waiting time and number of vehicles. An empirical case study in Chongqing city, China suggests that the proposed collaborative mechanism and transportation resource sharing strategy based on TS network can reduce cost, improve distribution efficiency, and contribute to efficient, smart, intelligent and sustainable urban logistics and transportation systems.

Introduction

The development of convenient online shopping and the resulting low cost of products have caused the rapid growth of online consumption, thereby generating a sharp increase in pickup demand (PD) and delivery demands (DD) that must be met by logistics facilities (Wang et al., 2018, Peng et al., 2019). In a traditional pickup and delivery network, conventional paper packages made from wood pulp are used as transport carriers to deliver products from suppliers to customers. However, most traditional packages are discarded directly without being recycled, thereby causing a substantial waste of resources. At present, green sustainability concepts and sharing modes are highly regarded. Thus, eco-packages have become the new preference of logistics companies to comply with the requirements of green sustainability and to reduce packaging costs.

In comparison with conventional packages, eco-packages are a type of commodity transport carrier that can be recycled many times. Although the cost of manufacturing a single eco-package is relatively high, multiple recycling greatly reduces the packaging cost and environmental damage because less raw material from wood is cut and manufactured. Statistics from the State Post Bureau of China (SPBC, 2018) indicates that the 2018 “Double 11” Shopping Festival generated a demand for more than one billion packages, resulting in the wastage of over 50,000 tons of raw paper because of using traditional paper packages. By contrast, ShunFeng Express uses recyclable eco-packages, called Feng Boxes. A total of 10 million Feng Boxes can replace approximately 500 million traditional packages, 1.4 billion m of tape, and 2.25 million cu. m. of filling. Moreover, this material effectively solves the problems of high cost, damage, low operation efficiency, and resource waste caused by the non-recyclable nature of traditional packages because each eco-package is relatively well made and can be used more than 100 times (Cruz et al., 2012).

However, an important challenge in fully promoting eco-packages is the efficiency of pickups and deliveries (Hao et al., 2019). The conventional multi-depot vehicle routing with pickups and deliveries (MDVRPPD) for operational cost optimization (Baldacci et al., 2011) is inapplicable to the situation with eco-packages, due to the requirement for collaboration and the time window and TS network constraints in optimizing eco-package pickups and deliveries (Wang, Peng et al., 2018). Moreover, the relatively independent operation mode among logistics facilities causes the eco-packages to undergo a long recycling process because the package can only be recovered by the initial delivery company (Yu et al., 2017). Therefore, optimizing the pickup and delivery network of eco-packages collaboratively is crucial. Collaboration among logistics facilities not only greatly reduces the overall operational costs of the logistics network (Dai and Chen, 2012, Fernández et al., 2018), but also improves the synchronization for the pickups and deliveries of eco-packages effectively; for example, the waiting time is reduced when vehicles collect eco-packages at facilities or deliver eco-packages to customers (Wang, Peng et al., 2018). However, establishing a collaboration mechanism (i.e., how to motivate logistics companies to collaborate) is costly, and compensation from the government and a fair profit allocation are important factors in building an acceptable collaborative alliance (Wang et al., 2017, Wang et al., 2018). To clarify the specific research of this paper, Fig. 1 shows the optimization mode of the collaborative logistics pickup and delivery problem with eco-packages (CLPDPE).

In Fig. 1, the implementation of the eco-packages’ pickups and deliveries is presented by the shared trucks and facilities in a two-echelon collaborative logistics network. For the feasibility and convenience, eco-packages are classified as large eco-packages (L-EPs) and small eco-packages (S-EPs) for pickups and deliveries between first and second echelons. L-EPs/S-EPs are transformed at the delivery satellite or pickup satellite. For example, a L-EP can be split into several S-EPs at the delivery satellite and several S-EPs can be merged into a L-EP at the pickup satellite. At the first echelon, a shared truck carries four L-EPs consisting of multiple S-EPs and departs from the logistics delivery center (LDC) with the state of [1, 1, 1, 1]. The truck then unloads the L-EPs at the delivery satellites with the state of [0, 0, 0, 0] and drives to the pickup satellites in the collaborative alliance to collect large eco-packages. When the shared truck is fully loaded with the state of [1, 1, 1, 1], it returns to the logistics pickup center (LPC) to complete the pickup of L-EPs. At the second echelon, the L-EPs that are delivered to the delivery satellite by the shared truck are split into multiple S-EPs, which are then delivered to customers via a shared vehicle. Similarly, the L-EPs collected by the shared truck at the pickup satellite come from the merger of the S-EPs collected by the shared vehicles from the customers.

Therefore, our research focuses on achieving efficient pickups and deliveries of eco-packages while minimizing cost in the collaborative two-echelon logistics networks. Existing studies have contributed substantially toward the optimization of a multi-echelon logistics network, but exhibited four deficiencies or aspects that need further exploration. First, the optimization of multi-echelon multi-center pickup and delivery problem (MEMCPDP) has been widely investigated. However, the lack of a collaborative mechanism has caused the network optimization to be conducted only within the facility or the enterprise itself rather than across all the logistics facilities. Second, previous studies on MEMCPDP emphasized that network structure optimization, including routing optimization, reduced operational costs but ignored the effect of recycling packaging materials on cost optimization and resource sharing. Third, in previous studies, the optimization objectives mostly focus on economic cost reduction and considered the time cost of waiting insufficiently. Thus, the timeliness of pickup and delivery is not guaranteed. Fourth, heuristic algorithms are widely used to solve the optimization problem of large-scale logistics networks. However, the advantages of combining exact and heuristic algorithms in the optimization of multi-echelon pickup and delivery problems should be further studied on the basis of the time–space (TS) network.

In compared with the previous research and combined with the actual logistics operation process, this study presents the following innovative contributions to theory and practice. First, the collaboration mechanism is introduced into the traditional MEMCPDP to establish the collaborative logistics pickup and delivery problem (CLPDP), which is an improvement from the local optimal networks under independent operations to the global optimal networks by establishing collaborative alliances. Second, in addition to network routing optimization, the recyclable eco-packages that replace traditional paper packages and the additional enhancements from eco-package pickup and delivery decrease costs. Third, the performance of a mixed integer programming model based on the TS networks in terms of satisfying the customers’ requirements for service timeliness while reducing the total network cost and improving resource utilization is examined. Fourth, in cases where the first and second echelon networks simultaneously consider routing optimization, the combination of the time-dependent forwarding dynamic programming algorithm based on the TS network and reference point-based non-dominated sorting algorithm-II (RP-NSGA-II) is used to achieve the Pareto optimal solution search under multi-objective requirements. In addition, the state of trucks/vehicles can be intelligently handled with L-EPs’ or S-EPs’ pickups and deliveries, and the entire logistics network can be iteratively optimized with the help of intelligent algorithms combined with exact and heuristic algorithms. Therefore, studying the CLPDPE is conducive to providing theoretical support for managing the operation of collaborative logistics pickup and delivery network, and then propelling the sustainable development of urban logistics and intelligent transportation systems.

The reminder of this article is structured as follows. The studies related to CLPDP with eco-packages (CLPDPE) based on the TS network are reviewed in Section 2. Then, relevant descriptions and theoretical proofs of the research problems are illustrated in Section 3. Meanwhile, modeling formulation based on the research problem and algorithms for solving the constructed model are described in Section 4 and Section 5, respectively. In addition, the improved Shapley value method for achieving fair allocation of the profits is described in Section 5. Subsequently, a practical case in Chongqing is used to confirm the validity of the proposed model and designed algorithms in Section 6. Finally, conclusions and potential research directions are presented in Section 7.