A Flexible Logistics Distribution Hub Model considering Cost Weighted Time

Abstract

The delivery time of order has become an important fact for customers to evaluate logistics services. Due to the diverse and large quantities of orders in the background of electronic commerce, how to improve the flexibility of distribution hub and reduce the waiting time of customers becomes one of the most challenging questions for logistics companies. With this in mind, this paper proposes a new method of flexibility assessment in distribution hub by introducing cost weighted time (CWT). The advantages of supply hub operation mode in delivery flexibility are verified by the approach: the mode has pooling effects and uniform distribution characteristics; these traits can reduce overlapping delivery time to improve the flexibility in the case of two suppliers. Numerical examples show that the supply hub operation mode is more flexible than decentralized distribution operation mode in multidelivery cycles.1. IntroductionToday, as the development of electronic commerce drives the increase of consumer demand, improving the flexibility of distribution hub to reduce the order-to-delivery time has become the primary challenge for distribution centers [1]. Due to the high time and financial costs, order delivery is considered an important activity of logistics distribution centers. Some studies have shown that the delivery time accounted for more than half of logistics services time [2].The distribution hub is closely related to the efficiency of order delivery, so it has been a hot research topic in logistics researches for the last several decades, and scholars have used a variety of methods in their attempts to study distribution hub, including logistics hub location, optimization of logistics network, region partitioning, and supply chain performance evaluation of logistics hub. (i) Logistics hub location: Ebrahimi-zade et al. [3] used covering radius, a decision variable, to build new mixed integer model to hub covering problem of multiperiod single distribution hub. Mohammadi et al. [4] proposed a mixed integer programming model to reduce failure cost for disruption problem; a new hybrid metaheuristic algorithm is designed to solve and prove reliability of the model. Correia et al. [5] put forward a modeling framework to minimize the total expected transportation costs between logistics hubs for determining the best multiple allocation hub locations. Some scholars even used the geographical level and family choice to construct structuration criteria for sequential assessment or simultaneous assessment of hubs location [6]. (ii) Optimization of logistics network: Ghodratnama et al. [7] proposed a single distribution hub location model with the objective of total transportation and installation costs and also considered the effects of different modes of transport on greenhouse gas emissions. Yang et al. [8] used the new fuzzy random simulation technique and the multistart simulated annealing algorithm to optimize the intermodal hub-and-spoke model to obtain the minimum mixed traffic cost and travel time. (iii) Region partitioning: the purpose is to divide the hubs in a certain area into several zones according to some rules. According to priority, the multistage logistics distribution network is divided into two-echelon logistics distribution region [9]. Anwar et al. [10] used spectral theory based graph partition to divide the density peak graph and obtain the different subnetworks, and results show that the proposed method is superior to the existing approach based on normalization. Franceschetti et al. [11] divided the distribution hub network by customer density to obtain the best combination of distribution hub partitions and number of vehicles. (iv) Supply chain performance evaluation of logistics hub: Bichou and Gray [12] proposed an approach that evaluates hub efficiency by directing port strategy towards relevant value-added logistics activities. Then Bichou [13] designed an integrative framework including traders affiliated with terminals as integrative benchmark to measure logistics hub performance by conceptualising hub from a logistics and supply chain management (SCM) approach. However, the most logistics centers usually measure output and resource utilization, while the use of comprehensive performance indicators is rare. Some scholars have made breakthroughs: Barros and Dieke [14] enhanced the performance of the least efficient airports by using data envelopment analysis (DEA). Cheng and Wang [15] constructed a platform that can be used to assess the integrated supply chain activities of enterprises, infrastructure, and institutional stakeholders in logistics hub.To improve the efficiency of order distribution, some researchers studied the performance of logistics activities from the perspective of flexibility. Logistics flexibility is defined as an ability that reasonably uses a variety of resources in the storage, transportation, distribution, and other aspects to improve the response speed for customer’s demand [16–19]. In particular, Nigel [20] and Upton [21] have made pioneering works in the study of flexibility. The former defined the flexibility from three dimensions (range, time, and cost): (1) range, the state or behavior of systems; (2) time, the time consumption of transition between states; (3) cost, the cost of transition between states. Based on the study [20], Upton [21] proposed the uniformity of flexibility, which can keep the performances of the whole system consistent. Now, the research of logistics flexibility is mainly focused on the transportation planning and management, facilities management, and other logistics activities related to information processing, material transportation, inventory management, reverse logistics, tracking, and delivery [22]. In order to improve the manufacturing flexibility, Mansoornejad et al. [23] considered the product line configuration in the integration of supply chain. Papageorgiou [24] proposed an optimized supply chain model considering the effect of uncertainty on system flexibility. Even the two-stage stochastic programming method is used to improve the flexibility of logistics system [25, 26].Nevertheless, the existing researches have some limitations: (i) most of them defined the definition and condition of delivery flexibility or used delivery flexibility as one of the assessment indicators to solve other problems. For example, the impact of the manufacturer’s promise strategy on the flexibility of delivery was discussed in [27]. Li et al. [28] designed a pricing strategy to maintain a high delivery flexibility. Chen [29] made a path planning model to solve fuzzy flexible delivery and pickup problem. However, few studies have responded to the question: how to determine the delivery flexibility under known distribution center operation mode? (ii) There are some previous works which considered weight in the supply chain management; Li et al. [30] constructed the TOPSIS model based on entropy weight considering the short-term target, the design concept of strategic target, and the characteristics of synchronous supply chain. Kocaoğlu et al. [31] designed a TOPSIS evaluation model based on strategic objectives and business indicators and used analytic hierarchy process (AHP) to determine the weights of indicators. Ha et al. [32] focused on interdependencies among logistics hub performance measures and built a model to quantify the supply chain performance of hub by introducing weights of interdependent measures. In [33], the “slack time” was used to evaluate the flexibility of enterprises’ supply chain. However, they constructed the indicators of flexibility evaluation only taking into account the company strategy, external environment, competitors’ reflection, and so on. Compared with these macrofactors, the microfactors such as delivery cost and delivery time interval have been neglected; they are flawed for assessing the flexibility of the distribution center [20]. Therefore, in order to resolve the research gap, this paper proposes a new method of flexibility assessment in distribution hub considering cost weighted time. The cost weighted time introduces the delivery cost factor of order into slack time that can be more accurate to measure the effect of delivery cost on the delivery time. In multidelivery cycles, the numerical examples show that supply hub is more flexible than decentralized distribution operation mode.The rest of the paper is structured as follows. Section 2 introduces the proposed new method. Section 3 presents our numerical example and discussion of results. The conclusions are presented in Section 4.2. The Method of Flexibility Assessment in Distribution Hub2.1. Problem StatementSupply hub is also known as supplier hub, vendor-managed inventory (VMI) hub, which refers to the logistics distribution center located near the manufacturer and used to store the raw materials of supplier [34]. Unlike decentralized distribution mode, supply hub changes the relationship between manufacturers and suppliers from one-to-many to one-to-one, simplifying the operation processes [34]. Since the utilization of supply hub has already achieved good results in enterprise management practices [35–37], we validate the effectiveness of the new approach by comparing and analyzing the flexibility performance of decentralized distribution mode as Mode 1 and supply hub operation mode as Mode 2. In order to facilitate the study, we consider the simple supply chain structure of 1 manufacturer and 2 suppliers. Each distribution center of Mode 1 only delivers one raw material as shown in Figure 1. The hub of Mode 2 concentrates on two kinds of raw materials for management and distribution as shown in Figure 2.Figure 1: The decentralized distribution mode.Figure 2: The supply hub operation mode.In Mode 1, two suppliers of enterprises deliver raw materials to manufacturers through distribution center 1 and distribution center 2. The manufacture