Robust optimization of a bi-objective closed-loop supply chain network for perishable goods considering queue system

https://doi.org/10.1016/j.cie.2019.07.018Get rights and content

Highlights

  • Modeling closed-loop supply chain for perishable products.

  • Considering queue system in closed-loop SCN.

  • Solving with robust optimization and benchmarking with recent works.

Abstract

Supply chain of the perishable goods such as food material, diary, drugs, and blood products has drawn attention recently due to their impact on human lives. Supplying adequate and healthy food materials, drugs, and blood products and their management has always been one of the major concerns for humans Therefore, collection- and distribution-management of them, require comprehensive and accurate management and planning. The current research trend is to design a supply chain network for the perishable goods while taking into account uncertainties associated with their corruption. Queuing paradigm was utilized for wait-time reduction in the distribution centers. The objective function considered for the proposed model includes two sub-objectives: 1. To minimize the total network costs, and 2. To minimize greenhouse gas emissions. Some model parameters such as (demand, operational cost, goods transportation cost, and permissible capacity of the goods distribution center) were considered as uncertain parameters in order to control these uncertain parameters, an optimization method was used. Finally, for solving a two-objective model, three multi-criteria decision-making methods, namely overall weighting method, and Torabi-Hassini method were employed. The results indicate that there is a significant difference between the mean of the first, second objective functions and computational time. The TOPSIS method was used to select the most efficient method. Torabi-Hassani method rendered to be more effective to finding a solution,

Introduction

Today, the competitive-business environment has led to increased collaboration between companies as part of the supply chain network. In this regards, designing the supply chain, given its effect on the efficiency and accountability of the chain, is one of the main strategic issues. In addition, attention to the environmental issues and economic benefits resulting from the revival activities have led to paying closer attention to the reverse supply chain. To this end, integration of the reverse and direct logistic network designs is of great highly important (Talaei, Moghaddam, Pishvaee, Bozorgi-Amiri, & Gholamnejad, 2016). Supply chain is a collection of connected organizations under flows of the material, information, and financial currents. These organizations may include the institutions that produce the raw materials and products, and supply services such as distribution, storing, wholesale, and retailing. In this spectrum, the ultimate costumers lie at the last level of the chain, and they should be considered as one of the members of these organizations (Amin & Zhang, 2013). This type of supply chain network is known as a forward supply chain. On the other hand, in the reverse supply chain, the goods being consumed by the end-user flow towards the manufacturing centers. The integration of the forward and reverse supply chains, leads to the closed loop supply chains (Zhalechian, Tavakkoli-Moghaddam, Zahiri, & Mohammadi, 2016). In other words, a closed-loop supply chain network consists of forward and reverse supply chains as an integrated entity (Ran, Chen, Wu, & Liu, 2016). Considering these definitions, it can be concluded that a supply chain management system should emphasize on the integration of the chain members via an optimization technique. Integration of the decisions is one of the main factors that can lead to reduction of the chain cost, increase of the profit, and last but not least, increased customer satisfaction. Since the role of the two components of locating the facility and optimal product flow allocation is clear in the lifetime of a supply chain, the integrity of these two elements will result in an efficient and effective supply chain (Kaya & Urek, 2016).

Many researchers in the food, drug and blood products supply chain have focused on supply chain management in recent years (Zarei, Fakhrzad, & Paghaleh, 2011). Supplying the healthy food, medicine and blood, especially those with a limited life expectancy such as perishable materials, is one of the concerns of the corporations involved in the chain management (Morganti & Gonzalez-Feliu, 2015). The dangers of perishable materials can appear in each stage of the supply chain, and it is necessary to determine the critical control points for obtaining information about materials, construction, expiration dates, etc., as well as providing it in a transparent manner for the participants and consumers of the supply chain (such as climatic conditions, seasonality of raw materials, corruption at a given time, and special storage conditions) lead to greater uncertainty. For example, it is cited in the literature that food industries supply chain is constantly evolving to address aforementioned-issues.

Changes in the food preferences, purchase habits, and people’s life style increase the world-wide need for facilitating preparation of the products. On the other hand, the population increase goes hand in hand with increasing demand for the fresh products and high value added products. In addition, given the variety of the products available in the market, the challenge of prediction and accountability to the customer demand gets more critical for the producers and retailers. Furthermore, for providing customer satisfaction, the producers and retailers should focus on the delivery time more rigorously. Thus, given significance of the supply chain issues and presence of related challenges, such as: Customer demand changes, delivery time, inventory shortage, etc., and considering that the perishable materials have more specific conditions compared to non-perishable food materials, the supply chain management for the perishable materials is of greater importance.

On the other hand, in recent years, due to advances in technology, strong competitive markets and hard-core customers, the focus has been shifted towards the management of inventory of perishable products. Perishable products are normally categorized into two general categories. The first category includes products that operate within their specified lifespan without any change in their operation, but they expire rapidly, quickly decay and exhaust the cycle. In other words, by reaching the expiration date, these products will be unnecessarily consumed. Products such as calendars and anniversaries are part of this category, with no change in the appearance of these types of products, ending their useful life, one-on-one, with corruption and misplacing, and in a non-usable way they will be. Other products, such as milk, yogurt and medicine, are also included in this category because they are unusable at the end of their lifetimes. The distinction between products such as calendars and milk is that (despite the common feature of these products that go away from the cycle of consumption), the former will become outdated rather than corrupt, where the latter can perish. The second category of perishable products includes goods that they corrupt over time, which means that they lose their edge over and before they reach the expiration date until they are completely corrupted at the time of expiration. Foods and fruits fall into this category. In this research, both types of corruptions are considered for commodities.

Considering the high cost of storing goods and consequently the importance of the inventory control system in the supply chain, one can claim the management of inventories as one of the important issues in trade. Undoubtedly, given the special circumstances of perishable goods, this issue can become even more important. In general, perishable goods are referred to as goods that lose their value over time, such as dairy products, fruits, vegetables, blood samples, chemicals, etc. In today’s competitive market, providing a proper way to better manage the demand for perishable items is considered necessary. Producers of these goods can collect crude goods from the retailer in order to meet environmental concerns and reduce system cost by returning them to the production line and repair them to resell these items. Therefore, the importance of commodity corruption has led to the study and modeling of a closed loop supply chain network for perishable goods.

The operational model of the supply chain is considered as a powerful tool that can be effective in understanding and studying the effect of supply chain on the performance and management of the tactical determinants as well as their relationships. Thus, an operational supply chain would comprise of strategic and tactical decisions to make. Strategic decisions are related to the long-term organizational decisions and cover the one to several years of span time. Overall, these include decision makings about the number and location of the facilities (locating), situation, and capacity of the product stores, and the material flow in the supply chain network (Costantino, Pellegrino, & Tauro, 2016). Tactical decisions are related to the mid-term organizational decision, and cover a time span of several months to a year. These decisions include purchasing and production, coordination of production and distribution, inventory policy, transportation strategy and discounts (Meepetchdee & Shah, 2007).

At the strategic level, decisions related to the configuration of the supply chain network, including the location and number of facilities, are taken, where at the tactical level, assuming that the supply chain network configuration has already been practiced, and at this stage, optimization of production, inventory, production and distribution coordination, transportation, determination of price level and so on are done. Given that the establishment, opening and closing of the facility is a time consuming and costly process, then, after making long-term decisions, it is not possible to change it in the short run. On the other hand, the tactical and operational decisions are determined following the strategic decisions; thus, configuration of the supply chain network is a constraint for decision making at operational and tactical levels (Özceylan, Paksoy, & Bektaş, 2014).

The novel concept proposed in this work is to achieve all the aforementioned objectives simultaneously. That is, to minimize the system cost and also the greenhouse gas emissions. To achieve these goals, the queuing theory approach is used. The queuing approach has various models depending on the way of entrance and service, including exponential and super-exponential models and group descriptions. (Hajipour, Rahmati, Pasandideh, & Niaki, 2014). Majority of the research work cited in literature uses exponential model for its simplicity. However, it may not realistically reflect on the true nature of the SM. In this problem, we propose a comprehensive model that encompasses a number of the existing models. For example, when using a queuing group; in most organizations the customers are assumed to be served in groups. If the number of customers in each group is considered as 1, the model reduces to an exponential model.

In designing the supply chain model, the parameters are often considered as uncertain (Yousefi-Babadi, Tavakkoli-Moghaddam, Bozorgi-Amiri, & Seifi, 2017). Considering the above-mentioned issues, integration of major strategic and tactical decisions including locating, inventory, and discount issues in a supply chain network can cause coherence among different levels of decision making at the supply chain network, and in the long term, it can lead to cost saving or increased profitability of the supply chain network (Kaya & Urek, 2016). The other case in the investigation of supply chain network is to pay attention to the reduction of customer wait time at the queue. Thus, proper management should be done on the inventory of the products at the store-house so that the customers do not wait a long time at the queue for receiving the products. In summary, it can be stated that if the store-house inventory is increased, the queue will be correspondingly reduced, and customer wait time will decrease, while it leads to increased network cost and therefore, reduced profit (Sarkar & Giri, 2018). Thus, considering this important issue known as the Jackson networks, it is imperative to take this into account when it comes to the design of the supply chain network.

Pertinent literature implies that various researchers have investigated and modeled the closed loop supply chain networks. Most of the works cited in the literature have investigated the strategic and tactical decisions simultaneously, and the main focus is mostly on optimization of the flow rate between the supply chain members. Furthermore, a few numbers of studies focus on modeling the decision-making on location, inventory, and exponential discount in the closed loop supply chain network. Through review of these papers it is found that there is no comprehensive decision making model for locating, inventory, and exponential discount at uncertainty conditions considering the Jackson network for reducing customer waiting time. Therefore, a robust optimization method for designing a two-objective closed loop supply chain network for the perishable goods would be essential for the queuing system.

The main objective of this paper is to develop policies that would lead to reduction of the cost of supply chain and a decrease in greenhouse gas emissions in a closed loop supply chain network subjected to uncertainty conditions. For a better reflection on reality, demand, operating cost, transmission cost, and capacity of distribution centers are considered as uncertain entities in the chain as well. Hence, the necessity of using a robust optimization method to control the uncertain parameters is one of the essential requirements of the paper.

Section snippets

Review of literature

Baird and Wilmsmeier investigated the factors facilitating transportation of perishable products. They aimed at determining the factors that increase the utilization of international transportation for fresh fish to Norway from Europe. Norway's experience in the aquaculture industry shows that multi-faceted shipping solutions need to be expanded and that long-haul railway connections should be made from a central hub in Europe. This can only be achieved with a balanced flow of products, if the

Definition of problem and modeling

Considering production of perishable products including food, dairy, drugs, and blood products and need for momentary production, distribution, and management of these goods and products, proper design of supply chain network for management in production and distribution, and proper transfer of these goods and products in the storehouse is paramount Hence, the producers should both consider decay time of the goods in one hand, and attempt for purchasing required raw material (discount) on the

Computational results

In this part of the paper, the solution to the optimization problem is provided and analyzed. To this end, first, a sample problem at very small size is designed, the model was solved using three solving methods of multi-objective optimization problems including comprehensive criterion method, weighted sum method, and Torabi-Hassini method. Furthermore, sensitivity analysis was conducted on several model parameters.

Conclusions

Current research addressed the design steps of a supply chain network for the perishable goods under the uncertain conditions such as: the decay in raw material. The queuing paradigm was used for reduction of customer wait-time in the distribution centers. The objective functions considered for this model include two contrasting objectives: 1. minimizing the total network cost, 2. Minimizing greenhouse gas emissions. Thus, first an uncertain model of the problem was designed and parameters of

References (32)

  • A. Yousefi-Babadi et al.

    Designing a reliable multi-objective queuing model of a petrochemical supply chain network under uncertainty: A case study

    Computers & Chemical Engineering

    (2017)
  • B. Zahiri et al.

    A multi-stage stochastic programming approach for blood supply chain planning

    Computers and Industrial Engineering

    (2018)
  • M. Zarei et al.

    Food supply chain leanness using a developed QFD model

    Journal of food engineering

    (2011)
  • M. Zhalechian et al.

    Sustainable design of a closed-loop location-routing-inventory supply chain network under mixed uncertainty

    Transportation Research Part E: Logistics and Transportation Review

    (2016)
  • S. Alavi et al.

    Integrated production, inventory, and location-allocation decisions in designing supply chain networks

    International Journal of Information Systems and Supply Chain Management (IJISSCM)

    (2016)
  • A. Ben-Tal et al.

    Robust convex optimization

    Mathematics of Operations Research

    (1998)
  • Cited by (0)

    View full text