Journal of Advances in Industrial Engineering
Volume:56 Issue: 2, Summer and Autumn 2022

In this paper, the problem of sustainable bi-level closed-loop supply chain network design under uncertainty is modelled and solved. The competition is between two supply chains (existing and new), on the selling price of new products in the forward supply chain and purchasing price of returned products in reverse logistics. Demand is price sensitive and the quantity of returned products is dependent on the price of the competitor. Dealing with uncertainty is possible here using fuzzy theory. The bi-level supply chain model is considered to be a leader-follower game. At the upper level, the leader optimizes decision variables of the network design and At the lower level, the follower deals with a non-linear problem with continuous variables. Then Benders decomposition approach is applied to solve the single-level problem. Finally, a case study in the battery industry and a numerical analysis are presented. Results show that increasing the forward elasticity coefficient entails a decline in supply chain profit and as the elasticity coefficient of the competitor increases, the demand becomes more sensitive to the competitive price, and the demand for the competitor declines, leading to a decrease in the rival's profit and an increase in the current supply chain's profit.

A model of a two-level single-producer multi-buyer supply chain (TSPMBSC) is focused on in this article with a single product made by the producer (or vendor) given to the buyers. The operational form of vendor managed inventory (VMI) is utilized by vendors and buyers. We assume the economic production quantity (EPQ) model used by the producer for inventory control with a limited production rate. Sales quantity and sales price are the parameters of each buyer as well as a certain production rate. Two objectives are considered for the model; the first objective is the maximization of channel profit while the second objective is the maximization of the production periods variances whereby the required storage space is minimized. Because of NP-hardness, the weighted sum multi-objective genetic algorithm (WSMOGA), the multi-objective particle swarm optimization algorithm (MOPSO) and the Non-dominated sorting genetic algorithm II (NSGA-II) are the three distinct heuristics embedded for tackling the problem. The instances of the considered problems with small, medium and large sizes are used to compare these heuristics. Considering the metrics of comparison, The MOPSO-based heuristic outperformed the other heuristics

Dynamic systems have always attracted much attention from researchers as a significant part of the various types of systems, and modeling of supply chain processes is considered as one of these due to the nature of its change over time, the volatility of customer demand is considered as one of these problems that have many effects on the system and its costs. In the present study, the SCOR Supply Chain is first modeled with the Dynamic Systems Approach (DSA) under specific parameters. We determine the control parameters of the studied policy using the DEA-SCOR model. We also Improvement the basic five-stage model to investigate models incorporating advanced demand information and evaluate the influence of demand variability on the system performance.  Then, the evaluation and ranking of the supply chain network of several distribution companies have been analyzed using the indicators of information sharing, based on the opinion of managers and experts familiar with the subject and by a combination of the data envelopment analysis (DEA)-SCOR and stochastic frontier analysis (SFA). By calculating total efficiency according to DEA-SCOR model in SCOR supply chain supply network of oil products Distribution Companies, Falavarjan Branch and Tehran Office Branch, have the highest performance, and the lowest performance is observed in Lordegan, Shahriar branch. According to the results of the SFA method, Tehran Office Branch and Isfahan Branch had the highest performances, and Shahriyar and Tiran branches had the lowest performances. The performance level calculated using the SFA method is approximately the same as the performance level calculated using the DEA method. The performance calculated by the DEA method is less than that calculated by the SFA method in some cases. The average calculated performance in the DEA method equals 0.80, and the average for the SFA method is 0.82. Given the inadequacy of indicators and the improvement of these indices at each stage, the calculated efficiency in each dynamic period gradually improves, and the average total performance in the dynamic period is 0.90.

Governments use online platforms to keep track of transactions in the supply chain (SC) of subsidized foods to prevent fraud. Although regular checks of warehouses and documents were conducted, current platforms failed to resolve the issue. Blockchain technology (BT) provides governments with the ability to access transparent and real-time data to address these challenges. In this paper, we examine the key challenges influencing the implementation of a BT platform for managing subsidized food products in Iran. The barriers appear to be interconnected. We present a model that integrates the Best-Worst method (BWM) for obtaining independent weights and the Weighted Influence Non-Linear Gauge System (WINGS) using a rescaling scheme for considering the interrelatedness between the criteria. Expert opinions and literature reviews are used to identify critical factors. According to the findings, the costs of implementing and maintaining the system, as well as the regular restructuring of government rules regarding the data to be collected, are the two main challenges of implementing this new technology. Moreover, there are concerns about the cooperation with downstream entities of SC, cultural differences among partners, and their knowledge level, which may affect the complexity of downstream implementation. The results of sensitivity analysis show that WINGS gives greater weight to factors that have more impact on others. Conversely, the weight of factors that are interwoven with other factors and factors that aren't influenced by other factors is reduced as compared to the independent relative importance obtained from BWM.

The health service network has problems such as a shortage of medical equipment and human resources. Due to the need for high expertise in supplying these facilities, this problem is much harder to be solved than other industrial ones. In the COVID-19 pandemic, maintaining tranquility in society is the most important factor. The tranquility is obtained by providing medical facilities in the health care network. Also, the COVID-19 pandemic imposes new restrictions on the network because of preventive guidelines. In this situation, the problem of resource allocation will become more sophisticated and will reduce system efficiency. In this paper, the problem of transferring hospital beds to patients infected by COVID-19 considering a predetermined capacity level is considered. To cope with these problems, a mixed-integer mathematical programming model is suggested. In addition, to consider the uncertainty in the demand of patients that occurs in the pandemic, the fuzzy programming approach is used. The suggested model is solved with the Benders decomposition algorithm (BDA) and applied for assigning beds in two samples. The results show that proper management of resources in crisis situations such as the COVID-19 outbreak is very effective. As a result, this issue causes to overcome pressure on medical staff and lack of hospital facilities, during pandemic conditions.

In this paper, emphasizing the real conditions prevailing in production industries, a new optimization model is developed in order to optimally schedule preventive maintenance and repair activities in a multi-component maintainable manufacturing system addressing a novel approach. It is assumed that failures or inspections are not only causes of stopping devices but also some other activities (non-failure stops) may interrupt the production process. The presented mathematical model utilizes these interruptions as opportunities to perform some maintenance activities. Failure interaction between components is also considered and the rate of failure of each component due to shocks from other components may be increased by a certain percentage. In addition to preventive maintenance and repairs, in the case of sudden failure of any component, corrective maintenance is implemented. Besides, the cost of stopping the system for performing maintenance activities is considered dependent on the duration of maintenance execution. Due to the complexity of the structure of the proposed model, the Genetic algorithm is adapted as the solution approach and its parameters are adjusted by the Taguchi method. A numerical example is solved and analyzed. Finally, a comparison between the exact method and the developed algorithm is provided to examine its efficiency and the impacts of the rise in problem sizes on the performance of the algorithm.