Robustness of supply chains

Abstract

In case of major disruptions, significant decrease in supply chain robustness value can be observed. Recently, many companies experienced it with the covid-19 pandemic, and it created big losses in companies’ revenues. One of the major reasons was these companies could not be supplied by their suppliers. Addressing this problem, in this study, the solutions are investigated to keep robustness value high in case of major disruptions by considering supplier selection problem. Real case of a food product supply across countries is simulated under uncertainty by developing the mathematical model. This model includes three steps which are plausible scenario creation, candidate design creation and evaluation & selection process. Firstly, plausible future scenarios are created by using Monte Carlo approach in order to estimate future minor and major disruptions. Disaster frequency data of suppliers’ zones are used. In addition, Covid-19 case, which can be considered as hazardous event, is created by using real active cases of suppliers’ zones. In second step, candidate designs are created. Candidate designs represents the selected suppliers for specific years and warehouse capacity. Designs are simulated by using randomly chosen plausible scenarios among all plausible scenarios, and candidate designs are determined by optimizing the cost and robustness values of designs. In last step, firstly performance measures are determined in order to compare candidate designs. Candidate designs are simulated by using all plausible scenarios and Covid-19 case scenario, and the performance measures are calculated. According to result values of performance measures, the best design is chosen. In this study, cost and robustness are chosen as performance measures. Two different candidate designs show the best performance for the plausible scenarios and Covid-19 case, respectively. Although increasing warehouse capacity shows good performance for plausible scenarios, having multiple suppliers shows better performance for Covid-19 case. Lastly, the combination of these two cases is simulated and possible strategies are discussed