A cost minimization supply chain design model considering pandemic uncertainty and infection trends

• Main Authors: Antonio, Fernando V., Atayde, Juan L., Yamzon, Martin Aloysius R.
• Format: text
• Language: English
• Published: Animo Repository 2022
• Subjects: Business logistics; COVID-19 Pandemic, 2020—-Influence; Industrial Engineering
• Online Access: https://animorepository.dlsu.edu.ph/etdb_induseng/9; https://animorepository.dlsu.edu.ph/cgi/viewcontent.cgi?article=1014&context=etdb_induseng
• Institution: De La Salle University

Supply chain design has been a widely used area of research especially given how companies all around the world have shifted to focusing on improving its systems. In the past years however the focus has been on creating lean supply chains in order to reduce costs, but due to the uncertainty brought by the COVID-19 pandemic it has challenged global supply chains with constraints on labor and capacity. Supply chain leaders now are focusing on creating a balance between agile and lean systems because of this. One of the largest disruptions a supply chain faces is preparing for the various restriction levels, and it is critical to prepare for this uncertainty. To better utilize readily available data such as infection counts on a daily basis, a mixed integer nonlinear programming (MILP) mathematical model for supply chain design, with the objective of minimizing set up & allocation costs. Involving major constraints on expected allocation, demand and capacity as restriction levels change. Given that the majority of uncertainties a pandemic brings hinges on the restriction levels local governments may employ, it has been seen that restriction levels may rely heavily on the most recent infection trend of the area, hence the constant change of restriction level a supply chain has to navigate through. The uncertainties brought about by restriction levels are demand, capacity, costs, and facility connections, to name a few. These all have been tackled in the past, however, the unique feature of this model involves the input of infection trends; with this data, restriction level probabilities are then calculated and are used when making decisions on the supply chain design. The model was run through various scenarios wherein different infection trend types were inputted to test model behavior and variable relationships. The infection trends tested were: steady increase, steady decrease, sharp increase, sharp decrease and steady. These were also tested with various starting restriction levels to encapsulate trends- types and expected behaviors. The model yielded results as expected, wherein it would hedge itself for whenever infection trend was increasing, this involved selecting additional facilities with low setup costs but higher operating costs. These were oftentimes selected as insurance if a capacity for the main facilities would decline or if connections were to break between the cheaper operating facilities and also if demand were to abruptly spike. A major finding also showed that the model formulation allowed for robust solutions in case of a change in restriction level were to suddenly occur. For future research on Supply Chain Design considering pandemic-like scenarios it is recommended that a wider range of factors and indices are to be considered in order to yield results that could be applied in the industry. Another area of improvement as well would be to add a service level objective into the model that way the optimum balance between service level and costs may be achieved. Incorporating these changes will not only make the model more dynamic and accurate.