An Exercise for Teaching Transportation Problem Using Spatial Data

Abstract

We present an exercise for teaching the transportation problem using a mix of spatial and randomly generated data. It illustrates the potential of using qualitative and quantitative data and is suitable for undergraduate or introductory business school courses on operations research (OR), logistics, and supply chain management. It poses two challenges: (i) given the demand locations and volume, open a certain number of warehouses to ensure customer responsiveness and (ii) given those warehouses with capacity limits, determine an optimal distribution plan that minimizes the total distribution cost. This exercise is developed with the active participation of MBA students in an introductory OR course. The participants, attending the class online from different parts of India during the COVID-19 pandemic, helped generate realistic customer locations by sharing their location data. Visualizing this spatial data (after masking) in Google My Maps helps the students decide on suitable warehouse locations by considering the proximity to customers as well as diverse socioeconomic, political, and environmental factors. Then, using these warehouse and customer data, the optimal distribution plan is obtained by employing OpenSolver. Students appreciate the exposure—starting from data set generation to deriving an optimal solution—offered by this data-driven decision-making exercise.