A Hybrid Memetic and Set Partitioning Optimization Framework for Decision Support in Industrial Transportation: A Case Study of Employee Shuttle Routing

Designing cost-effective shuttle services for large-scale industrial companies presents a significant challenge in the transportation industry.This challenge arises from the need to balance high-quality service with cost-effectiveness while considering various practical constraints.In this context, we introduce a novel approach to help decision-makers address Employee Shuttle Bus Routing Problems (ESBRP).Our method combines the Memetic Algorithm (MA), a metaheuristic, with the Set Partitioning Problem (SPP) model, an exact algorithm.The proposed framework consists of two phases: (1) generating routes that adhere to the real-world constraints of the ESBRP using the MA, and (2) allocating these routes to a heterogeneous fleet of vehicles by optimally solving the SPP Model.A unique feature of our approach is the extension of the framework to enable the transition from addressing the single-load scenario of the ESBRP problem to solving the mixed-load scenario.This transition is achieved by implementing the Single to Mixed Loads Heuristic (SMH).This paper presents the results of thorough computational tests conducted on multiple data instances of varying sizes.Additionally, we develop a mixed-integer programming (MIP) model for the ESBRP to compare and evaluate the results of the proposed framework.By assessing solution quality and execution times on small and moderate-sized data instances, the experiments demonstrate that the proposed approach is efficient and often generates near-optimal solutions.