Content area
Abstract
As manufacturing processes grow more complex, efficient scheduling and resource coordination have become central challenges for operations managers across industries.
This dissertation addresses the Job Shop Scheduling Problem with Transportation (JSSPT), extending the traditional framework by incorporating realistic transportation constraints, specifically focusing on Automated Guided Vehicles (AGVs) with multi-load capacity. While most prior work assumes single-load vehicles, our study introduces and investigates a more practical scenario where the AGV fleet is limited and each vehicle can transport multiple jobs simultaneously.
We develop a Biased Random-Key Genetic Algorithm (BRKGA) tailored for JSSPT with multi-load AGVs and validate its performance on benchmark instances. Computational results demonstrate that multi-load AGVs lead to substantial improvements in system performance, notably reducing makespan by minimizing empty trips and better synchronizing transportation with production. We also analyze the trade-off between increasing fleet size and vehicle capacity, finding that a smaller fleet of higher-capacity AGVs can outperform a larger fleet of single-load vehicles, yielding additional benefits such as reduced floor space and lower operational costs.
Our findings underscore the importance of integrated scheduling, as the configuration and coordination of transportation resources directly influence system efficiency. This dissertation fills an important gap in the literature by formulating and solving the JSSPT with multi-load AGVs and provides practical insights for managers seeking to optimize operational processes.
