The Job Shop Scheduling Problem (JSSP) is a complex NP-hard optimization problem in which multiple jobs, each consisting of several operations with distinct processing routes and requirements, must be scheduled on a limited set of machines. Inefficient scheduling can lead to increased makespan, tardiness, and poor resource utilization. This study addresses these challenges by integrating classical dispatching rules with a meta-heuristic Genetic Algorithm (GA) to improve scheduling performance in a practical manufacturing context. A job shop model is developed using real production data obtained from an aluminum and glass manufacturing facility and implemented in LEKIN simulation software. Several dispatching rules, including First Come First Serve (FCFS), Shortest Processing Time (SPT), Longest Processing Time (LPT), Earliest Due Date (EDD), and Critical Ratio (CR), are applied to establish baseline performance measures such as makespan, mean flow time, and tardiness. Subsequently, a GA-based optimization model is implemented to generate improved production schedules by minimizing makespan and enhancing machine utilization. Comparative simulation results demonstrate that the GA-based approach outperforms traditional dispatching rules across key performance indicators, providing a more efficient and flexible scheduling solution. The findings confirm that combining dispatching-rule-based benchmarking with GA optimization offers an effective and scalable approach for improving operational performance in job shop manufacturing environments.

Keywords

Job Shop Scheduling, LEKIN Simulation, Genetic Algorithm, Dispatching Rules, Meta-Heuristics.