Multi Objective Energy Efficient Jopshop Scheduling by Modified NSGA II

Energy conservation has become a key concern in job shop manufacturing, where machines consume considerable energy even during idle periods. Traditional scheduling methods often prioritize makespan while treating energy use, machine utilization, and WIP as separate goals, overlooking their interrelated effects. These objectives frequently conflict, with improvements in one area potentially leading to drawbacks in another. To address this challenge, the study proposes a multi-objective scheduling model that simultaneously optimizes makespan, energy consumption, WIP inventory, and machine underutilization. To solve this complex optimization problem, a Modified Non-Dominated Sorting Genetic Algorithm (MNSGA-II) is introduced. The algorithm incorporates a hybrid crossover strategy that blends Partially Matched Crossover (PMX) and Order Crossover (OX), enabling improved population diversity and convergence behavior. The performance of MNSGA-II is benchmarked against NSGA-II and SPEA2 across twelve Lawrence instances and a real-world automotive job shop case study. Evaluation metrics, including diversification matrix, mean ideal distance, and normalized objective values, are used to assess solution quality. Experimental results show that MNSGA-II consistently outperforms NSGA II and SPEA2 by generating superior Pareto-optimal solutions across key scheduling objectives. The algorithm optimally sequences jobs that support machine power-down during idle periods, effectively reducing energy consumption while also optimizing makespan, utilization and WIP across multiple solution fronts. In the automotive industry case study, MNSGA-II achieved improvements of 11% in makespan, 12% in energy consumption, 11% in machine underutilization, and 17% in WIP inventory compared to NSGA-II and SPEA2. By offering a range of Pareto-optimal solutions and interective dashboard, the framework enables production managers to make informed trade-offs based on specific goals, supporting more sustainable and efficient job shop operations.