Many-Objective Mixed-Model Parallel Assembly Line Balancing Utilizing Normal Workers, Disabled Workers, and Robots

Abstract

This research focuses on parallel assembly line balancing that involves the use of robotics, normal workers, and disabled workers. A parallel assembly line is a combination of more than two assembly lines in a parallel position to improve wasted time, excessive workstations, and redundancy of the labour force. When some process overlaps the time-defined for a certain process which could affect the overall productivity, such that this kind of arrangement would be significant for improving the efficiency of production outputs. Therefore, this paper aims at three objectives to provide a proper solution for balancing mixed-model parallel assembly lines. However, multiple objectives, known as NP-hard, may struggle in providing optimal solutions in a limited timeframe. As a result, a Multi-Objective Evolutionary Algorithm, the so-called “metaheuristic”, will be conducted to develop the quintessential result for each objective. The algorithm includes Multi-Objective Evolutionary Algorithm based on Decomposition (MOEA/D), Non-dominated Sorting Genetic Algorithm III (NSGA-III). All in all, the result demonstrates that NSGA-III performs considerably better than MOEA/D in almost every dimension.