Soft computing in optimizing assembly lines balancing

Abstract

As part of manufacturing systems, the assembly line has become one of the most valuable researches to accomplish the real world problems related to them. Many efforts have been made to seek the best techniques in optimizing assembly lines. Problem statement: Since it was published by Salveson in 1955, some methods and techniques have been developed based on mathematical modeling. In recent years, some researches in Assembly Line Balancing (ALB) have been conducted using Soft Computing (SC) approaches. However, there is no comprehensive survey studies conducted regarding the use of SC in ALB problems, which is became the aim of this study. Approach: This study reviewed published literatures and previous related works that applied SC in solving ALB problems. Main outcomes: This study looks into the suitability of SC approaches in several types of ALB problems. Furthermore, this study provides the classification of ALB problems that can facilitate distinguishing those problems as fields of research. Result: This study found that Genetic Algorithms (GAs) are predominantly applied to solve ALB problems compared to other SC approaches. This high suitability in ALB refers to GAs’ main characteristics that include its robustness and flexibility. These SC approaches have mostly been applied to simple ALB problems, which are not problems that are covered in a real complex manufacturing environment. Conclusion/Recommendations: This study recommends that future researches in ALB should be conducted with regard to other issues, beyond the simple ALB problems and more practical to the industries. Besides the advantages of GAs, there are still opportunities to use other SC approaches and the hybrid-systems among them that could increase the suitability of these approaches, especially for multi-objective ALB problems. This study also recommends that human involvement in ALB needs to be considered as a problem factor in ALB.