Incorporating shifting bottleneck identification in assembly line balancing problem using an artificial immune system approach

The manufacturing industry has evolved in the past few years due to the competitive global economy where the performance of its assembly line operations is primarily dependent upon optimum resource utilization. The assembly line operations are balanced among the available resources to obtain an equal amount of workload to achieve optimum resource utilization, called the assembly line balancing (ALB) problem. Various approaches have been proposed to solve the ALB problem, which is broadly categorized as exact, heuristic, and meta-heuristic approaches. Although solving the ALB problem is crucial, a bottleneck may still occur over the next operation stages. By using problem-specific information (bottleneck identification), it is expected to improve the solution quality of the ALB problem. As such, the contribution of this study is the computational method, namely as the swarm of immune cells with bottleneck identification (SIC+) approach, where both the ALB and bottleneck identification problems are addressed. In addition to the flexible problem representation, the SIC+ approach is equipped with a discrete bottleneck simulator to simulate the bottleneck scenario and bottleneck-specific operators to redistribute the machine workload of the identified bottleneck machine. The approach was tested on 24 benchmark data sets of the ALB problem, and the impact of incorporating bottleneck identification was illustrated. The experimental results show that the proposed SIC+ approach has achieved a total of 66.12% optimal solution over all instances of the benchmark data sets and has been compared with approaches from the literature where high-quality solutions were statistically justified.