Assignment of cross-trained workers with learning effect

Abstract

This dissertation addresses the problem of the practical application of worker assignment among workers of highly varying skill levels. Generally, a heuristic method for the assembly line balancing problem is applied in garment industry, which is that the tasks or grouped tasks are first pre-determined based on a standard processing time and then are assigned to the workers. The process of grouping and assigning tasks without considering the skill level of the worker limits the quality of the solution. Consequently, an integrated approach to the assembly line balancing and worker assignment problem has been developed. Due to the nature of fashion industry, new designs and new styles are launched every season and the workers in the industry are continuously required to learn new processes. This dissertation concerns the skill level of workers of both constant skill levels and learning ability. MIP models for both problems were developed and the objective is to minimize the cycle time and makespan respectively. To find a solution, a heuristic was proposed and a lower bound and the upper bound were determined. The performances of the heuristics were tested by comparing the solution obtained to the optimal solution. The result of the comparison of the solution from the heuristic and in practical application confirms the disadvantage if the practical application is applied for workers of highly varying skill levels. The constant skill level of workers learning ability is not appropriate in situations where workers have differing learning slopes.