Resource aggregation issues and effects in mixed model assembly

In this paper we consider the problem of grouping machine and human resources in mixed model assembly process so that sequencing and scheduling decisions may be performed more efficiently. Resource aggregation can potentially and significantly improve production efficiency and throughput by balancing production rates, and by minimising resource deficiencies and idle time inefficiencies. This resource aggregation problem in particular involves determining how many groups to have and what number of machines and workers should be assigned to each. Not only is the number important in the aggregation but also the type (identity). Hence which specific machines and workers should be assigned to each group must also be answered. There are numerous factors that affect the aggregation. For example worker experience level and preferences, job processing requirements, current and future jobs and workloads, travelling distances and adjacency conditions. Very little theory however exists to indicate what a good resource aggregation is. Few if any experimental results exist to indicate what level of improvement is also possible. In this paper we therefore provide a mathematical framework for answering these questions that includes an analysis of the complexity of the problem. A general mathematical model is formulated that is suitable for any machine scheduling environment and for any combination of distinct and indistinct resources. A number of alternative measures of performance may be used as the objective function for this model. These include workload and experience balancing and minimal travelling time and distance objectives, although the choice of objective is very much dependant on the particular process being addressed. However, due to the complexity of the problem, significant numerical investigations are left as a source of continuing research.