A COST BASED GROUPING HEURISTIC IN GROUP TECHNOLOGY

Abstract

This thesis proposes a cost based grouping procedure for Group Technology. The procedure is a two-stage graph partitioning approach to determine machine cells. Costs of work-in-process inventory, machine depreciation, machine setup and material handling are used as a basis for grouping. The procedure begins by selecting economic lot sizes for each part type and converting these values to machine requirements. The first stage assigns parts to machines of a common type. The second stage groups individual machines to cells, with each machine cell having a number of machines between chosen lower and upper limits. An additional check is given to the final groupings to see if adding a machine of any type can result in savings on the intergroup material handling cost. The process may iterate by next obtaining updated setup times and restarting.