Abstract
This paper illustrates the potential for the improved economic provision of tooling in flexible machining cells, through the use of a multi-stage algorithm for the allocation of jobs to machines and the efficient exchange of the associated tool kits. The tool-oriented manufacturing algorithm uses a combination of clustering techniques to identify groups of jobs which have some similarity of operational and hence tooling requirements. These clustered groups are allocated to a particular machining cell configuration with respect to the additional constraints of balancing the overall work load per machine and managing tool kit exchange under limited tool magazine capacity. The performance of tool-oriented strategies has been investigated using a computational study and indicates a consistent advantage compared to alternative work-oriented strategies. This is measured in terms of the quantity and frequency of tool exchange activities required over a given manufacturing period and the degree of necessary disturbance to the sequence of job throughput.