Principles for Partial Resource Flexibility Decisions in Flow Shop Environments

Resource flexibility refers to the ability to dynamically reallocate resource units from one stage of a production process to another in response to shifting system bottlenecks. Recent research has demonstrated that substantial improvements in operational performance can be realized in both serial and parallel production environments through the effective utilization of resource flexibility. In these contexts the resource was assumed to exhibit complete flexibility, i.e., each resource unit can be assigned to any stage of the system. This research explores the extent to which the operational benefits associated with resource flexibility can be achieved in a flow shop environment using a partially flexible resource. Focusing on labor flexibility, we propose corresponding metrics for partial flexibility and formulate a model for flow shop scheduling with partial resource flexibility (FSPRF). We also present a branch-and-bound algorithm and a heuristic for FSPRF. And on the basis of a set of computational experiments, we suggest the importance of the distribution of flexibility on system performance, and characterize important attributes for those flexibility distributions that yield superior results. The conclusions drawn from this research provide significant insight into the management of flow shops with a work force that is cross-trained to achieve partial flexibility. Moreover, we adapt the principles developed by Jordan and Graves (1995) for partially flexible manufacturing plants to the flow shop scheduling environment, and we link these principles in a novel way to recent research on self-buffering flow lines.