This research examines the performance of CONWIP versus "push" workload control in a simple, balanced manufacturing flowline. Analytical models and simulation experiments are used to evaluate the tradeoffs between throughput and inventory performance. Tradeoff curves based on inflating the inventory level for the CONWIP system and the arrival rate for the "push" system are generated. As well, the variability of interarrival and processing times are considered as experimental factors. Results show that, contrary to what some previous studies have indicated, CONWIP efficiency is not inherently superior to "push" system efficiency. Instead, the release mechanism used for the "push" system has a significant impact on which system will perform better. Utilization levels and processing time variability also affect the relative performances.
S. T. Enns, Paul Rogers