A note on “Stochastic unpaced line design: Review and further experimental results”

Abstract

Executive SummaryA 1985 study in this journal examined the “bowl phenomenon” and claimed to demonstrate that Hillier and Boling's original results 1967 were due to an assumption of unrealistically large variations in processing times. A closer look at these findings reveals that they are derived from a flawed experiment that cannot possibly serve to verify whether or not average task times at work stations on an assembly line should be selected in a bowl distribution. We offer further arguments in clarification of Hillier and Boling's original work based both on our own simulation experiments as well as recent studies by other authors. We also comment on the need to orient future efforts toward practical integration of the bowl phenomenon result with other findings that relate to improving the throughput of unpaced assembly lines.