Transfer batch sizing in trip-based material handling systems.

Trip-based material handling systems such as AGV systems, lift trucks, etc. are often designed with a given flow matrix (or FROM-TO chart) which indicates the number of loaded trips that the devices must perform per unit time between the workstations. In reality, the number of handling trips that would result from parts flow in a facility is dictated by the transfer batch size (TBS), i.e., the number of parts that are transferred from one workstation to the next in one trip. In the dissertation, we present analytical and simulation results aimed at determining optimal or near-optimal transfer batch sizes to minimize the expected total WIP in a manufacturing system. For that purpose, we develop an analytical model (i.e., the "WIP model") to estimate the expected total WIP in the system and show via enumeration and simulation that the WIP model performs well in determining "optimal" transfer batch sizes. We also show that "conventional wisdom" which dictates that the number of handling devices should simply be minimized is seriously flawed since our results indicate that the total expected WIP in the system can be reduced substantially by adding one or more devices and decreasing the TBSs accordingly. Furthermore, since the input parameters such as layout, part routings, part processing times, and part throughput requirements are subject to change, we perform sensitivity analyses to investigate how the above input parameters affect the expected total WIP in the system and the "optimal" transfer batch size. We also explore the effects of the empty device dispatching rule on the "optimal" transfer batch size and observe that different dispatching rules may yield different "optimal" transfer batch sizes. Finally, we present simulation results based on variable transfer batch sizes aimed at moving parts to the next station with minimal delay.PhDIndustrial and Operations EngineeringUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/104489/1/9527663.pdfDescription of 9527663.pdf : Restricted to UM users only