Robustness evaluation of an IP-based cross-docking schedule using discrete-event simulation
Résumé
Integer programming (IP) models are powerful tools for operations planning, but may encounter problems when applied in a stochastic environment. We propose a methodology to evaluate the robustness of an IP model using discrete-event simulation. In our case, the IP model determines truck and pallet schedules for a cross-docking facility. A portion of the output from the IP model is used as input to a simulation model that represents the facility. The decisions regarding pallet transfers and truck departures are left to the simulation model to ensure flexibility. Random events are induced through the simulation in order to see how robust the IP-based schedules are against the uncertainties in: (i) the transfer time of pallets through the cross-dock, (ii) the unloading time of pallets from trucks, and (iii) the earliness or delay of the trucks arriving at the platform. The difference between the deterministic and the stochastic scenarios are evaluated regarding the number of pallets put into storage, truck docking and sojourn times. Robustness measures are proposed based on numerical experiments.