The use of discrete event simulation techniques to optimize a proposed factory layout.

Abstract

This project has proved conclusively that discrete event simulation techniques can be used to simulate, on computer, a complex stochastic materials handling system. The packing, automatic palletising and warehousing departments of a large powders manufacturing factory was used as an example to investigate the capabilities of computer simulation. The company intends to increase the number of packing machines from seven to eleven, and has embarked on productivity improvement projects that aims to increase the average packing efficiency from the current 50%, to 60% with a long term goal of 70%. Due to the stochastic nature of the run and stop durations of the packing machines, it was impossible to predict the effect of the increased throughput on the palletising system by conventional means. The system was modelled on computer using the SIMAN simulation language. Extensive research was initially carried out in order to determine the operating parameters of the system. The generation of cases from the packing machines in the program was verified against actual production runs. Various alternatives were analyzed to assist in decision making on the expansion of the palletising system in order to accommodate the increased throughput expected from the packing floor. The simulation was therefore used to increase the capacity of the automatic palletising system at minimal cost while meeting demands from the packing floor. It was established that the only capital expenditure required would be about R500 000 to increase the capacity of a palletiser and to provide a pallet conveyor to transport 40% of the pallets to direct despatch.