An operating simulation tool for modelling and managing a job shop system

This study analyses the machinery department of a major Italian company, which manufacturers equipment and plants for the food industry. At present, the machinery department is arranged as a pure job-shop and the firm adopts a push strategy: every week the accepted orders are released in the shop floor and sequenced accordingly to the Earliest Due Date dispatching rule, with no pre-emption. Yet, the resulting performance is not satisfactory: often the jobs are not completed within the planned week and work in progress accumulates in the shop floor increasing time and costs. Aiming to improve the current situation, especially in terms of Makespan minimization, the job shop has been reproduced and analysed using the simulation environment SIMUL8TM. Different dispatching rules have been assessed but, due to the extreme variability of the jobs (both in terms of times and routings), none of these rules performed much better than the other ones. It was then decided to turn the simulation into an operating tool that, depending on the current systems state, may support the production manager in selecting the most favourable dispatching rule. To this end, every week the simulation environment is preloaded with the current systems state and with the list of the incoming orders. Next, different dispatching rules can be selected and simulated, and for each rule, the queuing times, machines saturation and makespan are returned as performance. To validate the model, several tests based on historical data were carried on; obtained results demonstrated the superiority and the robustness of the proposed approach.