Production improvement using Discrete Event Simulation : case study of Volvo Penta engine production line

Abstract: The global market forces manufacturers to minimise costs, meet production schedules, and improve productivity to stay competitive. The produced products require higher customisations, which are affecting the complexity of the production system. Companies seeking methods and tools to improve their production system. One of these is Discrete Event Simulation (DES), which can be used to analyse the production system without affecting the real system. The limitations of the systems, are often known as bottlenecks. The bottleneck limits the performance of the production system to reach its maximum efficiency. This study, investigates the use of simulation in manual assembly lines, at Volvo Penta in Vara. The production system produces over 300 variants of engines, divided into three product families. The addressed problem, is that manual work and difference in processing times for each variant cause variation, which in turn, leads to bottlenecks. The required data was collected at the company and a simulation model was developed. During the analysis of the current state model, the main bottleneck of the system was found, by analysing buffer allocation-, utilisation-, and shifting bottleneck detection charts. Based on the analysis, three scenarios were developed to minimise the waiting portion of the bottleneck of the system. In the third scenario, one operator was allocated to the bottleneck, which in turn, eliminated the waiting potion. Additional four scenarios were developed with the objective to increase the throughput (TH) of the system. However, the bottleneck moved after each additional scenario. Based on the analysis of the performed scenarios, three proposed improvements were identified. The first improvement was to reschedule one engine family from the production line. The second, was to dedicate one operator directly to the bottleneck and finally, the third, prioritise and relocate operators to certain areas in the production line. In conclusion, simulation can be used to analyse manual production lines. However, simulating variation caused by manual work, requires a high amount of data. Furthermore, the combination of the proposed improvements has increased the TH of the system by 12 %, in comparison to the current state results and similarly used existing resources.