Production optimization using Discrete Event Simulation : Case study of Volvo Penta engine production line

Simulation allows decision-makers in modern industries analyse the outputs of specific systems, and predict future ones based on certain decision variables when in combination with optimization and Lean. This project studies the production line of boat engines in Volvo Penta at Vara (Sweden), carrying out different experiments based on a verified and validated model of the real production system using FACTS Analyzer in order to improve the key performance indicators whilst comparing the optimization performance of four genetic algorithms (NSGA-II, NSGA-III, PSO and DE) and extracting knowledge using data mining. Following the experimental design methodology, four main conclusions are obtained regarding the production line: simulation can be a successful tool to detect bottlenecks, optimization based on the resources such as the number of operators; forklifts and buffers can improve the outputs of the current system 22% just by rearranging the assets around the bottleneck and also helps predicts the behaviour of the system when these available resources are increased or decreased, when optimizing as they do not have the same performance it is important to cross-check the results of different algorithms to ensure the validity of the results, and lastly knowledge extraction can help decision[1]makers by providing sets of rules that selected solution areas of the optimization follow