A combined optimization and simulation model approach to enhance mixed model assembly line balancing in prefabricated house industry

The construction industry is shifting towards off-site construction for improved labor conditions, better quality and less environmental impact. The prefabrication of timber framed houses contains of several unique wall-elements, which are assembled on one or more assembly lines and then put together on site. To stay competitive with the conventional construction industry the processes need to be improved. The aim of this study is to generate new improved task configurations for the assembly line and test them for real-life application. This is done by developing a mixed model assembly line balancing algorithm for prefabricated houses to create the configurations. Subsequently, these configurations are implemented in a validated discrete event simulation model to test for real-life feasibility and compare the results based on four indicators, cycle time per station, waiting time per station, time in system per wall-element and produced wall-elements, with the current production plant. The results show that the optimized task configurations generate more balanced cycle times, less waiting times and a time in system improvement for each wall element of around 20%. However, due to the complex and varying properties of the wall elements queuing effects can be observed with discrete event simulation and the outcome compared with the real-life production can only be improved by 1,8%. This research approach demonstrates the significant potential that the combination of optimization and simulation offers to operations managers in mixed-model planning within prefabricated housebuilding. However, it also reveals the necessity for an integrated procedure of assembly line balancing and sequencing to fully utilize these possibilities.