A Simulation Study of Robotic Welding System with Parallel and Serial Processes in the Metal Fabrication Industry

This paper presents the usefulness of simulation in studying the impacts of system failures and delays on the output and cycle time of finished weldments produced by a robotic work cell having both serial and parallel processes. Due to multiple processes and overlapped activities, process mapping plays a significant role in building the model. The model replicates a non-terminating welding fabrication system with duplicate stochastic events caused by system failures and delays. A full factorial model is employed and analyzed to examine the main and interaction effects of five major types of system failures and delays via multiple regression analysis. The analysis derived from the full factorial model shows that material handling carrier delays have the most significant impact on the cycle time. This case study illustrates a modeling approach with system verification and validation revealing fundamental system design flaws which cause a significant loss of production.