Product-specific Process Time Estimation from Incomplete Point of Production Data for Mass Customization

Abstract To realize cyber-physical production system on mass customization, process time estimation of each product is critically important. However, actual process time is often unobvious because POP (Point of Production) data is incomplete for operational reasons for example process start/end time of multiple products are registered simultaneously. Therefore, new work combination analysis method was developed. In this method, process time is modeled by product-specific value and model parameters are identified through correlation analysis between total work time and worked products combination in arbitrary time period. This method has been applied to real process of custom-made electronic device production.