Early Indication System for Critical Situations in Job-shop Production

Abstract

Job-shop production can be very complex and extremely volatile, especially when part variability is high and orders require an increasingly large amount of processing steps. Achieving a manufacturing performance that is robust against volatilities is of great importance in such systems. We therefore develop an early indication system to detect potentially problematic situations (e.g. processing delays) in a job-shop and to forecast a set of key performance indicators with the aim of increasing the robustness of manufacturing performance by improved reaction possibilities. We compare the results predicted by the early indication system with a discrete-event simulation. Our findings show that 67.8% of the processes that developed issues in the simulated environment were detected by the system beforehand, while the schedule deviation forecasting had an error rate of 34.4%.