Using Process Mining in Real-Time to Reduce the Number of Faulty Products

Process mining is a field of research whose tools can be used to extract useful hidden information about a process, from its execution log files. The current problem is that there is no solution available to track the formation of faulty products in real-time, both in time and space, to make it possible to reduce their number. The aim of this study is to find an effective solution for real-time analysis of manufacturing processes. The solution is considered to be effective if it helps to detect the error source points as soon as possible, and thus helping to eliminate them, it contributes in reducing the number of faulty products. Our previous solution, the "Time and Space Distribution Analysis" (TSDA), can analyze production data in time and space, but not in real-time. As a further development, we created the "Real-Time and Space Distribution Analysis" (RTSDA), which is capable of observing manufacturing process log data in real-time. It was implemented in software and tested with real process data. Real-time process mining can increase the productivity by quickening the detection process of the potential error source points, thus reducing the number of faulty products.