Change-point visualization and variation analysis in a simple production line: a process mining application in manufacturing

In production systems, digital twins must be always aligned with the real system to guarantee an effective decision making process in a continuously changing environment. To allow the alignment, digital models can be updated with process mining techniques through data collected by sensors. This paper addresses the issue of detecting changes in the production system by analyzing data collected from sensors. Using raw collected data, a procedure is proposed to compute and plot relevant system measures that could help change identification. Simulation is used to test the effectiveness of the procedure in a realistic medium size production line.