Process Mining with Programmable Logic Controller Memory States

This paper presents novel contributions to the field of process mining in Programmable Logic Controllers (PLCs). One significant contribution is the proposal of a new event source for process mining, which can be used when PLC logs are absent or insufficient. This research also demonstrated the feasibility of converting the memory states of PLCs into event logs, which can be used to construct a process model of the PLC. This approach requires neither detailed knowledge about control logic nor source code of PLC. The paper also evaluates the use of process confrontation as a means of detecting abnormal or cyber-attacks on PLCs, which can provide an effective method for identifying abnormal behaviors and deviations from the normal process flow. Finally, the paper explores other optimization ideas for process mining in PLCs, provides useful insights, and offers potential solutions for extending research in this area.