PLAT: An Automated Fault and Behavioural Anomaly Detection Tool for PLC Controlled Manufacturing Systems

Abstract

Operational faults and behavioural anomalies associated with PLC control processes take place often in a manufacturing system. Real time identification of these operational faults and behavioural anomalies is necessary in the manufacturing industry. In this paper, we present an automated tool, called PLC Log-Data Analysis Tool (PLAT) that can detect them by using log-data records of the PLC signals. PLAT automatically creates a nominal model of the PLC control process and employs a novel hash table based indexing and searching scheme to satisfy those purposes. Our experiments show that PLAT is significantly fast, provides real time identification of operational faults and behavioural anomalies, and can execute within a small memory footprint. In addition, PLAT can easily handle a large manufacturing system with a reasonable computing configuration and can be installed in parallel to the data logging system to identify operational faults and behavioural anomalies effectively.

Implementing the Concept of Alarm Flexibility During Production

“…The proposed method provides options for suspending operations in the event of an impending system failure. However, the overall impact on the process whenever a fault occurs is sometimes ignored by the ATPs defined by various techniques [26] [31]. The loss of a valuable batch in progress is inevitable when a process line is tripped due to a fault.…”