The Fraunhofer Institute for Secure Information Technology realized a conceptual study to detect attacks and other anomalies by means of Machine Learning, and report those to production control based on monitoring of industrial plants.
- Contrast to conventional Intrusion Detection Systems, this version does not need any type of defined attack patterns
- The continuous evaluation of field bus-, sensor-, production and ERP-data is the basis of the anomaly detection
- Means of autonomous process of machine learning, the system has the capability to recognize the standard range and raises an alarm by leaving it
Director of Industrial Security Solutions
04:00PM - Day 1