Towards a Secure SCADA Environment: Intelligent Detection of Cyber Threats in Industrial Control Systems

Abstract

Supervisory Control and Data Acquisition (SCADA) systems form the core of industrial automation and infrastructure control, enabling centralized monitoring and management of critical processes. However, their increasing connectivity with open networks has exposed them to a range of cyber threats targeting data integrity and operational reliability. This research focuses on developing a machine learning-based intrusion detection model to identify anomalies in SCADA network behaviour. The model was trained and tested on the BATADAL dataset, which represents realistic water treatment plant operations. The approach demonstrated strong performance, achieving an overall accuracy of 99.88% and an F1-score of 0.9888 for attack detection. The results suggest that machine learning can effectively enhance the detection of cyber threats in SCADA environments with minimal false positives. The study contributes toward improving industrial control system resilience through intelligent, data-driven security monitoring while maintaining interpretability and operational practicality.