Anomaly Detection in Digital Twins: Leveraging AI for Real-Time Insight
Erilda Muka, Emiliano Mankolli, Galia Ilieva Marinova · 2025
Digital twins provide real-time virtual replicas of physical systems, enabling predictive insights that enhance decision-making and operational efficiency. By integrating artificial intelligence (AI), these digital models can proactively detect and mitigate anomalies before they lead to system failures, reducing downtime and ensuring optimal performance. This paper explores the role of AI-driven anomaly detection in digital twins, emphasizing its significance in maintaining reliability and operational continuity. Various machine learning techniques, including clustering and outlier detection, are examined in the context of a hybrid approach for their effectiveness in analyzing the vast data streams generated by Digital Twins.