Real-Time Anomaly Detection in Industrial IoT: AI-Based Predictive Maintenance
Ayushi Chopra, Jyoti Gupta, Vanshika Bhatia, Sheel Nidhi Tripathi, Jagriti Basera, Sakshi Aggarwal, Rinku Sethi, Raghav Trivedi · 2025
The increasing adoption of the Industrial Internet of Things (IIoT) has enabled real-time monitoring of industrial systems, yet unexpected failures remain a critical challenge. This paper presents an AI-based anomaly detection framework for real-time predictive maintenance in IIoT networks. The proposed approach leverages machine learning and deep learning techniques, including Isolation Forest, Autoencoders, and Long Short-Term Memory (LSTM) networks, to detect anomalies in sensor data streams. The system is evaluated using publicly available IIoT datasets, achieving an accuracy of 97.3%, an F1score of 96.9%, and a false positive rate of 1.9%. The hybrid LSTM-CNN model outperformed traditional methods, reducing detection latency to 120 milliseconds and optimizing predictive maintenance schedules. This indicates that real-time AI-based anomaly detection enhances predictive maintenance in the industrial environment.