Graph Neural Networks for PredictiveMaintenance in IoT Sensor Systems UsingDevice Telemetry Data

Abstract

Industrial IoT systems generate massive telemetry streams, requiring intelligent predictive maintenancemodels to detect failures early, reduce downtime, and improve operational reliability and safety. Traditional approachesemploy statistical analysis, sequence segmentation techniques, CNN-LSTM hybrids, and graph-based classificationmodels to capture spatial-temporal dependencies and identify abnormal device behaviour patterns. These methodstypically achieve high classification accuracy but often exhibit moderate RUL estimation performance, demonstratingstrong fault detection capability across industrial, energy, and smart infrastructure applications. However, static graphstructures, limited temporal attention, imbalanced fault distributions, and poor generalization under noisy conditionsrestrict robustness and real-world deployment scalability. This paper proposes a dynamic graph-based GAT-BiLSTMwith cross-attention and gated fusion, achieving 99.50% accuracy and superior RUL prediction stability under noisyconditions. The framework incorporates adaptive adjacency learning and multi-task optimization to enhance predictivemaintenance accuracy and robustness in IoT sensor networks.