MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202521086902 A) filed by Pandit Deendayal Energy University on September 12, 2025, for Ai-Driven Fault Detection And Predictive Maintenance System For Smart Grids.
Inventors include Dr. Alok Jain; and Dr. Ravi Kant.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention relates to an artificial intelligence driven fault detection and predictive maintenance system for smart grids. The system is configured to continuously collect real-time electrical and operational data from loT-enabled sensors, SCADA systems and grid monitoring devices. The collected data includes voltage, current, temperature, equipment status, load variation and other grid performance parameters. The system processes the acquired data using artificial intelligence algorithms, including deep learning models, to identify abnormal operating conditions, classify electrical anomalies and predict potential equipment failures before they result in major faults or grid interruption. The invention further employs edge computing to perform critical fault analysis locally, thereby reducing latency and enabling rapid response to dynamic grid events. Historical fault records and real-time sensor inputs are used to train and refine predictive models for proactive maintenance planning. The system also includes computer software with data visualization, anomaly detection graphics and predictive maintenance indicators. A three- dimensional graphical representation of grid sections and fault-prone areas may be generated to assist rapid fault localization, maintenance dispatch and decision- making. The invention enables early detection of faults, reduces downtime, improves grid reliability, lowers operational and maintenance costs and supports uninterrupted power supply while being compatible with existing smart grid and SCADA infrastructure. Figure -1
Disclaimer: Curated by HT Syndication.
