MUMBAI, India, Dec. 13 -- Intellectual Property India has published a patent application (202511100592 A) filed by National Institute Of Technology, New Delhi, on Oct. 17, for 'multi-agent edge-ai based predictive safety system for preventing lithium-ion battery fires and explosions.'
Inventor(s) include Balram Kasniya; Tirupathiraju Kanumuri; Vivek Shrivastava; and Surya Karan.
The application for the patent was published on Dec. 12, under issue no. 50/2025.
According to the abstract released by the Intellectual Property India: "The present invention relates to a multi-agent Edge-AI based predictive safety system (10) for lithium-ion batteries, which enables real-time cell-level detection. The system comprises several cell-level monitoring units (101) that measure voltage, current, and temperature, a CAN bus interface (106) for encrypted telemetry transmission, and a central edge-AI node (800) running a multi-agent safety network module (300). Each cell acts as an independent agent to perform time series anomaly detection, predict the RSOT, and classify safety states. The agents use distributed consensus for fault-tolerant cooperation, eliminating cloud dependencies and allowing real-time decisions. A safety actuator module (400) initiates fire suppression, isolation relays, and thermal control within 10 milliseconds. Additionally, it enables the implementation of federated learning for OTA model updates and ensures the safety of falling back to threshold-based BMS operation logic. It targets multiple markets, such as electric vehicles, drones, grid storage systems, aerospace, and industrial battery platforms."
Disclaimer: Curated by HT Syndication.
