MUMBAI, India, Aug. 17 -- Intellectual Property India has published a patent application (202641096638 A) filed by Vellore Institute Of Technology on August 10, 2026, for Microprocessor-Based Electricity Theft Detection System Using Finite State Machine Modeling And Ai-Assisted Human Verification.
Inventors include Dr. V. Sivakumar; Shreemathi E; and Kirubakaran R D.
The application for the patent was published on August 14, 2026, under issue no. 33/2026.
Abstract: The present invention relates to a microprocessor-based electricity theft detection system and method utilizing Finite State Machine (FSM) modeling and AI-assisted human verification for identifying unauthorized electricity consumption. The system comprises a data acquisition module for real-time monitoring of voltage, current, power, and energy consumption parameters, a consumption modeling module for generating expected usage profiles, an anomaly detection module for identifying deviations from expected consumption, and an FSM-based decision module configured to classify operational conditions into normal, minor deviation, suspicious, and theft states. The system further includes an alert generation module, a human-assisted verification interface, and an adaptive learning module that continuously updates consumption models and detection parameters based on historical data and verified events. By combining real-time monitoring, structured computational analysis, explainable state-based classification, and adaptive learning, the invention improves theft detection accuracy, reduces false positives, and enhances operational reliability in electrical distribution networks.
Disclaimer: Curated by HT Syndication.
