MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641091300 A) filed by Srm Institute Of Science And Technology, Ramapuram Campus on July 27, 2026, for A Systems-Theoretic Fall Prediction System Using Non-Deterministic Finite Automaton And Non-Deterministic Petri Net For Smart Assistive.
Inventors include Dr. R. Krishna Kumari; Dr. V. Padma; and Dr. K. Janaki.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: Abstract: The present invention relates to a fall prediction and detection system for elderly individuals using non-deterministic computing models. The system integrates a Non-deterministic Finite Automaton (NFA) with five risk states (Normal, Elevated Risk, Pre-Fall Warning, Fall Likely, Fall Detected) and a Non-deterministic Petri Net (NPN) for formal verification. The NFA uses a non-deterministic transition function to handle ambiguous sensor readings from a smart walking stick equipped with accelerometer, gyroscope, heart rate variability (HRV), blood sugar, SpO2, and pressure sensors. The NPN verifies critical properties: boundedness (safe net, maximum token count = 1), liveness (all transitions enabled), and deadlock freedom. Confidence levels for fall prediction are computed from event sequences, achieving 0.87 for three or more concurrent anomalies and 0.32 for isolated anomalies. The system uses thresholds from established medical literature (American Heart Association, WHO, ADA), requiring no large training datasets. Steady-state token distribution shows highest occupancy in intermediate risk states (S1 and S2), enabling early intervention. The invention provides a verifiable, interpretable, and data-efficient solution for fall detection in IoMT (Internet of Medical Things) environments, suitable for embedded deployment with low computational overhead.
Disclaimer: Curated by HT Syndication.
