MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641090784 A) filed by Vellore Institute Of Technology on July 26, 2026, for Behavioural Entropy Engine For Cognitive Fatigue Estimation And Determination.
Inventors include R Sujatha; and Jaivardhan Singh Chauhan.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: “BEHAVIOURAL ENTROPY ENGINE FOR COGNITIVE FATIGUE ESTIMATION AND DETERMINATION.” A system and method for estimating cognitive fatigue using a deterministic, privacy-preserving behavioral entropy engine. An input capture daemon interfaces with an operating system to capture raw key-down events from a keyboard, recording millisecond timestamps and discarding key identity to preserve privacy. A statistical processing engine calculates inter-keystroke intervals (IKIs) from the timestamps, grouping them into sliding windows to compute windowed means and population variance. An entropy computation module maps the windowed IKIs into a discrete probability distribution across predefined temporal bins to calculate Shannon entropy, first-order entropy rate, and entropy acceleration. A temporal analysis module monitors temporal gaps between key-down events to classify session states and triggers a write operation to log session metrics locally upon detecting state transitions. The system operates locally without machine learning models or external sensors. [FIG. 1]
Disclaimer: Curated by HT Syndication.
