MUMBAI, India, June 26 -- Intellectual Property India has published a patent application (202621051445 A) filed by Dharmender Singh Yadav; Ruchita I Yadav; and Komal E Patil on April 22, 2026, for Ai Smart Driver Safety System.
Inventors include Dharmender Singh Yadav; Ruchita I Yadav; and Komal E Patil.
The application for the patent was published on June 19, 2026, under issue no. 25/2026.
Abstract: An AI-Based Smart Driver Safety System is designed to detect and mitigate dangerous driving conditions in real time using a multi-modal sensing and risk assessment framework. The system continuously captures the driver's facial image via a webcam and processes it using MTCNN face detection and a Keras Convolutional Neural Network to classify seven emotional states. Drowsiness is detected through frame-by-frame Haar cascade eye detection and Eye Aspect Ratio analysis. Voice aggression is assessed by measuring the RMS amplitude of microphone audio. Vehicle speed is obtained from a slider control or can be integrated with an OBD-II interface. All four signals are fed into a weighted risk scoring engine that produces a composite 0-100 risk score and a three-level classification: SAFE, MEDIUM RISK, or HIGH RISK. Upon HIGH RISK detection, the system triggers an audible alarm and dispatches an automated WhatsApp alert to a registered emergency contact via the Twilio API. If the driver continues to ignore warnings over four consecutive monitoring cycles, an Emergency Safety Mode activates, progressively reducing vehicle speed to 30 km/h and engaging seatbelt lock. The system provides a real-time CustomTkinter dashboard displaying live camera feed, metric cards, risk history chart, and event log. The invention is cost-effective, software-deployable without vehicle hardware modification, and represents a comprehensive advancement over single-modality driver monitoring systems described in prior art.
Disclaimer: Curated by HT Syndication.
