MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641091796 A) filed by T C Swetha Priya; T Kruthika; A. Namratha Rao; P Hari Chandana; G Tanusha; and N Shiva Kumar on July 29, 2026, for A Mobile Centric Multi Modal Framework For Adaptive Alerting In Human Automation Takeovers.
Inventors include T C Swetha Priya; T Kruthika; A. Namratha Rao; P Hari Chandana; G Tanusha; and N Shiva Kumar.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: In the transition from fully manual to semi-autonomous driving, ensuring timely and reliable human intervention is _critical to safety. This paper Presents a._ novel sthartphone-based monitoring and alert framework specifically designed for Level 2-3 autonomous vehicles, where control may shift between vehicle automation and the driver Following a brief calibration phase, the system continuously captures driver behavior-including eye aspect 1.atio, blink frequency, head-pose dynamics, and inertial Motion—using the smartphone's front Camera and IMU sensor§. Lightweight, on-device models detect signs of drowsiness, distraction, or delayed responsiveness, triggering a tiered alert protocol: (1) subtle visual cuesfor mild inattention, (2) combined visual-audio prompts for moderate fatigue, and (3) urgent audio=visual-haptic alarms in critical takeover scenarios. Optional integration with wearable heart-rate variability data further enhances detection accuracy. Alerts are delivered through vehicle's.heads-up display or the smartphone interface when no HUD is available. A Figma-based simulation demonstrates the app's adaptive alert logic and interface flow. A webcam-based prototype was developed to validate the detection logic, achieving an accuracy of 88%, with a precision - Of 0,91 for the Alert class and 0.87 for the -Drowsy class. The adaptive alert Mechanism demonstrated effective escalation based on real-time driver state, and the Figrna-based simulation illustrated the planned mobile app interface and alert flow. By leveraging widely available mobile hardware, the proposed system provides a cost-effective, scalable safety layer that forms a critical bridge between automated vehicle functions and essential human control.
Disclaimer: Curated by HT Syndication.
