MUMBAI, India, June 30 -- Intellectual Property India has published a patent application (202521059837 A) filed by Parul University Parul Institute Of Engineering Technology; Dr. Jaimeel Manojbhai Shah; Anthony K. Ardiabah; and Naval Captain Kwama Osei on June 23, 2025, for A System And Method For Intelligent Deep Learning-Based Road Accident Prediction And Prevention For Road Safety.
Inventors include Dr. Jaimeel Manojbhai Shah; Anthony K. Ardiabah; and Naval Captain Kwama Osei.
The application for the patent was published on June 26, 2026, under issue no. 26/2026.
Abstract: The present invention relates to road safety and employs a sensor and imaging module featuring high-resolution CMOS cameras and radar sensors arranged for 360- degree real-time data capture. Captured data is preprocessed—resized, normalized, and denoised—to extract spatial features such as vehicle type, position, speed, and obstacle presence. A hybrid machine learning prediction engine, comprising a CNN module (optionally integrated with transformer-based modules) and an RNN module with LSTM units, analyzes both spatial and sequential data. An object detection module localizes and classifies hazardous objects, while a sensor integration and data fusion module compiles a multidimensional dataset. A SHAP-based explainability component and a communication and alert interface support risk prediction and provide real-time alerts via a central control unit, thereby enhancing predictive accuracy and overall road safety.
Disclaimer: Curated by HT Syndication.
