MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202641122762 A) filed by Mr. Shyam Kondeti; Aparna Sandesh Patil; Deepak Kumar Kushwaha; Bsh. Shayeez Ahamed; R. Arul Jose; Dr. A. S. Muthanantha Murugavel; Dr. Sumit Kushwaha; Dr. R. Monisha; Mrs. Raja Rajeswari M; Dr. Koppula Satish Kumar; Dr. V. Kameswara Rao; and Ms. Ch L N Deepika on October 13, 2026, for Ai-Powered Multimodal Risk Intelligence System For Real-Time Health, Environmental And Industrial Hazard Prediction Using Iot And Explainable Ai.
Inventors include Mr. Shyam Kondeti; Aparna Sandesh Patil; Deepak Kumar Kushwaha; Bsh. Shayeez Ahamed; R. Arul Jose; Dr. A. S. Muthanantha Murugavel; Dr. Sumit Kushwaha; Dr. R. Monisha; Mrs. Raja Rajeswari M; Dr. Koppula Satish Kumar; Dr. V. Kameswara Rao; and Ms. Ch L N Deepika.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: An artificial intelligence powered multimodal risk intelligence system is disclosed for predicting health, environmental, and industrial hazards in real time using Internet of Things devices and explainable artificial intelligence. The system comprises physiological sensors, environmental monitoring devices, industrial equipment sensors, imaging devices, an edge gateway, and a computing platform. Acquired observations are associated with source identifiers, timestamps, and monitoring context. A pre- processing module validates, synchronizes, normalizes, and assesses the quality of heterogeneous observations while retaining missing data indicators. A multimodal fusion module extracts complementary features and combines them according to their quality, availability, and learned relevance. A predictive analytics engine evaluates fused representations and preceding observation windows to generate hazard categories, risk estimates, and forecasting intervals, with confidence information where supported. An explanation module identifies contributing features and associates them with source measurements and observation periods, enabling users to inspect the evidence supporting each prediction. An alert management module prioritizes warnings according to severity, likelihood, forecast proximity, and confidence, consolidates related notifications, and supports acknowledgement and escalation. Processing is distributed between edge and central resources, with buffering during communication interruptions and controlled validation of model updates. A historical repository records observations, predictions, explanations, and response actions under configured access controls and retention policies. The system integrates physiological changes, environmental exposure, and equipment conditions within a coordinated analytical workflow. It supports anticipatory monitoring and interpretable warnings across healthcare settings, workplaces, industrial facilities, and surrounding environments, providing a technical platform for reviewing emerging hazards and initiating appropriate preventive responses by authorized personnel.
Disclaimer: Curated by HT Syndication.
