MUMBAI, India, Aug. 10 -- Intellectual Property India has published a patent application (202641089727 A) filed by Vignan'S Nirula Institute Of Technology And Science For Women on July 23, 2026, for Ai Emergency Response Assistant.
Inventors include Prathipati Silpa Chaitanya; Mearuga Prasanna Kumar; Appireddy Chennakesava Reddy; Gorikapudi Manidheer; Kakarla Pavani; Darla Siva Koteswari; Duddu Prasanna; and Rajana Ashritha.
The application for the patent was published on July 31, 2026, under issue no. 31/2026.
Abstract: The invention describes an intelligent deep learning framework designed to detect and coordinate response to emergencies, and it is known as the AI Emergency Response Assistant. Multimodal data collected from IoT-enabled devices, wearable devices, and communication tools are considered, including voice, text, image, videos, and sensor data. Hybrid deep learning models with the combination of CNN, LSTM, and Transformer-based architecture are used to identify and analyze linguistic, visual, auditory, and context features for accurate emergency incidents classification and severity rating. The system is implemented in a cloud computing environment and guarantees scalability and real-time monitoring while keeping the computational load at minimum levels for the end-user devices and emergency responder's device. Adaptive learning is introduced to keep updating the model according to new incident patterns and changing report patterns. From the experimental evaluation process, it can be concluded that the system proposed in the present invention has been able to realize substantial improvements in terms of precision, recall, and robustness in contrast with traditional detection methods that rely on manual techniques or a single modality approach. With its integration of multimodal analysis, hybrid deep learning models, and cloud computing-based implementation, the present invention creates a powerful framework for emergency management.
Disclaimer: Curated by HT Syndication.
