MUMBAI, India, Aug. 12 -- Intellectual Property India has published a patent application (202641093796 A) filed by Vidyavardhaka College Of Engineering on August 03, 2026, for Ai And Iot-Based Smart Canine Detection And Adaptive Deterrent System.
Inventors include Prof. Kavyashree B; Kumarswamy M; Mukul M L; Navajeevan S; and Nischal Mayur Gowda D M.
The application for the patent was published on August 07, 2026, under issue no. 32/2026.
Abstract: The increasing number of stray and roaming dogs in urban and semi-urban areas leads to serious problems for the infrastructure, public health and public safety. This work introduces a smart adaptive dog detection and deterrent system using Internet of Things (IoT) and Edge Artificial Intelligence (Edge AI) to detect dogs in real time and take the required preventive measures. The proposed system implements a Mobile Net-SSD deep learning model on a Raspberry Pi 4 for fast edge-based object recognition. The model was trained on the COCO dataset and a custom dataset that included various environmental variables. The resulting model achieved an accuracy of 87.4%, precision of 89.2%, recall of 86.5%, and an average inference latency of 180ms. The system combines an adaptive multi-modal deterrence mechanism, which dynamically chooses deterrent responses based on detection confidence, repeated detections, and temporal conditions. The device uses IoT connectivity to send real- time notifications through the Telegram platform and combines optical detection with programmable deterrents through sound, light and water. Experimental evaluations under different environmental conditions demonstrate that the proposed adaptive approach provides better effectiveness and responsiveness than conventional static deterrence methods, making it a promising solution for intelligent urban safety and animal management applications.
Disclaimer: Curated by HT Syndication.
