MUMBAI, India, Oct. 26 -- Intellectual Property India has published a patent application (202611100493 A) filed by Madan Mohan Malaviya University Of Technology on August 19, 2026, for Real-Time Multi-Modal Crowd-Pressure Forecasting And Automated Flow-Control System For Stampede Prevention.
Inventors include Dr. Ninni Singh; Dr. Rakesh Kumar; Dr. Rajkumar; Dr. Manish Kumar Gupta; Sanjay Kumar; Dr. Anuraj; and Dr. Amit Kumar Dwivedi.
The application for the patent was published on October 23, 2026, under issue no. 43/2026.
Abstract: A public-safety system forecasts dangerous crowd-pressure build-up and automatically controls pedestrian flow to prevent stampedes at mass gatherings. Distributed edge nodes, each combining a camera with an anonymised device-density counter, process imagery locally to derive a crowd-density map and velocity field and to compute a crowd-pressure / turbulence metric, discarding raw frames so that no personal identity is stored. A fog forecasting engine fuses metrics from all nodes into a spatiotemporal machine-learning model that predicts the crowd-pressure field 30 to 90 seconds ahead and classifies risk per zone. A controller then selects, by optimisation or reinforcement learning, a coordinated set of flow-control actions - opening alternate routes, metering entry at gates, updating dynamic signage and issuing zone-targeted public-address guidance - to relieve pressure before a crush occurs. The edge/fog architecture is resilient to poor connectivity, and an operator dashboard provides interpretable decision support with automatic escalation and post-event analytics.
Disclaimer: Curated by HT Syndication.
