Memory of 2011 Sabarimala stampede drives development of AI-based crowd monitoring tool

#K Unnikrishnan
Devotees queue at Sabarimala as authorities step up crowd management, boost darshan flow, and expand facilities amid rising pilgrim arrivals. Photo: PTI
Devotees queue at Sabarimala as authorities step up crowd management, boost darshan flow, and expand facilities amid rising pilgrim arrivals. Photo: PTI

Kochi: A firsthand experience of the 2011 Pullumedu stampede has inspired a researcher to develop an AI-powered camera system designed to alert authorities and help prevent future crowd disasters. The system was granted a national patent last week.

One of the inventors, Dr K Ramesh Babu, Principal of Palakkad Government Engineering College, was present in Sabarimala during the Pullumedu tragedy and said the incident inspired the research. The other team members include Dr Preetha KG, Dr Saritha S, and Binu A, faculty members at the Rajagiri School of Engineering and Technology, Kakkanad.

Dr K Ramesh Babu, Dr Preetha KG, Binu A and Dr Saritha S

Early warning system

The technology monitors crowd behaviour in real time, identifying unusual patterns and sudden increases in crowd density. When required, the system can also perform AI-assisted face detection from submitted photographs to identify individuals moving within a crowd.

The system is built on an AI framework called NEXARIS (Next Gen Agenting Realtime Vision and Intelligence System for Large Scale Monitoring) and comprises a camera, computation unit, battery, power unit, and communication module. The high-speed computation unit continuously analyses video captured by the camera and flags any dangerous or alarming footage. These alerts are sent instantly to a central server, allowing officials to review the situation and intervene before a tragedy occurs.