Bengaluru city police use AI to curb post-10 pm cracker bursts

Bengaluru: Bengaluru police have successfully used Artificial Intelligence (AI) to track and identify incidents of firecracker bursting across the city.

Over 2,000 such incidents were detected in five days during Deepavali, with 41 per cent occurring after 10 pm.

The initiative used the city’s extensive Safe City surveillance network, comprising cameras and Integrated Command and Control Centre infrastructure.

A Firecracker Detection Video AI app analysed live streams from over 200 cameras located in high-density areas such as Srirampuram, KR Market, HSR Layout, Haralur and Marathahalli.

"The AI algorithms were designed to detect firecracker flashes, smoke and crowd hyperactivity in real time, automatically generating alert packets containing location, timestamp and visual evidence. These alerts were sent instantly to the Command Control via WhatsApp, enabling police operators to verify events and coordinate immediate field responses," said Kuldeep Kumar Jain, Joint Commissioner of Police (Administration).

The areas were selected based on high population density and earlier complaints of large-scale cracker bursting after 10 pm.

The Supreme Court prohibits bursting crackers after 10 pm. Until now, police depended on complaints from the public to act against persons bursting crackers after 10 pm.

"The Hoysala teams would reach as and when complaints were received. With the AI module integrated with our cameras, the Hoysala teams get an alert as soon as such an incident is identified. This helps us to act quickly and prevents further bursting of crackers," Kuldeep added.

The initiative has helped police proactively prevent large-scale bursting of crackers after 10 pm and has significantly reduced noise and air pollution, police said.