CCTV integration, AI for policing: Gujarat Police announce ₹37-lakh hackathon

The Gujarat Police have announced what they describe as the country’s largest hackathon focused on CCTV integration and AI-based video analytics, aimed at developing technology for policing and public safety.

According to hackathon website under the state home department, the Gujarat Police invited students, researchers, freelancers, startups and companies to participate and develop innovative solutions for a robust CCTV surveillance mechanism in the state. A prize pool of ₹37 lakh has been announced, while the hackathon is scheduled for September 10-11 at iHub Gujarat in Ahmedabad.

Here are the four main objectives of the hackathon, as outlined by the Gujarat government:

• Integrate diverse CCTV systems into a unified platform.
• Cross-reference live video feeds with government databases.
• Use AI to identify persons, vehicles and events of interest.
• Generate real-time alerts to support faster and smarter law enforcement.

The Gujarat government has said different departments currently operate independent CCTV systems, with some using cloud-based storage and others relying on local storage infrastructure. The government intends to integrate these cameras into a unified video management and analytics ecosystem.

What is the challenge?

The main challenge is to develop a system that can scale to more than 80,000 CCTV cameras across Gujarat and enable AI-powered analysis of live feeds.

The system is expected to support capabilities such as real-time suspicious-activity detection, automatic number plate recognition (ANPR), vehicle tracking, watchlist matching and cross-camera search. The challenge will test these capabilities using live CCTV feeds and real-world policing scenarios.

How to register and what happens next?

Those willing to participate can register on the Sentinel Gujarat portal. After completing registration, participants will be able to access the details, resources and live camera feeds from about 50 geographically distributed cameras through the Resources page for technical evaluation, according to the website.

Participants are then required to submit their prototype, documentation and architecture details. The submission requirements include:

• An unlisted YouTube link, with video visibility set to “Unlisted”.
• A Google Drive or OneDrive link with access enabled for “Anyone with the link — Viewer”.
• A URL to their hosted platform, along with test login credentials for the screening committee, if applicable.
• A GitHub or GitLab repository link containing the platform source code or relevant components, if applicable.

The technical evaluation will require participants to integrate approximately 50 heterogeneous cameras, track a designated vehicle and provide real-time alerts and GIS visualisation. The system will also be expected to generate movement history and searchable events.

“All eligible submissions will first be assessed against the common evaluation areas. Bonus consideration may then be awarded for meaningful capabilities demonstrated beyond the mandatory requirements,” the site says.

What happens after submission?

The process covers choosing an integration model, building and testing the solution, submitting the prototype and planning its scale-up.

Participants will test their systems on about 50 CCTV feeds, tracking a designated vehicle and generating real-time alerts, GIS visualisation and searchable movement data. They must also submit a plan to scale the system to about 80,000 cameras across Gujarat.

Solutions will be assessed on the test case, architecture, analytics, scalability and proof-of-concept readiness, with bonus consideration and special recognition for innovative capabilities.

What is the schedule?

Registration opened on August 4 and the last date to apply is September 7. Shortlisting is also scheduled for September 7.

The hackathon is scheduled to take place over two days, September 10-11, with the results to be announced on the second day.