Archive · 4 Sep 2026
AI Hackathon Produces Three Prototypes Targeting Bengaluru’s Traffic Challenges
Flipkart held a competition to find better ways to manage Bengaluru’s traffic.
More than 33,000 people took part.
They created over 1,100 computer-based prototypes.
Three teams were chosen as winners.
Gridlock Oracle helps predict when one traffic problem could cause larger traffic jams.
PRAHAR suggests where traffic officers and equipment should be sent.
ParkSight studies illegal parking to find places where parking causes congestion.
The teams used traffic information from the Bengaluru Traffic Police’s ASTraM platform.
Their ideas could help officials respond to traffic problems more quickly.
Flipkart selected Gridlock Oracle, PRAHAR and ParkSight as winners of its Gridlock Hackathon 2.0.
More than 33,000 participants created over 1,100 prototypes addressing Bengaluru’s traffic and mobility problems.
Gridlock Oracle’s model predicts how individual incidents can trigger wider congestion across the city.
PRAHAR recommends deployment of officers, barricades, diversions, routes and dispatch stations.
ParkSight uses parking violation data to identify congestion and enforcement hotspots and optimize patrols.
- Who
- Flipkart, Bengaluru Traffic Police, MapmyIndia, Arcadis and HackerEarth organized the hackathon; Gridlock Oracle, PRAHAR and ParkSight were the winning teams.
- What
- Gridlock Hackathon 2.0 selected three AI and machine-learning prototypes designed to address congestion, traffic-resource deployment and parking-related congestion.
- Where
- The winners were announced at Flipkart’s Bengaluru headquarters.
- Why
- The challenge aimed to crowdsource technology-based solutions for Bengaluru’s traffic and urban mobility problems.
- Participants
- More than 33,000 people from across the country
- Prototypes
- More than 1,100 prototypes were developed
- First place
- Gridlock Oracle, which predicts how traffic incidents may spread congestion
- Second place
- PRAHAR, a platform for predictive traffic-resource allocation
- Third place
- ParkSight, which analyzes parking-related congestion
- Data platform
- ASTraM, developed and implemented by Arcadis for the Bengaluru Traffic Police
- Competition structure
- Two stages: an online machine-learning challenge followed by real-world urban mobility problems