Patna: Turning AI into a weapon against urban traffic chaos, three IIT-Patna students have clinched first position in Flipkart GRiD 2.0 with a solution that can slash unnecessary deployment of traffic personnel and barricades by 86%.The national-level technology and innovation challenge, with an AI-powered solution, could significantly improve traffic management and reduce wastage of resources.The winning team comprises Kashish Rai and Bhumi Garg, both students of mathematics and computing, and Pragya Mahajan of computer science. All three are in the third year of the BTech-MTech integrated dual-degree programme at IIT-Patna.Competing under the theme ‘Event-driven traffic congestion’, the team developed an innovative system called ‘Gridlock Oracle’ to help traffic control rooms predict whether reported incidents are likely to escalate into major disruptions requiring road closures.The hackathon was organised in collaboration with Bengaluru Traffic Police, which provided participants with a dataset of 8,057 real-world traffic incidents.Traffic authorities often face difficulties in assessing the severity of reported incidents in advance. Uncertainty forces them to deploy additional personnel and barricades as a precaution, sometimes resulting in substantial wastage of manpower and resources.IIT-Patna’s registrar A K Thakur said the team’s AI-driven decision-support system addresses this challenge by predicting which incidents are likely to require road closures and identifying suitable locations for positioning traffic management resources.During evaluation on previously withheld data, the solution demonstrated its potential to reduce unnecessary resource deployments by 86%, highlighting the practical application of AI in addressing urban traffic congestion.
