Hit-and-Run Case: The man lacked specifics and only recalled the red lines on the truck that struck his wife in Nagpur. He could not identify its size or make. Despite this, police used artificial intelligence (AI) to apprehend the accused within 36 hours. On Raksha Bandhan, August 9, a speeding truck hit the man and his wife in Nagpur. The woman fell on the road, and the truck ran her over before fleeing the scene. The man then strapped his wife’s body to his bike and transported her to their village in Madhya Pradesh. A chilling video of the incident later went viral.

Nagpur Rural Police Superintendent Harsh Poddar told the media that the victim had provided “very little information.” When asked how police identified the suspect with such limited details, Poddar explained that they gathered CCTV footage and processed it through AI algorithms.

MARVEL Police AI System:

Mr. Poddar explained that investigators collected CCTV footage, or metadata, from three toll nakas located 15 to 20 kilometers apart and subjected it to two AI algorithms for analysis. He said both algorithms relied on Computer Vision technology. The first algorithm scanned the entire footage, pulled out trucks with red markings, and identified them. The second algorithm analyzed the average speed of these trucks and determined which one might be involved. Based on this process, investigators identified a single truck.

Police traced the vehicle to the Gwalior–Kanpur route, nearly 700 kilometers from Nagpur. A Nagpur rural police team located the truck and arrested the accused the day before yesterday. The officer confirmed, “We have now arrested the accused and cracked the case using AI within 36 hours.” Maharashtra developed MARVEL (Maharashtra Research and Vigilance for Enhanced Law Enforcement) as the nation’s first state-level police AI system, and the state owns it entirely.

Also Read: Hyderabad Nightlife to Become ₹26,000-Crore Economy by 2031

The police officer stated that officials developed the system to implement and deploy AI in Maharashtra government departments, including law enforcement agencies. He explained that the core idea is to build AI solutions owned by the Maharashtra government instead of relying on external sources, enabling them to analyze and utilize government metadata effectively.

He added that authorities collaborated with up to 13 different departments to create these solutions. Since data forms the backbone of AI, the government has been analyzing its vast datasets to build AI-driven solutions for the entire ecosystem. He further noted that Maharashtra became the first state to establish this model, and many large states have already shown interest in duplicating Marvel. When asked how AI differs from traditional inquiry, the officer said that technology accelerates the process by analyzing data within minutes and providing solutions rapidly.

“We receive terabytes of metadata; they could include organized crime or sophisticated financial crimes that involve a lot of telecom data, for example. In the previous model, seasoned law enforcement officers would examine the data to find trends. In these situations, human error was quite likely to occur. Additionally, analyzing big data sets could take weeks or months. We can now analyze this data in a matter of minutes because of the power of artificial intelligence and fast computers. For instance, in this instance, 12 hours of CCTV data might have been analyzed in 12 to 15 minutes, when in the past, it might have taken over a day,” the officer stated.

Share

• 
• 
• 
• 
• 
•