Gurugram cops to use AI system to flag traffic violators

Gurugram police said the system can identify vehicle attributes like colour and model, and verify pollution and fitness certificates

The district’s police have partnered with a private firm to implement an artificial intelligence-powered system that can detect fake vehicle registration plates through the city’s CCTV network.The system, called JARVIS (Joint AI Research for Video Instance Segmentation), scans number plates and cross-references them with Regional Transport Office (RTO) records to identify forged plates and stolen vehicles.

Police officials said the system can identify vehicle attributes like colour and model, and verify pollution and fitness certificates. (Parveen Kumar/HT Photo)

Police officials said the system has yielded results in law enforcement’s vehicle inspection efforts. “The technology has helped us identify defaulters using fake number plates and stolen vehicles within weeks of implementation,” said Gurugram police commissioner Vikas Arora.

The cameras cover major intersections including Hero Honda Chowk, Iffco Chowk, Rajiv Chowk and Shankar Chowk, along with main carriageways and internal roads.

Police officials said the system can identify vehicle attributes like colour and model, and verify pollution and fitness certificates. However, details about data privacy safeguards and storage protocols were not shared. According to Staqu Technologies, which developed JARVIS, the system helped identify 4,800 vehicles violating traffic rules at two checkpoints in the first week. The company’s CEO Atul Rai said the technology is being used by police departments in 11 states.

According to police officials, the JARVIS platform cross-references scanned vehicle number plates with government records, enabling the immediate identification of discrepancies or violations. ”The technology also scans vehicles’ pollution and fitness certificates, sending alerts to traffic police if these documents are expired, thereby streamlining enforcement efforts. The system’s ability to recognise vehicles based on physical attributes like colour and model further enhances police efficiency in tracking suspicious vehicles,” Gurugram police commissioner Arora, said.

Meanwhile, Rai said, “Given that our headquarters is also based here, this initiative holds special significance for us.” By analyzing audio-video data, the AI can efficiently identify vehicles violating traffic laws or attempting to evade detection using fraudulent number plates, Staqu Technologies officials added.