Story so far: On February 20, Dick Schoof, former Prime Minister of the Netherlands, visited the Traffic Management Center (TMC) of the Bengaluru Traffic Police to talk about the Actionable Intelligence for Sustainable Transformation Management (ASTraM) system. Developed in collaboration with Dutch company Arcadis, ASTraM aggregates data from CCTV footage and open data sources to monitor and forecast real-time trends on Bengaluru’s congested roads. Cities like Delhi, Hyderabad and even Dubai have shown interest in the technology, the source said.
How does ASTraM and other predictive systems work?
Google Maps and other prominent mapping apps have been providing users with real-time traffic congestion data for years. They also inform about traffic incidents and affected regions. However, these systems do not provide predictive services. In contrast, ASTraM identifies congested areas, doses them, and then alerts the appropriate officials at fifteen-minute intervals. By capturing both recurring and non-recurring congestion points, this application provides insight that can be used for predictive traffic control and incident reporting.
Automatic Number Plate Recognition (ANPR) is another strategy used in Indian cities to identify criminals. The Greater Chennai Traffic Police also uses an Integrated Traffic Regulation System (ITRS) that includes artificial intelligence and live feeds for effective traffic management, enabling them to track down repeat offenders.
What are the benefits of technology-focused traffic policing?
Intelligent traffic policing systems allow authorities to quickly process data across multiple media formats to get a consolidated picture of which areas require immediate intervention and urgent traffic policing solutions. Police can also build on this data to prepare for future events such as processions, riots and traffic jams. This is much more efficient than relying on existing app-based GPS systems or waiting for users to phone or post their complaints on social media before taking action.
Additionally, because these intelligent traffic police systems have more localized data to work with, they can potentially prevent accidents that occur due to a mixture of Google Maps and human error. Take, for example, past incidents where Google Maps allegedly took drivers into danger zones such as broken bridges, causing several deaths.
Surveillance/intelligence policing can also help authorities identify violators in high-risk areas without officers having to be physically present at the scene.
In early February, it was reported that the Udupi police planned to implement a “contactless system” to monitor traffic violations by installing 150 surveillance cameras in the coming months. This will help them flag dangerous driving, signal-jumping drivers and seat belt violations. According to another recent report, 25 Intelligent Traffic Management System (ITMS) cameras have been installed on the Mysuru-Bengaluru state highway to detect traffic violations by motorists.
What are the risks of predictive traffic policing?
Digital rights advocates warn of mistakes made by news-based systems. They also warn of the privacy risks that result from increased surveillance. As more data on citizens is collected for traffic policing purposes, greater security measures and investments are needed to secure sensitive and personally identifiable information.
“Although a controversial issue in Western countries (and some US cities have banned it), predictive policing is widely deployed in Asia,” took note of the Deloitte report and added“Both surveillance and predictive policing are seen as undesirable in more privacy-conscious geographies such as the EU and North America. Latin America and Asia have shown greater acceptance.”
For example: last year in Delhi, authorities deployed AI cameras to enforce a fuel ban – when old vehicles were banned from the roads – to tackle air pollution. Cameras with an automatic license plate reader were used for this purpose.
As previously stated, the implementation of advanced AI traffic policing solutions is often undermined by a lack of effective human intervention. In Kerala, AI-powered cameras were deployed in 2023, with 726 of them operationalized to detect various traffic and driving rule violations in accident-prone areas. However, implementation was far from flawless: multiple sightings sometimes led to multiple fines for the same vehicle; and there was confusion about the various speed limits. This led to a public backlash, which in some cases resulted in fines being waived at the initial stage. Some drivers have even tried to avoid the AI cameras by covering up their license plates or using fake numbers.
However, a year later, the state’s Motor Vehicles Department (MVD) imposed a fine of ₹437 million, with nearly 68 million motorists caught violating traffic rules between 5 June 2023 and 22 June 2024. However, it managed to collect only around ₹80 crore. This collection of fines is also a problem for traffic violators in Karnataka. Even so, Kerala’s initiative was the first of its kind and attracted the interest of other states, including Tamil Nadu.
AI surveillance techniques and live data analysis can improve traffic policing and road safety, but they bring their own risks – data management, privacy and cyber security. In fact, Mr. Schoof was surprised during his visit to see the amount of information that was made available to the authorities, including open data and surveillance infrastructure.
Published – 27 Feb 2026 08:30 IST
