The story so far: On February 20, Dick Schoof, former Prime Minister of the Netherlands, visited Bengaluru Traffic Police’s Traffic Management Centre (TMC), where he spoke about the Actionable Intelligence for Sustainable Transformation Management (ASTraM) system. ASTraM, developed in collaboration with the Dutch company Arcadis, pools data from CCTV footage and open data sources to monitor and predict trends in real-time across Bengaluru’s congested roadways. Cities like Delhi, Hyderabad, and even Dubai, have expressed interest in this technology, a source said.
How do ASTraM and other predictive systems work?
Google Maps and other prominent mapping applications have been providing users with real-time traffic congestion data for years. They also report on traffic incidents and affected regions. But these systems do not provide predictive services. In contrast, ASTraM identifies congested areas, batches them, and then alerts relevant officers at fifteen minute intervals. By capturing both recurring and non-recurring congestion points, this application provides insights that can be used for predictive traffic policing and incident reporting.
Automatic number plate recognition (ANPR) is another strategy that is used across Indian cities to identify violators. The Greater Chennai Traffic Police also uses an integrated traffic regulation system (ITRS) that involves AI and live feeds for efficient traffic management, allowing them to track down repeat offenders.
What are the advantages of tech-focused traffic policing?
Intelligent traffic policing systems allow the authorities to quickly process data across many media formats in order to get a consolidated picture of which areas require immediate intervention and urgent traffic policing solutions. The police can also build on this data to prepare for future events like processions, unrest, and traffic chokeholds. This is far more effective than relying on existing app-based GPS systems, or waiting for users to telephone or post their complaints on social media before taking action.
Furthermore, since these intelligent traffic policing systems have more localised data to work with, they can potentially prevent accidents that occur due to a mixture of Google Maps and human errors. For instance, take past incidents where Google Maps allegedly led drivers into dangerous zones such as non-functional bridges, causing multiple deaths.
Surveillance/intelligence policing can also help the authorities identify rule violators in risky areas, without officers having to be physically present at the scene.
In early February, it was reported that the Udupi police was planning to implement a ‘contactless system’ of monitoring traffic violations, by installing 150 surveillance cameras in the coming months. This will help them flag dangerous driving, drivers skipping signals, and seatbelt offences. As per another recent report, 25 Intelligent Traffic Management System (ITMS) cameras were installed on the Mysuru-Bengaluru National Highway, to detect motorists’ traffic violations.
What are the perils of predictive traffic policing?
Digital rights advocates warn of errors caused by intelligence-based systems. They also caution against privacy risks that arise from increased surveillance. As more citizen data is collected for traffic policing, greater cybersecurity safeguards and investments are required to secure sensitive and personally identifiable information.
“Although a controversial issue in Western countries (and some cities in the US have banned it), predictive policing is being deployed widely in Asia,” noted a report by Deloitte, adding, “Both surveillance and predictive policing are considered undesired in more privacy-aware 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 the fuel sale ban—when old vehicles were banned on the roads—to tackle air pollution. Automatic Number Plate reader cameras were used for this purpose.
That being said, the implementation of advanced, AI-powered traffic policing solutions is often undercut by the lack of effective human intervention. In Kerala, AI-powered cameras were deployed in 2023, with 726 of them being operationalised to detect different traffic and driving rules violations in accident-prone areas. However, the implementation was far from flawless: multiple sightings led to multiple fines for the same vehicle sometimes; and there was confusion over differing speed limits. This led to a public backlash, leading to the waiving of fines in certain instances in the initial phase. Some riders have even tried to evade AI cameras by covering their license plates or using fake numbers.
A year later, however, the state’s Motor Vehicles department (MVD) imposed ₹437 crores as a fine, with close to 68 lakh motorists caught breaking traffic rules between June 5, 2023 and June 22, 2024. However, only around ₹80 crores could be collected. This collection of fines is also an issue with traffic rule violators in Karnataka. Even so, Kerala’s initiative was the first of its kind and attracted interest from other States, including Tamil Nadu.
AI surveillance techniques and live data analysis can enhance traffic policing and road safety, but they come with inherent risks—data management, privacy, and cybersecurity. In fact, Mr. Schoof, during his visit, was surprised to see the amount of information being made available to the authorities, including open data and surveillance infrastructure.
Published – February 27, 2026 08:30 am IST
