For years, Dhaka’s streets have symbolised chaos, frustration and poor traffic management, with endless gridlock and widespread disregard for zebra crossings turning daily commuting into a nightmare for millions.

Against this backdrop of chronic congestion and mismanagement, police have begun introducing AI-based CCTV cameras and modern traffic signal system at major intersections in the capital — a move that is already showing early signs of restoring some discipline on the roads.

Alongside the smart surveillance system, the launch of an e-prosecution or digital case management system has also started reducing incidents of traffic signal and zebra crossing violations by drivers, according to officials.

Traffic officials  believe that once fully implemented, the technology-driven system could significantly reduce the need for direct police intervention on roads, improve transparency in law enforcement and make traffic management more efficient in the overcrowded capital.

They also said the long-term success of the initiative will depend on technological capacity, adequate manpower and greater public awareness.

According to officials of Dhaka Metropolitan Police (DMP), AI-powered CCTV cameras have already been installed at 22 key intersections, including Bijoy Sarani, Gulshan-1 and 2, Jahangir Gate, Hotel InterContinental, Shahbagh, High Court Crossing, Kadam Foara, Matsya Bhaban, Kakrail Mosque Crossing, Police Bhaban, Old Ramna Thana Crossing, Banglamotor, Karwan Bazar, Rampura Traffic Box and the Mirpur National Heart Foundation area.

Under the new system, AI-driven cameras automatically detect vehicles and drivers violating traffic signals and zebra crossing rules.

The footage is then reviewed by the DMP Traffic Technical Unit (TTU) which files digital cases against violators.

Although the AI-based CCTV system formally became operational on April 7 enforcement activities were intensified further from May 7.

Officials said the initiative remains at an early stage and that visible improvements will take time to become fully noticeable across the city.

Still, traffic police and commuters say there are already encouraging signs of behavioural change among drivers.

Visits to several major intersections recently showed that many drivers are now stopping behind zebra crossings instead of occupying pedestrian lanes or crossing stop lines — a common scene on Dhaka roads in the past.

Some residents also told UNB that traffic management at certain intersections appears to be improving gradually.

At Banglamotor intersection, driver Milon Mia, 45, said the presence of AI-based CCTV cameras has made drivers far more cautious.

“Earlier, if I used to violate  a signal intentionally or unintentionally, I could often drive away. But after seeing news reports and social media coverage about AI cameras at different intersections, I now remain very careful because of fear of cases. Even if the road is empty, I do not dare cross during a red signal,” he said.

DMP Deputy Commissioner (Media) N M Nasiruddin said AI-based CCTV cameras recorded nearly 5,000 incidents of traffic signal and zebra crossing violations between April 29 and May 16.

“Among them, 478 cases have already been filed, and more cases will be processed after verification,” he said.

He added that extensive media and social media coverage of the initiative has already created a positive psychological impact among drivers.

“The number of traffic signal violations is gradually decreasing. At the same time, violations involving zebra crossings are also showing a declining trend. Driver’s tendency to follow traffic rules has increased,” he said.

According to the DMP official, the system is also helping reduce confrontations between traffic police and drivers.

Earlier, drivers often fled after violating traffic signals while attempts by police to stop vehicles frequently triggered arguments and created additional congestion on already busy roads.

Passengers of public buses also became frustrated when vehicles were stopped for on-the-spot prosecution, further complicating traffic management.

“With AI-based CCTV cameras, many of these complications can now be avoided. The system has created a more comfortable situation for both police and drivers,” Nasiruddin added.

He said AI cameras would gradually be installed at traffic intersections across the city under a broader smart traffic management plan.