Newsletters
Bangladesh’s capital has launched its first artificial intelligence-powered traffic enforcement system. The new system connects existing road cameras to software that automatically detects violations and issues fines.
This marks a major change in how one of the world’s most congested cities manages its streets.
Dhaka, with over 22 million residents, has faced severe traffic jams for years. A joint study by the World Bank and Bangladesh University of Engineering and Technology found that the city’s average traffic speed is just 4.8 kilometers (2.98 miles) per hour, slower than walking.
Another study by the U.S. National Bureau of Economic Research ranked Dhaka as the world's slowest city.
Dhaka police introduced the system in April. It links traffic camera feeds to AI software that can spot different violations, such as running red lights, improper lane changes, and illegal parking.
When a vehicle is flagged, the system automatically issues a fine and sends a text message to the registered owner. Hannan Rahman Jibon, 28, was one of the first drivers caught by the new system after running a red light.
"The owner of my car, who was sitting at home, received a text message saying that the vehicle had violated traffic rules," Jibon told Agence France-Presse (AFP).
He received an automatic fine of 2,000 taka ($16). "I am more careful now, with cameras installed in many different parts," he said. Traffic sergeant SM Nazim Uddin said he has noticed a change in how drivers behave.
"Those who violate the rules turn against us," Uddin told AFP. "But since AI was introduced, people behind the wheel have started obeying the law, and we have been spared the everyday quarrels."
The city had previously relied almost entirely on manual enforcement, with officers stretching ropes across roads before signals turned green and stopping vehicles individually to check documents and issue on-the-spot fines, a process that frequently led to arguments and physical confrontations.
Inside a control room at police headquarters, analyst Sharmin Afroze monitors live camera feeds connected to the AI system.
Human operators review violations flagged by the software before fines are confirmed. Afroze said the system recorded nearly 800 violations in a single day.
"Before, police used to stop vehicles, check papers and determine fines," she told Agence France-Presse (AFP).
Police spokesperson N.M. Nasiruddin said authorities have prosecuted at least 300 vehicles since the system went live. "We have started getting results," he said.
For now, police are only fining the most serious offenders and giving warnings to others as the system is gradually introduced.
The system does not yet cover the whole city, and manual enforcement will be reduced over time.
Officials admit the system still has technical challenges. "Some number plates are blurred or too small to identify," Afroze said.
Police are working with the road transport authority to fix this. They also plan to add features to detect vehicles driving on footpaths.
Right now, the system only targets motor vehicles, and authorities are still deciding how to handle Dhaka’s many pedal rickshaws.
Hasib Mohammed Ahsan, a professor at Bangladesh University of Engineering and Technology, warned that technology by itself will not guarantee the system’s long-term success.
"We have spent huge amounts of money on traffic signals and their upgrades, but those efforts were never sustained," Ahsan told AFP.
"We do not follow rules, there is no consistency in our planning, and there is no accountability for failure."
It is still unclear if the AI system will break this pattern, but early signs show it is already changing how people behave on Dhaka’s streets.
Set as preferred source
Set as preferred source
3 min read
