Vilnius uses artificial intelligence to tackle traffic jams

On one of Vilnius’ busiest roads, an artificial-intelligence traffic system has cut total driver waiting time by about 19 minutes during a single morning rush hour.

The system operates on Nemenčinė Highway, where it coordinates seven intersections and pedestrian crossings. It analyzes traffic-flow data, predicts near-term conditions and automatically adjusts traffic lights. At some pedestrian crossings, waiting times fell to as little as 30 seconds.

The system has also been tested to give public transport priority. It can turn a traffic light green earlier or extend the green phase for a bus, depending on the intersection and traffic flows. That reduces idle time and increases public transport speed.

“

Congestion cannot be solved with a single measure. Traffic must be managed comprehensively.

Evalds Morkūns, head of JUDU’s Traffic Management and Organization Department, said congestion cannot be solved with a single measure. The city is focusing on making existing roads and traffic lights work more efficiently, improving public transport and making alternatives to private cars more attractive.

Vilnius changed 60 traffic-light programs this year, mainly to improve coordination at intersections, increase road capacity and reduce queues. At the intersections where results were analyzed, the average number of vehicle stops fell by about 10%. Morkūns said the city plans to expand the use of real-time data and traffic forecasting, especially on coordinated roads with heavy traffic flows.

“

One example of such a solution is already operating on Nemenčinė Highway.