By Peter Witten, AAU Communication and Public Affairs

You’ve probably experienced this: during the morning rush hour, cars are stuck in a long queue at a traffic light. Time crawls by while you and other drivers impatiently wait for green. "Will I make it on time?" you wonder as your petrol car idles - polluting.

Can it be done better - smarter?

The answer is yes. Research from the Department of Computer Science at Aalborg University has shown it - and now the solution is being used in Turkey.

Less pollution

By using AI to analyze data from traffic radars, traffic lights can be controlled intelligently, reducing waiting times, getting drivers through faster, and lowering fuel consumption. This also reduces pollution.

In a typical intersection, the green light period is predetermined regardless of the number of cars in each direction. With artificial intelligence, traffic can flow more smoothly through intersections.

Research at AAU led to the creation of the company ATS, Advanced Traffic Systems, with Andreas Berre Eriksen as CEO.

He was a PhD student at AAU and part of the research project "Controlling Signalized Intersections using Machine Learning". The project involved student Jakob Haahr Taankvist, Professor Kim Guldstrand Larsen from the Department of Computer Science, and Associate Professor Harry Lahrmann from the Department of Civil Engineering and Urban Planning.

Great results, but…

The promising trial results with AI-controlled traffic lights - including a major one in Aarhus - raised hopes for demand and a commercial breakthrough for Advanced Traffic Systems.

"We got great results from the Aarhus trial. We expected our phones to be ringing off the hook with customers," says Andreas Berre Eriksen.

But that didn’t happen. The flood of customers never came.

Turkish partner

Andreas Berre Eriksen and his colleagues didn't give up. They stuck with the idea and the company, and a few years ago Advanced Traffic Systems  connected with the Turkish firm Intetra, the world's largest in road toll systems - similar to paying when crossing the Great Belt Bridge.

After Intetra invested in Advanced Traffic Systems, the rollout began.

"Now Intetra is really pushing these systems in Turkey. So here, nearly seven years later, we’re finally starting our deployment journey," says Andreas Berre Eriksen.

The AI-system from ATS is now used in pilot projects at major intersections in five Turkish cities, including Istanbul. In neighboring Iraq - in Baghdad - a trial with ATS's traffic light control system is also underway.

Andreas Berre Eriksen notes that with Intetra, ATS has finally achieved a breakthrough.

"It’s awesome that it worked in Istanbul. But it's sad Denmark isn't part of it. We were originally four researchers from Aalborg University who started a company to help cars move faster and reduce CO₂ emissions - something our politicians here talk about all the time. In our part of the world, we're good at making speeches and having ambitions. But I lose respect when there are so many obstacles and procurement rules that make it hard to spread our technology," says Andreas Berre Eriksen.

A sea of obstacles

"We're not delivering a full traffic control cabinet for intersections. We're delivering a small computer that can make these calculations and instruct a cabinet on how to manage traffic. And here at home, we've faced a sea of obstacles from traffic cabinet suppliers," he explains.

According to him, municipalities either wouldn't or didn't dare challenge the traditional suppliers.

Encouraging

He is pleased that, with Intetra's help, the AI system is now used in Turkey.

"It's always exciting to create something new and succeed with it abroad. I've visited Intetra in Istanbul several times and toured various Turkish cities to see our traffic lights in action," says Andreas Berre Eriksen.