Luxembourg tests AI to ease traffic and boost public transport

Luxembourg is testing artificial intelligence tools to improve public transport and ease traffic congestion, with researchers at the University of Luxembourg developing systems aimed at making it easier for people to switch from cars to buses and trains.

The work is part of the EU-funded ACUMEN scheme (AI-aided deCision tool for seamless mUltiModal nEtwork and traffic maNagementproject), which uses AI and large datasets to better manage complex, multimodal transport networks involving buses, trains and other forms of mobility.

“AI is a good tool for improving traffic forecasts based on data,” said Francesco Viti, professor of transportation engineering at the University of Luxembourg.

University of Luxembourg academic Francesco Viti is researching how autonomous buses, AI and digital twins can make public transport more efficient. © Photo credit: LW-Archiv / Claude Piscitelli

However, he cautioned that the technology is not a solution on its own. It requires interaction and follow-up questions, because “the first answer is usually not the best.” AI should support human decision-making rather than replace it, he added, noting that “AI can save us a lot of time.”

Researchers are using the technology to analyse large volumes of mobility data and improve traffic flow. “We use all available data to improve traffic flow,” Viti said.

AI is a good tool for improving traffic forecasts based on data.

Francesco Viti

University of Luxembourg professor

Part of the project involves creating a “digital twin”, which is a real-time model of the entire transport system, allowing simulations of how disruptions or changes affect overall traffic patterns.

Arrival of the new autonomous shuttle in Esch three years ago. © Photo credit: Luc Deflorenne

Pilot projects

Two pilot projects are currently underway in Esch-sur-Alzette and Contern, where autonomous shuttle buses are being tested as part of efforts to improve connections between different modes of transport.

We use all available data to improve the organisation of traffic.

Francesco Viti

University of Luxembourg professor

“In Esch, the project served to demonstrate the technical feasibility of autonomous, on-demand shuttle operations,” he said.

The vehicles run along the city’s main shopping street at low speeds and are designed to navigate obstacles automatically, while still carrying a driver who can intervene if needed.

Initial findings suggest the service is mainly used by older people and those with reduced mobility, with users rating the experience positively. The project has also tested flexible routing, such as stopping only when passengers need to board or alight.

The research also highlights structural challenges in the system, including poorly coordinated timetables, inconvenient transfers and unclear information, which often push commuters towards using cars instead.

If you can make better use of time on the road, then it’s not wasted.

Francesco Viti

University of Luxembourg professor

The initiative comes as Luxembourg faces growing pressure on its transport infrastructure, with congestion driven in part by long commuting distances. “I wish people didn’t have to drive over 50 kilometers every day,” Viti said.

“People need to rediscover the benefits of active mobility, such as cycling or walking, as well as public transport,” Viti said.

(This article was originally published by the Luxemburger Wort. Translated using AI and edited by Kabir Agarwal.)