Artificial intelligence should reduce delays in train traffic

- Since train traffic is such a climate-smart alternative, it is good if it is a realistic alternative instead of, for example, trucks. Then it wants it to work, says Jan Lundberg, professor of operations and maintenance technology.

He is the leader of the project at LTU, Luleå University of Technology.

For two years, measurements will be made on a test rail where a train is towed back and forth to collect data on rail damage - damage that scientists themselves sometimes create to teach the computer how the vibration curves look when the rail is damaged.

If train traffic gets rid of the interruptions that are linked to unplanned stops due to noisy train switches, Professor Jan Lundberg reckons that the delay hours can be halved.