What is it about?
When a tram stops at a station in the middle of the road, its long body hides the pedestrians crossing in front of it from drivers in the next lane. The tram itself can see these pedestrians, because its cameras look straight at the crossing. This paper proposes that trams and buses use the cameras they already have to spot people and send a short wireless warning to nearby connected cars. To keep the network from being flooded with messages, a tram sends a warning only when a nearby driver could still react to it, and it pays extra attention at places where sudden braking has often happened before. Because trams are run by a transport operator, their equipment can be managed and checked, which makes their warnings easier to trust than messages from anonymous cars.
Featured Image
Photo by Michael Pointner on Unsplash
Why is it important?
Pedestrians are not connected to any vehicle network. Their safety depends on whether someone can see them in time, and tram stops are some of the most dangerous places for them. We tested the idea with the real layout of the Mariánska tram stop in central Bratislava. In our calculation, a driver approaching at the 30 km/h limit would not see a pedestrian crossing in front of the tram until it was too late to stop. With a warning from the tram, the same driver would have about 2.4 seconds more time and could stop with several metres to spare. This holds for slower walkers, such as older or mobility-impaired people, who are most at risk.
Perspectives
This work grew out of our earlier research on where trams in Bratislava brake hard. The first idea was to have the tram warn others when it was about to hit someone, but that does not help, since the tram already knows. The idea became useful when we turned the question around and asked who the tram is hiding people from. The tram can see what drivers beside it cannot. The next steps are testing the idea with real camera data, measuring how it affects network load, and studying how drivers and automated vehicles should respond to these warnings. We also want to be careful that a missing warning is never read as “no one is there.”
Dominik Šalgovič
Slovak University of Technology
Read the Original
This page is a summary of: Public Transport as Trusted Mobile Collective Perception Infrastructure: Resolving Transit-Induced Occlusion for Vulnerable Road Users, December 2026, ACM (Association for Computing Machinery),
DOI: 10.1145/3843201.3856199.
You can read the full text:
Contributors
The following have contributed to this page
