DyNaMo - Safe and sustainable mobility in the city of tomorrow - How does artificial intelligence help cycling safety?

Cycling promises many benefits such as lower noise and fewer air emissions and is therefore set to play a more important role in many cities. A central problem in this regard is the increasing number of traffic accidents. Among children and young adults up to the age of 29, traffic accidents are one of the most common causes of death worldwide.

The "DyNaMo" project is committed to making cycling safer. To prevent traffic accidents in the long term and increase the safety of cyclists, DyNaMo is researching the foundations for:

• implementing suitable cycling infrastructure measures,
• developing training measures for cyclists, each based on an AI-based analysis of the behaviour of road users,
• developing new, legally compliant AI-based assistance systems for cyclists,
• even better recording and anticipatory consideration of cyclists' behaviour in AI-based autonomous driving.

To address these goals, the expertise of computer science, traffic psychology, transport planning and legal sciences is combined. The recording of riding behaviour, e.g. raising an arm to turn, is carried out using vehicle- and wearable-based sensors and is researched using various AI methods (e.g. transformers, hyperparameter optimisation, transfer learning, sensor fusion).