If a dangerous situation occurs on the road, nonverbal signs, such as gestures or eye contact, help road users to communicate and avoide accidents. The ability of humans to communicate in nonverbal form poses a major challenge to automated vehicles. This raises the question of what future communication between (partially) automated road users and weaker road users such as pedestrians, cyclists and non-automated vehicles could look like?

Objectives and approach 

INITIATIVE stands for intelligent man-machine communication in mixed traffic (German: INtellIgenTe Mensch-Technik KommunIkATIon im gemischten VErkehr). The project aims to develop and implement AI-based communication to integrate automated vehicles in mixed traffic scenarios. The goal is to equip automated vehicles with communication interfaces for external road users (external HMI) and for occupants of the vehicle (internal HMI).

To ensure the success of the communication process, information needs to be interpreted and synchronized for all traffic participants. For the synchronization process  relevant road users need to be identified and their intentions predicted in order to be able to transmit appropriate signals in the context of the overall traffic scenario.

The behavior of the road users should be monitored iteratively and signals should be adapted in real time accordingly. To identify relevant participants within a communication, scientist plan to access metadata from a networked infrastructure. AI methods will be used to synchronize signals, detect the intentions of communication participants, and preselect relevant road users.

Is there data protection in the data flow? Limitations of AI-based communication

What information can infrastructure provide to the vehicle? How should the information, data collected by cameras for example, be processed in compliance with current data protection laws? What information is needed? are the key questions considered by the Institute for Legal Informatics. The project further investigates whether existing regulations ensure the protection of transferred data sufficiently and where there is a need for further development of the legal framework. In addition, it is investigating who is liable in the event of an accident caused by faulty data (processing).