Sign
Sign is an interactive, gesture-based language used between pedestrians and autonomous vehicles to ensure that people have an equal say in urban traffic
A student project by Judith Brünner, Ali Haidari and Henrike Rähse
3rd semester, 2019
Supervising lecturers: Prof. Claudius Coenen, Prof. Andrea Krajewski, Prof. Tsunemitsu Tanaka, Stephan Voigt
Autonomous driving is a widely discussed topic in the world of mobility. In the course of our research, we identified a lack of communication and interaction between autonomous vehicles and pedestrians. Our team, CommuniCouch, therefore decided to tackle the problem by developing a language consisting exclusively of gestures.
The vehicle recognises the gestures and responds accordingly by interacting with users through visual and haptic elements. This creates a balanced distribution of power between the two entities.
The system
Using a machine learning algorithm, we have programmed a gesture recognition system that can distinguish between three different gestures.
These gestures are used to signal to the vehicle that we either wish to cross the road, travel with it, or cancel one of the previous interactions should we have changed our mind.
The aim here, once again, is to give pedestrians back control over their own journey.
Interaction
Within this concept, the car can then be redefined as a means of transport. The windscreen now essentially serves as a view for the passenger and a display for the pedestrian. At the same time, the wing mirrors are replaced by small ‘ears’, which also react to the pedestrian.
In this way, we transform a conventional potential hazard into a trustworthy interactive partner that respects people’s freedom of movement and communicates with its surroundings and with other vehicles of its kind.