Road user behaviour modelling based on data driven methods
A paradigm shift within the transport system is here. Historically the human driver has been responsible for both the car and the driving. As machines are introduced, they will take over the responsibility of interpreting the surroundings and other road users. For this we need methods to capture road user behavior and models that can predict them.
Transport is becoming commodity and vehicle automation and servification are the enablers. Building a transportation system for everybody requires understanding of the behavior of the different users. At RISE, a broad knowledge within road user behavior is developed from a multidisciplinary approach.
Camera analysis to predict behaviour
We design interaction protocol for cooperative and automated vehicles to perform e.g. cooperative platoon merge, or automatically negotiate free-of-way in an intersection. We monitor traffic behavior using camera-based sensors to obtain trajectories that can be used to predict future behavior for e.g. action and intention prediction, both from a single user perspective as well as in interaction between two or more road users.
Visual and sonically aspects
We also explore methods to e.g. visually or sonically interact between an autonomous vehicle and surrounding road users. In particular we explore behavior in interaction with platoons of trucks that drive with short inter-vehicular distance and how they can communicate with the surrounding traffic their intentions. Another important use-case is the one with visually impaired road users and how they can interact with (autonomous) vehicles.
How to measure and interpret interactions
RISE can provide guidelines on how to measure interaction, how to design systems to become interpretable and what variables to study to understand intentions of road users.
Researchers at RISE have long experience from the transportation domain and in close collaboration with our partners we deliver valuable insights around products, businesses and innovations.