A Maneuver Based Interaction Framework for External Users of an Automated Assistance Vehicle.

Automated vehicles become gradually available for restricted environments. Fully Automated Vehicles (FAV) operate without a driver and need to cooperate and interact with other road users of any kind. This article illustrates an interaction framework, which allows a human user outside the car to interfere with the FAVs guidance. This is achieved by communicating a desired maneuver, where the external user is asked to choose among a set of possible maneuvers. This set of maneuvers is communicated by the FAV to the user and has been checked for execution feasibility by the FAV, based on its perception. To this end, the environment is represented as an occupancy grid and a path search without distinct goal is performed. A small set of paths will be selected and communicated to the external user in an abstract level. This article presents the planning framework, as well as basic implementations for suited path search algorithms. The conclusion addresses unsolved challenges.