Corridor-Based Motion Planning for Teleoperated Driving Tasks

Remote operation of road vehicles in complex urban scenarios causes a high workload for the human operator, which may lead to a stop-and-go driving behaviour. This contribution presents a supervisory control concept to raise the autonomy level, while the operator still remains the main decision-maker in all driving tasks. With the introduced approach, operators can define a corridor to the desired destination according to their perception to take into account in the motion planning process, for example, inadequate lane markings or obstacles still undetected by sensors. Within the specified corridor, the automated vehicle itself calculates the optimal motion in the dynamic environment. For this purpose, a new hybrid motion planning method has been developed that, in the first phase, determines an optimal path using a modified Constrained CHOMP algorithm. In the second phase, a collision-free trajectory is generated online along the optimal path by means of a linear time-varying model predictive control scheme under the consideration of time-varying system constraints. Real driving experiments confirm the effectiveness of the proposed approach and highlight both the achieved driving safety and the real-time capability.