Abstract
The development of collision avoidance systems which are based on active intervention into the steering process is a promising way of reducing accident fatalities resulting from collisions with other road users. An important prerequisite of such systems is the calculation of feasible manoeuvres. In this paper, a sampling-based path-planning algorithm from robotics is proposed for planning trajectories on highways. In particular, adaptive polar splines are presented as one suitable way of describing path segments. Furthermore, it is demonstrated how kinematic constraints can be incorporated into the path-planning process.