My research is centered around autonomous motion planning in environments with pedestrians and cyclists (so called vulnerable road users). The challenge in autonomously planning safe trajectories is that intentions and behavior of humans are not precisely known, especially when predicting their future motion. This gives rise to a probabilistic motion planning approach where the motion planner makes use of stochastic predictions to generalize the notion of safety to uncertain environments. To translate this to the real world, I am involved in the design of an experimental autonomous motion planner on a self-driving vehicle in cooperation with the Intelligent Vehicles group at the TU Delft (Demo: https://youtu.be/opSSsKPpXew).