Research Engineer — Reinforcement Learning for Robot Navigation

We build mobile robots for hospital logistics — transporting medication, linen and lab samples between departments. Our robots navigate using a hybrid approach: pre-mapped routes for known environments and reactive planning for obstacles. The reactive planning is our bottleneck. Rule-based obstacle avoidance fails in crowded corridors, near elevators and in dynamic clinical environments where humans move unpredictably. We want to explore RL for navigation in partially-known environments. You'd work in our ROS2 simulation environment (Isaac Sim), run training experiments, and evaluate in our physical test corridor. PhD preferred but not required. What matters is you've trained RL agents that work outside of benchmark environments.