This list contains only the countries for which job offers have been published in the selected language (e.g., in the French version, only job offers written in French are displayed, and in the English version, only those in English).
As a Reinforcement Learning Intern, you will help bring perception into Mirokaï's navigation stack. You will integrate depth camera and time-of-flight sensor models into our RL environments, and use these rich perceptual inputs to train navigation policies that can handle real-world obstacles and spaces. This internship offers deep hands-on experience at the intersection of sensor simulation, reinforcement learning, and sim-to-real transfer.
Job Responsibility
Integrate depth camera and time-of-flight sensor models into our Isaac Lab simulation environments
Design observation spaces and policy architectures that leverage perceptual inputs for obstacle-aware navigation
Train and evaluate reinforcement learning policies conditioned on simulated sensor data
Analyze navigation performance, robustness to sensor noise, and sim-to-real transfer aspects
Integrate trained policies into the Mirokaï software stack and validate them on physical robots
Stay up to date with recent research in perceptive RL and learning-based robot navigation
Requirements
Bac+5 in Robotics, engineering, computer science, or related field
Coursework or project experience in reinforcement learning or learning-based control
Strong Python skills and knowledge of a deep learning framework (PyTorch, TensorFlow, etc)
Familiarity with simulation environments such as Isaac Sim, Mujoco, or Gazebo is a plus
Solid analytical and problem-solving abilities
Nice to have
Experience implementing RL algorithms
Familiarity with robot control, dynamics modeling, or motor control
Prior work in sim-to-real transfer or domain randomization