Tech Stack

Description

You will design, train, and deploy reinforcement learning-based controllers for robots operating in complex real-world environments. Your work spans from simulation to deployment, ensuring robust and efficient policies under hardware constraints.

Requirements

• MSc or PhD in Robotics, ML, Control Engineering, or related field with focus on RL and robot control.
• Hands-on experience training and deploying RL controllers on real robotic hardware.
• Deep familiarity with modern RL techniques (PPO, SAC, TD3, model-based RL, sim-to-real).
• Solid understanding of robot dynamics, kinematics, and classical control theory.
• Proficient in physics simulators (MuJoCo, Isaac Gym, PyBullet, Gazebo) and strong software engineering skills.

Responsibilities

• Design, train, and deploy RL-based controllers for robots in unstructured real-world environments.
• Enable robots to perceive and understand surroundings by fusing sensory inputs including vision.
• Own the full pipeline from simulation to deployment of learned policies.
• Ensure robustness and efficiency of policies under tight hardware constraints.
• Iterate on physical robotic systems to achieve real-world performance.