Role overview
- Design, develop, and optimize reinforcement learning algorithms for real-time control and locomotion of humanoid robots.
- Integrate learned policies into real-world robot platforms with hardware-in-the-loop validation.
- Collaborate with mechanical, perception, and embedded systems teams to ensure tight integration between hardware and software.
- Apply advanced techniques such as curriculum learning, domain randomization, and sim2real transfer to improve policy generalization.
- Analyze and optimize control performance with a focus on robustness, energy efficiency, and adaptability.
- Contribute to the continuous development of our in-house RL training pipelines and tooling.
- 2+ years of experience in machine learning (NNs, LVMs) and reinforcement learning applied to robotics or similar realtime environments.
- Hands-on experience with physics simulation environments (e.g., MuJoCo, Isaac Lab).
- Proficiency in Python and C++ for algorithm development and deployment.
- Experience with deep learning frameworks (e.g., PyTorch, JAX, TensorFlow).
- Familiarity with ROS/ROS2 and real-time robotic systems.
- strong software development experience, including CI/CD, unit testing, etc.
- Strong understanding of classical and modern control theory, locomotion dynamics, etc.
- Experience deploying RL algorithms on physical robots.
- Experience with high-performance computing for distributed training.
- Contributions to open-source RL, ML or robotics projects.
- M.Sc. or Ph.D. in Robotics, Computer Science, Mechanical Engineering, or a related field.
Tags & focus areas
Used for matching and alerts on DevFound Fulltime Ai Engineer Robotics Ai