Robotics Motion Control Research Engineer

Focus: Learning-Based Control, Whole-Body Control & Robot Learning

Responsibilities

1. Design, train, evaluate, and deploy motion control policies (System 1 / System 0) for robotic mobile manipulation tasks.
2. Develop and advance learning-based control methods, including behavior cloning and reinforcement learning, with a strong emphasis on stability, safety, and real-time performance in real-world environments.
3. Build unified perception-to-action policies that integrate command tracking, vision-based obstacle avoidance, contact force sensing, and adaptive stiffness control.
4. Train robust motion policies that can handle real-world challenges, including sensor noise, uncertain contact dynamics, and changing environments.
5. Collaborate closely with foundation model, systems, and hardware teams to deploy learned policies on physical robots, enabling reliable, real-time, on-device execution.

Qualifications

1. Strong understanding of both classical robotics control and learning-based approaches, including their respective strengths, limitations, and practical trade-offs.
2. Hands-on experience developing and deploying algorithms on real robotic platforms, such as legged robots, dexterous hands, robotic arms, or drones.
3. Experience in robot motion control, with familiarity in robot learning methods such as behavior cloning and reinforcement learning.
4. Proficiency with robotics simulation environments such as Isaac Lab or MuJoCo, and machine learning frameworks such as PyTorch or JAX.
5. Ability to independently design experiments, troubleshoot failures, and rapidly iterate on algorithms for real-world robotic systems.

Nice to Have

1. Experience with multimodal alignment or fast-slow control architectures (e.g., System 1 / System 0).
2. Experience working with continuous, high-frequency, noisy sensor signals and bridging the sim-to-real gap, particularly with proprioceptive, visual, or tactile inputs.
3. Publications in embodied AI, robotics, robot learning, or related fields.