Job Summary
The Applied Robotics / Robot Learning Engineer will develop, optimize, and deploy reinforcement learning (RL) and Vision-Language-Action (VLA) models for complex robotic manipulation and manufacturing tasks. The role focuses on improving robot performance, sim-to-real transfer, failure recovery, and reliable execution through hands-on experimentation and deployment on physical robotic systems.
Key Responsibilities
Develop, fine-tune, and evaluate VLA models and learning-based policies on physical robotic systems.
Apply reinforcement learning techniques to improve manipulation success rates, execution efficiency, and recovery from failures.
Implement and evaluate RL approaches such as PPO, SAC, Residual RL, and policy adaptation.
Design reward functions, success detectors, and learned reward models for real-world manipulation tasks.
Apply sample-efficient training methods, including imitation learning, offline-to-online RL, and human-in-the-loop corrections.
Implement safe real-robot RL using action limits, workspace constraints, automated resets, and supervised exploration.
Identify and address sim-to-real challenges through domain randomization, system identification, and real-world fine-tuning.
Develop and evaluate policies for manipulation tasks, including bimanual and coordinated robot operations.
Conduct robot experiments, troubleshoot hardware and software issues, and validate model performance.
Analyze failure modes and evaluate success rates, cycle times, generalization, and repeatability.
Oversee data collection, identify data-quality issues, and coordinate improvements to demonstrations and datasets.
Collaborate with simulation teams on benchmarking, model evaluation, and engineering activities.
Required Qualifications
Ph.D. in Computer Science, Robotics, or a related discipline, or a Master's degree with at least 3 years of industry experience.
Strong applied expertise in reinforcement learning for robotics, including policy optimization, reward design, training, and evaluation.
Hands-on experience deploying or validating learning-based policies on physical robots; simulation-only experience is insufficient.
Experience with sim-to-real transfer and troubleshooting simulation-to-real-world performance gaps.
Proficiency in Python, PyTorch, Git, and ROS/ROS 2.
Experience with NVIDIA Jetson, robotics MLOps tools, experiment tracking, dataset versioning, and Foxglove.
Experience with sensor calibration, force-torque sensors, or teleoperation data collection.
Understanding of imitation learning, robot kinematics, control interfaces, and manipulation workflows.
Ability to work directly with robotic hardware and independently manage experiments, model training, evaluation, debugging, and deployment.
Preferred Qualifications
Experience combining reinforcement learning with pretrained VLA models or imitation learning.
Experience with Residual RL, online policy adaptation, failure recovery, and contact-rich manipulation.
Familiarity with VLA architectures such as Pi0/Pi0.5, OpenVLA, and Groot.
Experience with industrial or collaborative robots and real-world manufacturing applications.
Hands-on experience with FANUC or Universal Robots (UR).