Remote job
Machine Learning Engineer, Robotics
Job details
About this role
Role overview
Develop machine learning capabilities that help production robots plan, scan and act in real customer environments. You will turn practical robotics challenges into measurable learning problems, train models using fleet and simulation data, and take them from prototype through deployment and monitoring alongside established planning and perception methods.
Responsibilities - Define machine learning problems for robot planning, scanning and behavior. - Train models that complement classical approaches, such as learned trajectory costs, plan quality predictors and collision models. - Build viewpoint planning policies that choose where and how to scan and estimate scan quality before execution. - Apply imitation learning, reinforcement learning or generative methods when they outperform classical baselines. - Build production data and training pipelines using fleet logs and simulation, with clear metrics. - Evaluate performance in the loop, analyze failure modes and calibrate confidence so systems can fall back to classical methods when needed. - Own deployment of models to live robotic cells, including low-latency inference and monitoring.
Requirements - Master’s or doctoral degree in computer science, robotics, machine learning or a related field. - At least two years of applied machine learning experience in robotics or autonomy, in research or industry. - Strong Python skills and experience using PyTorch or JAX to train, evaluate and deploy deep learning models. - Solid foundations in linear algebra, 3D geometry, kinematics and optimization. - Experience running methods on physical robot hardware. - Ability to explain model performance and limitations clearly to planning, perception and field teams.
Nice to have - C++ and experience deploying into a C++ or ROS 2 system; motion planning, trajectory optimization or collision checking. - Background in active perception, 3D point clouds, uncertainty calibration or information-theoretic planning. - GPU inference optimization, simulation and sim-to-real transfer, or relevant research publications.
Benefits and work setup - Daily lunch, flexible paid time off, medical, dental and vision coverage, and a retirement plan. - Six weeks of paid parental leave, with an additional six to eight weeks for birthing parents.