Robotics Engineer - Manipulation focus

1 мес. назад

СШАOnsite

reinforcement learningsimulationdata collectiontrainingevaluationrobotic armsmobile manipulators

Engineer robotics manipulation methods for autonomous robots, from simulation to real hardware deployment.

We're hiring Robotics Engineers to bring state-of-the-art manipulation methods into the GRID platform. You'll work at the frontier of autonomous manipulation, building production-grade policies that generalize across embodiments — from fixed-base robotic arms and mobile manipulators to humanoids. This is a role for engineers who want their research to ship. You'll move between training runs, simulation, and real hardware in customer environments.
General Robotics is building the intelligence grid for physical AI — the platform that makes any robot, from robotic arms to humanoids, genuinely intelligent. Headquartered in Redmond, Washington, we're venture backed, including by Accenture, who invested in General Robotics in 2026 to advance Physical AI-powered robotics in manufacturing and logistics, and we're also part of Microsoft's Startups Pegasus Program. Our team's work spans some of the most widely adopted robotics and AI research to come out of Microsoft Research, Google Research and DeepMind — including AirSim, PACT, ClimaX, Tensorflow Object Detection and VideoPoet.

Design methods, tools, and infrastructure to push forward the state of the art in robotics and foundation models.
Define research goals informed by practical engineering concerns.
Train, adapt, and deploy manipulation foundation models — vision-language-action (VLA) policies, world action models (WAMs), reinforcement learning, and agentic robot policies.
Build synthetic data generation pipelines on high-fidelity simulation platforms, and drive sim-to-real transfer.
Develop data collection, training, and evaluation pipelines as production-quality code: extensible, well-documented, and usable by engineers outside your team.
Run experiments end to end — experimental design, reusable implementations, model evaluation, and clear reporting of results.
Contribute to publications and open-source releases.
Partner with customers to take manipulation capabilities from prototype to deployed product.

Bachelor's degree in Computer Science, Computer Engineering, relevant technical field, or equivalent practical experience.
Research experience in machine learning, robotics, and computer vision.
Experience with developing robotics algorithms or machine learning models at scale.
Programming experience in Python/C++. Good understanding of deep learning frameworks like Pytorch or Jax.
Must obtain work authorization in the country of employment at the time of hire, and maintain ongoing work authorization during employment.
MS or PhD in Robotics, Computer Science, Computer Engineering, or a related technical field.
Direct research experience in robot learning, computer vision, or machine learning.
First-author publications at peer-reviewed AI and robotics venues (NeurIPS, ICML, ICLR, CVPR, ICRA, IROS, CoRL).
Hands-on experience with modern manipulation policies — VLAs, world action models, reinforcement learning, imitation learning, diffusion or flow-matching policies, or agentic robot policies.
Experience with high-fidelity simulators such as NVIDIA Isaac Sim, MuJoCo, ManiSkill, AirSim, or CARLA, including domain randomization and sim-to-real workflows.
Familiarity with modern perception stacks and sensing modalities: RGB and RGB-D cameras, LiDAR, 3D representations (point clouds, meshes), and segmentation, detection, and tracking.
Judgment about systems constraints — latency, throughput, compute budget, memory — and the ability to factor them into model and architecture choices.

This role is open to candidates currently based in and authorized to work in the US.