Research Engineer, Robotics: Physics Simulation and Sim-to-Real

The Allen Institute for Artificial Intelligence•Seattle, WA
•$128,880 - $193,320•Onsite

About The Position

Ai2’s Robotics team is hiring a Research Engineer to own physics simulation stability and sim-to-real fidelity for robot foundation model research. The team develops scalable simulation domains and real-world capable robots, training and evaluating policies across MuJoCo, Isaac Sim/Isaac Lab on Newton, and other GPU simulators, then deploying them on bimanual and single-arm manipulators. The work supports manipulation, large-scale simulated training, sim-to-real transfer, robot foundation models, and skill learning, with research published and tools, models, and data released openly. The role improves contact, friction, and constraint solvers; diagnoses instability and non-determinism; makes articulated robots and closed-loop mechanisms stable across simulation engines; and measures the accuracy-throughput trade-offs of simulation settings at scale. It also builds stress and regression tests and prepares assets for simulation across URDF, MJCF, and OpenUSD. The engineer will calibrate simulation models against logged robot data, including dynamics, contact, payload, actuator, and latency parameters. They will maintain cross-simulator consistency and a sim-to-real correlation suite that replays recorded trajectories, reports quantitative errors, detects regressions, and informs domain-randomization ranges. The position also includes documenting robot and scene model parameters and their sources and confidence. The role is expected to work from Ai2’s Seattle offices; on-site requirements vary by position and team. Ai2 is a Seattle-based nonprofit AI research institute and emphasizes learning, inclusion, collaboration, and work/life balance. The stated base salary range is $128,880–$193,320, with bonus plans in addition to base salary.

Requirements

  • PhD in robotics, mechanical engineering, applied mathematics, physics, computer graphics, or a related field; alternatively, a Master’s degree plus 3 years of experience. Exceptional applicants with a bachelor’s degree and 3 or more years of relevant industry experience are also welcome.
  • Demonstrated depth in rigid-body dynamics and constrained multibody systems, including reduced- and maximal-coordinate articulations, contact and friction modeling, and numerical integration of stiff systems.
  • Solver or physics-engine code authored by the candidate in a research codebase, production engine, or widely used open-source simulator such as MuJoCo, Newton/Warp, PhysX, Drake, Bullet, or Genesis.
  • Track record diagnosing and fixing simulation instability, such as jitter, penetration, energy drift, or divergence, with an understanding of why the fixes work.
  • Experience fitting simulation parameters to measured data and quantitatively validating results against real-world recordings.
  • Proficiency in Python and C++.
  • Experience with robot description formats such as URDF, MJCF, and OpenUSD.
  • Knowledge of at least one major simulation stack at source level.
  • Ability to communicate information and ideas clearly and exchange accurate information.
  • Ability to observe details at close range.
  • Ability to work under deadlines.
  • Ability to remain in a stationary position for long periods of time.

Nice To Haves

  • GPU programming for physics using CUDA, NVIDIA Warp, or JAX, and experience scaling simulation for reinforcement or imitation learning.
  • Experience with deformable, cable, or FEM simulation, including XPBD, VBD, corotational or hyperelastic FEM, or Cosserat rods, and coupling these with rigid bodies.
  • Familiarity with differentiable simulation, Bayesian or gradient-based parameter estimation, or learned residual dynamics.
  • Experience in robot manipulation, grasping, or contact-rich tasks in simulation or on hardware.
  • Publications or open-source contributions in simulation, dynamics, or sim-to-real transfer, including at CoRL, RSS, ICRA, SIGGRAPH/SCA, or engine repositories.

Responsibilities

  • Design, implement, and improve contact, friction, and constraint solvers, including complementarity and soft-constraint formulations, friction cones, warm-starting, iteration counts, and tolerances.
  • Diagnose and resolve simulation failure modes including resting-contact jitter, interpenetration, tunnelling, energy gain, constraint drift, instability under stiff joint drives, and non-determinism across runs and GPUs.
  • Improve stability and accuracy of articulated robots and closed loops across MuJoCo, Newton, and PhysX, including parallel grippers, mimic and coupled joints, and tendons; contribute fixes upstream to open-source engines.
  • Select integrators, timesteps, and substeps based on evidence, measure convergence, and quantify accuracy-throughput trade-offs for GPU-batched training across many environments.
  • Build stress and regression tests for long-horizon rollouts, randomized scenes, extreme parameters, and contact-rich edge cases.
  • Prepare simulation-ready assets by addressing collision geometry, inertial properties, joint limits, and contact settings, consistently across URDF, MJCF, and OpenUSD.
  • Fit simulation models to logged hardware data, including joint friction, armature and reflected inertia, payload mass and center of mass, actuator drive gains, and latency; identify missing simulator behavior rather than hiding it in other parameters.
  • Calibrate contact behavior against recorded pushes, slides, grasps, and handovers.
  • Maintain consistency for robot models across simulators, including a common source of truth for frames, masses, and contact settings, and explain remaining discrepancies.
  • Own a sim-to-real correlation suite that replays recorded trajectories in each simulator and reports per-joint RMS, onset lag, velocity profiles, and contact-force error against predefined thresholds.
  • Catch regressions when a solver, asset, or parameter changes; convert measured uncertainty into domain-randomization ranges and measure how simulated policy success relates to real-world success.
  • Write a model card for each robot and scene documenting parameters, their provenance, and confidence.
  • Perform other job-related duties as assigned.

Benefits

  • Medical, dental, and vision coverage for team members and their families.
  • Basic life insurance and basic accidental death and dismemberment insurance.
  • Short-term and long-term disability coverage.
  • Employee assistance program.
  • Voluntary life insurance and voluntary accidental death and dismemberment programs.
  • Health savings account, healthcare reimbursement arrangement, and healthcare and dependent care flexible spending account plans.
  • Company 401(k) plan.
  • $125 per month for commuting or internet expenses.
  • $200 per month for fitness and wellbeing expenses.
  • Up to ten sick days per year, up to seven personal days per year, up to 20 vacation days per year, and twelve paid holidays per year.
  • Annual bonuses and participation in the long-term incentive plan.
  • Generous paid vacation and sick leave as well as family leave.
  • Ongoing learning and education opportunities.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service