Sim2Real Robotics Research Scientist

Robotics and AI Institute•Cambridge, MA
•$140,910 - $326,040

About The Position

The RAI Dexterous Manipulation team develops advanced manipulation systems for precise, contact-rich tasks at or above human cadence. It co-designs hardware, software, controls, and machine learning components to enable reliable, high-performance skills for complex manufacturing and construction tasks intended for real-world deployment. The Sim2Real Robotics Research Scientist supports the development, deployment, and testing of policies on novel robotic hardware. The role works closely with researchers and engineers on high-fidelity simulation-based reinforcement learning for dexterous manipulation, with a focus on testing and debugging learned policies on hardware, reducing simulation-to-real gaps, and developing testing techniques that support high-throughput experimentation.

Requirements

  • M.S. or Ph.D. in Robotics, Computer Science, Machine Learning, or a related field.
  • Hands-on expertise in real-world robot integration, hardware deployment, and physical testing setups.
  • Experience developing, adapting, fine-tuning, or deploying multimodal policies that integrate vision with tactile or force/torque sensing.
  • Solid background in deep reinforcement learning algorithms, simulation-to-real transfer methodologies, reward engineering or shaping, and curriculum design.
  • Experience incorporating state-of-the-art workflows and techniques to automate robotics training loops, simulation environments, or simulation-to-real transfer tuning.
  • Demonstrated experience applying simulation-to-real techniques to manipulation or locomotion.
  • Proficiency in Python, deep learning frameworks, and physics simulation engines such as PhysX, MuJoCo, or Newton, applied to complex robotic systems.

Nice To Haves

  • Direct experience with NVIDIA IsaacLab.
  • Experience applying reinforcement learning techniques to dexterous robotic manipulation.
  • Explicit experience investigating and applying simulation-to-real methods to dexterous, contact-rich robotic manipulation.
  • Expertise in simulation-to-real reinforcement learning for locomotion and whole-body loco-manipulation.

Responsibilities

  • Develop, deploy, test, and debug reinforcement learning policies on real-world novel robotic hardware platforms for dexterous manipulation tasks.
  • Investigate, analyze, and mitigate simulation-to-reality gaps in contact-rich manipulation using parameter identification, dynamic domain randomization, and real-to-simulation alignment.
  • Design and implement curriculum training frameworks, demonstration-led state guidance, and reward engineering strategies to address exploration bottlenecks in tight-tolerance manipulation tasks.
  • Integrate vision, tactile and contact sensing, six-axis force/torque sensing, and joint proprioception into real-time sensorimotor learning architectures and policies.
  • Explore policy-learning techniques for highly dynamic, contact-rich action spaces that regulate speed, impedance, and energy expenditure.
  • Develop high-throughput experimental workflows and collaborate with research scientists, hardware engineers, and control teams to co-design closed-loop learning systems and maintain technical documentation.

Benefits

  • Annual bonus may be included in the total compensation package.
  • Long-term cash incentive may be included in the total compensation package.
  • Monthly cell phone cost reimbursement may be included in the total compensation package.
  • Commuter cost subsidy may be included in the total compensation package.
  • Medical benefits may be included in the total compensation package.
© 2026 Teal Labs, Inc
Privacy PolicyTerms of Service