Member of Technical Staff, Microsoft Robotics (Robotics Simulation)

Design, develop, and maintain physics-based simulation frameworks for robotics applications, including accurate modeling of rigid-body dynamics, articulated mechanisms, contact and friction, deformable objects, and fluid interactions as required by target robot platforms. Implement essential robotics simulation features, including accurate sensor models (cameras, LiDAR, IMUs, force/torque sensors, tactile arrays), actuator models, controller interfaces, and communication protocols that mirror real robot hardware behavior. Build real-to-sim and sim-to-real workflows for dynamic environments and robotics tasks, implementing domain randomization, system identification, and physics parameter tuning to minimize sim-to-real gaps. Create and maintain asset toolchains supporting industry-standard formats (USD, URDF, MJCF, SDF) and integrate with 3D content pipelines for environment and robot model ingestion. Develop simulation infrastructure for robust autonomy test and evaluation, enabling the use of rigorous test methods and design of experiments for validation and verification-based of robotics technologies and algorithms. Develop simulation infrastructure for robot learning policies, including reinforcement learning training at scale, with parallelized environment instances, reward instrumentation, curriculum management, and integration with distributed ML training frameworks. Collaborate closely with robotics engineers, ML researchers, and platform engineers to enable large-scale robotics development, training pipelines, benchmarking suites, and automated evaluation workflows. Lead architectural decisions for simulation platform selection, customization, and extension, evaluating and integrating engines such as MuJoCo, Isaac Sim, Gazebo, Genesis or proprietary solutions based on fidelity, performance, and scalability requirements. Review product code and test code to ensure it meets team standards for simulation accuracy, performance, and maintainability, coaching and providing feedback to develop other engineers' skills. Drive efforts to integrate logging and instrumentation for gathering telemetry data on simulation performance, fidelity metrics, and training pipeline health. Remain current in simulation and physics engine developments—as well as emerging technologies such as physical world models—proactively evaluating new trends, technical solutions, and research advances, and assessing how to adapt them to the robotics platform. Bachelor's Degree in Computer Science or related technical field AND 4+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python Master's Degree in Computer Science or related technical field AND 6+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python OR Bachelor's Degree in Computer Science or related technical field AND 8+ years technical engineering experience with coding in languages including, but not limited to, C, C++, C#, Java, JavaScript, or Python OR equivalent experience. Extensive experience in physics simulation engineering, with demonstrated expertise in rigid-body dynamics, contact modeling, or real-time physics engines (e.g., MuJoCo, Newton, PhysX, Bullet, DART). Experience with robotics simulation platforms such as Isaac Sim, MuJoCo, Gazebo, Drake, or equivalent, including sensor simulation and actuator modeling. Strong background in 3D graphics, real-time rendering, or game engine development (Unity, Unreal, or proprietary engines). Experience building sim-to-real transfer pipelines, domain randomization frameworks, or system identification workflows for robotic manipulation or locomotion. Proficiency in Python and C++ with experience in distributed computing, GPU-accelerated simulation, or parallel environment management for RL training. Familiarity with ML frameworks (PyTorch, JAX, TensorFlow) and their integration with simulation environments for end-to-end training pipelines. Experience with asset pipeline toolchains supporting URDF, MJCF, USD, or OpenUSD formats for robot and environment model management.