Research Engineer, Robotics, DeepMind

Google•Cambridge, MA

5h•$147,000 - $210,000

About The Position

At Google, research-focused Software Engineers are embedded throughout the company to set up large-scale tests and deploy promising ideas. They work on real-world problems, prototype implementations, design architectures, and stay connected to the research community through collaboration with universities and paper publication. As a Robotics Research Engineer, you will connect physical and simulated robot platforms to the research and machine learning software stack. The role focuses on building a unified software interface across physics simulation and physical hardware so that models, teleoperation systems, and evaluation tools run in both domains. Working in the Cambridge lab alongside hardware, simulation, and machine learning teams, you will bring up robotic platforms, end effectors, and wearable devices and prepare them for research. Google DeepMind is an interdisciplinary AI lab focused on advancing AI to address complex global challenges and support product innovation. The organization emphasizes public benefit, scientific discovery, safety, and ethics. Individual pay is determined by job-related skills, experience, and relevant education or training. US compensation is $147,000–$210,000, plus a 15% bonus target, equity, and benefits.

Requirements

Bachelor's degree in Computer Science, Computer Engineering, Robotics, or equivalent practical experience.
At least 3 years of experience developing, integrating, and debugging software for simulated and physical robotic systems.
Experience with physics simulation environments, such as MuJoCo, NVIDIA Isaac Sim, or Gazebo, for modeling robots, testing software stacks, or validating sim-to-real transfer.
Experience integrating sensor streams (cameras, IMUs, encoders, or tactile sensors), actuator interfaces, and real-time control loops into a robotics software stack.

Nice To Haves

At least 5 years of experience building core components of a robotics software stack, SDK, or hardware abstraction layer.
Experience developing, deploying, or running machine learning policies, such as Vision-Language-Action models or reinforcement learning controllers, on physical robots.
Experience with system identification, actuator and sensor modeling, or closing the sim-to-real gap for manipulation or locomotion.
Experience with teleoperation systems, motion tracking hardware, wearable data collection devices, or multi-fingered dexterous hands.
Experience designing automated calibration verification, system health monitoring, or hardware-in-the-loop (HIL) test pipelines.

Responsibilities

Integrate robot control software and hardware interfaces with the common robotics SDK across simulated and physical platforms, end effectors, and wearable devices.
Build, validate, and maintain physics simulation environments and software-in-the-loop harnesses that accurately model the on-robot software stack, control loops, and hardware interfaces.
Run and maintain the on-robot execution harness to call local and remote machine learning action models in real time.
Connect platforms to the teleoperation and data logging stack to capture synchronized sensor streams, robot actions, and demonstrations for model training.
Develop platforms with fleet orchestration and evaluation tools to run standardized benchmarks and guide data collection.