Senior Robotics Software Engineer, Simulation

Bonsai Robotics • San Jose, CA

12h • $180,000 - $215,000 • Hybrid

About The Position

Bonsai Robotics is seeking a Senior Robotics Software Engineer to build and own its simulation platform. The platform supports robots operating in orchards, vineyards, and bedded crops under changing light and weather conditions and across multiple vehicle form factors. Simulation is used to regression-test software releases, reproduce and resolve field failures, and generate synthetic data for perception models. This role owns the platform end to end, from vehicle, sensor, and environment models through CI pipelines and metrics used to assess whether changes are safe to ship. The goal is to ensure that software that passes simulation also works in the field.

The role involves creating realistic simulation environments and integrating them with the ROS 2 autonomy stack. The engineer will develop software-in-the-loop (SIL) and hardware-in-the-loop (HIL) frameworks to run perception, planning, and controls in a closed loop; convert field logs and failures into repeatable regression scenarios; and develop sensor and vehicle models calibrated against real hardware and vehicle data. The role also includes building scenario generation, evaluation, and reporting into CI, keeping simulation fast and reproducible, assessing sim-to-real accuracy, and generating synthetic perception-training data in collaboration with the ML team.

The engineer will work with perception, controls, and platform teams to define simulation requirements and prioritize important sim-to-real gaps. Bonsai describes its workplace as hybrid for selective roles, with flexibility to split time between the office and home. The company also describes a supportive, collaborative, and mission-driven culture focused on robotics innovation and impact.

Requirements

• Bachelor's, master's, or PhD in Computer Science, Robotics, or a related field.
• At least 5 years of industry experience in robotics software engineering, with a significant focus on simulation.
• At least 2 years building SIL, HIL, or closed-loop evaluation systems used in production.
• Strong computer science fundamentals and systems fluency.
• Strong proficiency in C/C++ and Python, and ability to read and debug a large existing codebase.
• Experience building simulation environments using platforms such as NVIDIA Isaac Sim, MuJoCo, Gazebo, or similar.
• Familiarity with physics engines such as PhysX, Bullet, or MuJoCo.
• Deep understanding of ROS 2.
• Experience integrating simulation into CI/CD pipelines.
• Pragmatic engineering approach: choose the simplest model that answers the question and avoid over-engineering.
• Strong debugging, testing, and problem-solving skills, and ability to work independently and collaboratively.

Nice To Haves

• Rust programming experience.
• Exposure to agricultural, construction, or off-road robotics.
• Experience with digital twins or synthetic data generation.
• Experience using world models, such as NVIDIA Cosmos or similar generative video models, for synthetic data generation or scenario augmentation.
• Experience with neural reconstruction, including 3D Gaussian Splatting, of scenes from real sensor data.
• Familiarity with USD/OpenUSD scene formats.
• Experience with sim-to-real transfer techniques, including domain randomization.
• GPU-accelerated rendering or ray tracing for camera simulation.
• Experience with NVIDIA Jetson or other embedded deployment.

Responsibilities

• Build and maintain simulation environments representing field conditions, including orchards, vineyards, bedded crops, uneven terrain, and variable lighting and weather.
• Own the interface between the simulator and the ROS 2 autonomy stack, and build SIL and HIL frameworks that run the full perception, planning, and controls stack in closed loop.
• Replay recorded drives, reproduce field failures as regression scenarios, and expand a scenario library to cover real-world edge cases.
• Create and maintain camera, GNSS/RTK, IMU, and LiDAR sensor models, including noise and failure modes calibrated against real hardware.
• Model vehicle dynamics and terrain interaction for multiple robot form factors using system identification against real vehicle data.
• Build automated scenario generation, evaluation metrics, and reporting so autonomy-stack changes are tested in simulation in CI before deployment to a machine.
• Keep simulation fast, deterministic, and reproducible, and measure how well simulation results predict real-world behavior.
• Generate synthetic data for perception training and collaborate with the ML team on domain randomization and generative methods.
• Work with perception, controls, and platform teams to define simulation requirements and prioritize the most important sim-to-real gaps.

Benefits

• Hybrid flexibility for selective roles, including the option to split time between the office and home.
• Covered lunch on in-office days.
• Comprehensive health insurance, including medical, dental, and vision coverage.
• Mental health support.
• Wellness programs.
• Supportive work environment focused on work-life balance.

What This Job Offers

• Job Type: Full-time
• Career Level: Senior