Robotics Software Engineer

Terabase Energy•Woodland, CA

2d•$130,000 - $160,000•Hybrid

About The Position

Terabase Energy builds the automation and robotics systems that are transforming how utility-scale solar power plants are built. Our platforms combine robotics, controls, and software to increase construction speed, quality, and safety across the solar EPC lifecycle. As physical automation becomes central to how we deliver projects, we are building out a Physical AI capability that pairs machine perception and decision-making with the mechanical and controls systems already deployed in the field.

Requirements

Bachelor's degree in Computer Science, Electrical Engineering, Mechanical Engineering, Robotics, or a related field
5+ years of hands-on experience shipping automated, mechatronic, or robotic systems into production — spanning real-time control, perception, and the software that connects them
Proficiency in Python and computer vision tooling (OpenCV or equivalent); experience with deep-learning frameworks (PyTorch, TensorFlow) is a plus but not required
Working understanding of how AI perception connects to physical systems: sensors, actuators, motion control, PLCs, or robotic arms and end effectors
Experience with robotic middleware (ROS/ROS2) and/or industrial communication protocols (EtherCAT, EtherNet/IP, CAN, Modbus)
Demonstrated experience designing systems for reliability and supportability in harsh, uncontrolled, or outdoor environments (construction, industrial field equipment, automotive, aerospace, or similar) — not just lab or data-center conditions
Willingness and ability to travel up to 30%
Prefer experience in industrial manufacturing and/or construction industry
Self-starter, able to thrive in a fast-paced and continually changing environment.
Strong communication, customer relationship skills, and ability to communicate effectively and interact within a team environment.
Proven skill in MS Suite of software (Outlook, Excel, PowerPoint, etc.)

Nice To Haves

Experience with industrial machine vision systems (e.g., Cognex, Keyence) or camera/lighting setup for QC applications
Experience with edge AI deployment tooling (TensorRT, Jetson, or similar embedded inference platforms)
Use of robotic simulation software (e.g., KukaSim, Roboguide, Visual Components, NVIDIA Isaac Sim, Gazebo) for sim-to-real workflows and offline testing
Exposure to reinforcement learning, imitation learning, or vision-language-action (VLA) / generalist robot policy models
Experience taking an AI-enabled product from concept to shipped, including the software layer around it (web UI, APIs, data pipelines)

Responsibilities

Partner closely with the Controls and Mechanical Engineering teams on automated work cells, PLC-controlled systems, and end-effector tooling — owning the full loop from real-time control and perception through the operator-facing software that runs on the equipment
Engineer AI/vision systems to be robust and supportable in outdoor, active solar-construction environments — accounting for dust, temperature extremes, vibration, moisture, and intermittent connectivity — and maintainable by field personnel without ML expertise
Design, train, and deploy computer vision models (object detection, defect/anomaly detection, segmentation) for real-time material and component QC on automated assembly and production lines
Build perception pipelines that fuse camera, LiDAR, and other sensor data, and deploy low-latency inference at the edge (e.g., NVIDIA Jetson or similar) to support closed-loop equipment control and real-time vision-guided actions
Develop and validate defect-classification and QC models against production tolerances — including comprehensive model and system testing — working with manufacturing and quality engineering to define acceptance criteria
Collect, label, and manage training data from factory and field environments; build the data pipelines and tooling needed for continuous model improvement
Instrument automated cells and production lines with the telemetry needed to evaluate AI system performance, uptime, and impact on cycle time, scrap rate, and yield
Build lightweight operator-facing tools — dashboards, web UI — so field teams can monitor and interact with AI-enabled systems without ML expertise
Support integration, commissioning, and on-site troubleshooting of AI-enabled automation equipment in test and production environments, working closely with controls engineers, and document system performance and failure modes for cross-functional stakeholders
Support risk assessment and EHS review of newly designed AI-enabled machines, ensuring compliance with safety protocols