Perception Engineer — 3D Representation & Navigation

FieldAI’s Irvine team is where embodied AI meets real robots, real sensors, and real field deployments. Based in the heart of Southern California’s robotics ecosystem, we build risk-aware, reliable, field-ready AI systems that solve the hardest problems in robotics and unlock the full potential of embodied intelligence. If you want your work to ship, get tested on hardware, and improve through real deployments, Irvine is the place.

We go beyond typical data-driven approaches or pure transformer-only architectures, combining rigorous engineering with learning systems proven in globally deployed solutions that deliver results today and get better every time our robots run in the field.

We're looking for a Perception Engineer to help build the 3D scene and traversability representations our robots rely on to navigate real-world environments. You'll design learning-based approaches to spatial understanding from point cloud–based representations to learned traversability mapping and take them from research prototype to production on physical robots. This role sits at the intersection of perception, mapping, and autonomy, working closely with cross-functional teams to make sure our systems perform reliably in the field, not just in simulation.

1. Own 3D Representation for Navigation (core focus)

• Design and build the 3D scene/traversability representations our robots plan and act on — point cloud–based or learned implicit structures — optimized for real-time, on-robot use rather than offline reconstruction quality.
• Drive a long-term research agenda on learning-based approaches to building these representations — learned traversability, learned surface/reconstruction, geometry-aware embeddings — rather than only integrating existing classical pipelines.
• Stay close to the current literature on 3D scene representation for navigation — online navigation-mesh construction from streaming point clouds, learned elevation/traversability mapping, topologically-grounded navigation representations — and translate promising ideas into deployable systems.

2. Build and Maintain Perception Systems

• Design, implement, and maintain perception systems for autonomous robots operating in real-world environments.
• Develop localization and mapping capabilities that hold up in unstructured, off-road, and field conditions.
• Continuously evaluate and improve perception performance through testing, iteration, and field validation.

3. Develop and Integrate Sensor-Based Perception

• Implement perception algorithms that fuse data from multiple sensors — LiDAR, cameras, RADAR, inertial sensors.
• Support integration of new sensing modalities and configurations as platforms evolve.
• Ensure perception software behaves consistently across simulation and real-world deployment.

4. Deploy Perception Software on Real Robots

• Take representations and algorithms from research prototype to production on physical robots.
• Debug issues discovered during on-robot testing and field operations.
• Collaborate with autonomy, controls, and platform teams to integrate cleanly into the full autonomy stack.

5. Improve System Robustness and Scalability

• Contribute to code quality, testing, and long-term maintainability.
• Build tools, metrics, and regression tests for representation quality and downstream navigation performance.
• Help scale representation and perception solutions across multiple robots, environments, and missions.

6. Collaborate Across Teams

• Work with engineers, researchers, and field operators to define representation and perception requirements.
• Communicate technical tradeoffs clearly to both technical and non-technical stakeholders.
• Support field operations and customer demonstrations by keeping systems production-ready.