Skip to main content

Field AI

Perception Engineer — 3D Representation & Navigation

Irvine, CA$135k–$190kfull timemidAdded today

About this role

FieldAI seeks a Perception Engineer to develop 3D representation and navigation systems for embodied AI robots deployed in real-world field conditions. You'll combine classical robotics engineering with modern learning approaches to build reliable perception pipelines that enable autonomous systems to understand and navigate complex environments.

What you'll do

  • Design and implement 3D scene representation systems for robot perception
  • Develop navigation algorithms that integrate visual and sensor data
  • Optimize perception pipelines for deployment on resource-constrained hardware
  • Validate algorithms on physical robots in field environments
  • Collaborate with roboticists and ML engineers to improve system reliability
  • Debug and refine perception systems based on real-world deployment feedback

What they're looking for

  • 3D computer vision and scene understanding
  • Robot navigation and path planning
  • Sensor fusion and calibration
  • Python and C++ programming
  • ROS (Robot Operating System)
  • SLAM or visual odometry experience
  • Deep learning frameworks (PyTorch, TensorFlow)
  • Hardware integration and embedded systems
Apply with Autofill

Opens the application — the Jobs AI extension fills it for you. Set up autofill

Opens the official application on the employer’s site. No login required.

Field AI

Field AI develops embodied AI and autonomous robotics systems for real-world deployment in industrial environments like oil & gas and mining. The company is hiring software engineers to build web-based systems, perception and validation pipelines, test infrastructure, ROS-based robotic software, and customer-facing products that integrate AI with field-deployed hardware.

View all jobs at Field AI

Likely interview questions

  • Describe your experience building perception systems for physical robots—what were the biggest gaps between simulation and real-world performance?
  • How do you approach designing 3D representations that balance accuracy with computational efficiency on edge devices?