Real-Time Perception Engineer

Path Robotics, Inc.; Columbus (OH); On-site; USD 130,000 - 190,000; Full time

Benefits offered by this job

Daily free lunch

Flexible PTO

Medical, dental, and vision coverage

6 weeks parental leave + 12–14 weeks (

Job summary

Path Robotics, Inc. is building an elite robotics and perception team and seeks a hands-on Perception Engineer to own the real-time vision loop on embedded hardware. You will tackle end-to-end perception, from sensing to enabling the control loop, with a focus on reliability under degraded imagery and rigorous latency measurement.

You will defend sensing choices with data, push toward learned methods, and collaborate with hardware and software teams to shape the project’s technical direction.

Qualifications

• Real-time loop on camera data in hardware with measured latency end-to-end.
• First-principles depth in classical vision and 3-D geometry and error propagation.
• Production-level C++ and ROS 2 experience; capable of defending design choices with data.
• Experience with degraded imagery and robust feature extraction under constraints.

Responsibilities

• Own the real-time perception pipeline end-to-end, from sensing to the output used by control loops.
• Evaluate sensing options, run experiments, and defend choices with numbers and measurements.
• Ensure robustness when signals are degraded and manage per-frame budgets on embedded hardware.
• Set technical direction as perception evolves toward learned methods.

Skills

Real-time vision

3-D geometry

End-to-end perception

Uncertainty and error propagation

Camera-to-control loop experience

Tools

C++

ROS 2

GPU-accelerated inference

Job description

Path Robotics, Inc. is building an elite robotics and perception team and seeks a hands-on Perception Engineer to own the real-time vision loop on embedded hardware. You will tackle end-to-end perception, from sensing to enabling the control loop, with a focus on reliability under degraded imagery and rigorous latency measurement.

You will defend sensing choices with data, push toward learned methods, and collaborate with hardware and software teams to shape the project’s technical direction.