Build the first system
Perception Engineer
Build calibrated field perception and the data engine behind person safety, navigation, and machine awareness.
Apply nowThe mandate
Give the platform dependable awareness of people, rows, traversable space, obstacles, payload state, and operating context. Build the sensor, labeling, evaluation, and edge-inference system required to improve from every field run.
First-year outcomes
Evidence the role is working.
- 01
A versioned field dataset and evaluation suite organized around operational failure modes.
- 02
Calibrated person and obstacle detection across representative light and field conditions.
- 03
Traversability and row-understanding outputs that integrate cleanly with navigation.
- 04
An on-robot inference pipeline with measured latency, resource use, and confidence behavior.
What you will own
- Select and calibrate cameras and complementary sensors with the systems team.
- Develop perception for people, obstacles, rows, free space, and machine or payload state.
- Build data capture, triage, labeling, dataset versioning, and regression evaluation.
- Quantify coverage gaps by time of day, weather, field, and failure category.
- Optimize inference for the deployed compute and observe model health in the field.
- Define how uncertainty changes vehicle behavior with autonomy and safety owners.
What we are looking for
- Has shipped perception or machine-learning systems whose mistakes affected real-world behavior.
- Understands calibration, data quality, confidence, latency, and distribution shift.
- Can build the data and evaluation pipeline, not only train a model.
- Comfortable optimizing models and sensor processing for edge compute.
- Investigates false positives and false negatives in operational terms.
The field reality
The environment is part of the design.
The same row looks different by hour, weather, variety, canopy, and camera cleanliness. The goal is not benchmark accuracy in a curated dataset; it is calibrated perception that supports safe decisions and exposes uncertainty.
- Computer vision, multimodal sensing, 3D geometry, or edge ML.
- Outdoor, automotive, warehouse, construction, or agricultural perception.
- Dataset tooling, active learning, simulation, or synthetic data.
- Camera, lidar, radar, depth, or thermal sensor integration.
Apply for this role
Upload your résumé.
We are actively hiring for Perception Engineer. A professional link is optional.