Robotics Engineer (Perception)

Sabanto•Itasca, IL

About The Position

Sabanto is a venture-backed, Illinois-based technology company focused on bringing autonomy to agriculture through a robotic machinery platform and Farming-as-a-Service. The Perception-focused Robotics Engineer will build sensing and perception systems that enable autonomous tractors and other agricultural equipment to see and understand their surroundings. The role takes real sensors from bench development into production field environments, including challenging conditions such as dust, glare, rain, and crop canopy, and turns their data into a dependable world model for navigation, safety, and control systems. The engineer will design and harden perception pipelines, integrate and calibrate sensors, fuse data, deploy computer vision models on edge hardware, and develop data collection and evaluation processes. The role also involves investigating field issues, validating changes with real sensor data and on-vehicle testing, and occasionally visiting customer sites to observe and refine systems. Sabanto describes its work culture as fast-paced and demanding, with flexibility, autonomy, and career opportunities.

Requirements

Bachelor's degree in computer science, robotics, electrical engineering, or a related field.
Strong quantitative, analytical, and problem-solving skills, including comfort with geometry, linear algebra, and probabilistic estimation.
Entrepreneurial attitude and keen attention to detail.
Ability to code with and without AI assistance.
Organized and disciplined, with the ability to manage multiple priorities simultaneously.
Industrious; ideally able to drive, or willing to learn to drive, tractors and other four-wheeled vehicles.
Interest in learning the agriculture industry, understanding farmers' needs, and helping them succeed with Sabanto's offerings.
Applicants must be legally authorized to work in the United States; Sabanto is unable to sponsor work visas.

Nice To Haves

Five or more years building perception, computer vision, or other real-time sensing applications, with a track record of shipping to production hardware outside the lab.
Mastery of C++ and/or Python.
Mastery of ROS/ROS2, including sensor drivers and high-bandwidth message pipelines.
Hands-on experience integrating and calibrating real cameras, LiDAR, and/or radar in production environments.
Experience with computer vision and point cloud libraries such as OpenCV, PCL, or Open3D.
Experience training and deploying deep learning models for detection or segmentation on edge hardware, using tools such as PyTorch, ONNX, TensorRT, or NVIDIA Jetson.
Experience with sensor fusion, state estimation, and multi-object tracking, such as Kalman filters.
Experience working with geospatial data.
Familiarity with functional safety standards for autonomous machinery, such as ISO 25119, ISO 18497, or ISO 13849.
Experience working with Docker-based software and build pipelines.

Responsibilities

Design, build, and harden perception pipelines for obstacle detection, classification, and tracking to support safe autonomous agricultural equipment operation.
Integrate cameras, LiDAR, radar, and GNSS/IMU sensors onto tractors and implements, including drivers, intrinsic and extrinsic calibration, and time synchronization.
Fuse multi-sensor data into a dependable world model for navigation, safety, control, and sequencing systems.
Evaluate, select, and integrate third-party sensors and perception vendors from bench prototype through production deployment.
Deploy and optimize computer vision models on edge compute within real-time latency, thermal, and power constraints.
Build data collection, labeling, and evaluation pipelines using field logs to measure and improve performance across crops, seasons, lighting, and weather.
Triage, reproduce, and resolve field perception issues using recorded sensor data.
Travel occasionally to customer sites to observe and refine perception systems.
Write tested, hardened, and safe code, including unit tests, replay tests against real field data, and on-vehicle validation before changes ship.