Senior Robotics Perception Engineer

Senior Robotics Perception Engineer

As a Senior Robotics Perception Engineer at AgriRobot, you will own and advance the perception capabilities behind our obstacle detection and geofencing safety functions, ensuring the reliability, robustness, and safety of autonomous agricultural machinery operating in harsh, unstructured field environments. In this role, you will lead the development, testing, and optimization of the perception stack that fuses sensors such as 3D LiDAR, 4D imaging radar, monocular cameras, GNSS, and INS into production safety systems deployed across agricultural OEM platforms, from tractors to specialized robots.

Your Responsibilities

In this role, you will:

• Implement multi-sensor fusion across 3D LiDAR, 4D imaging radar, monocular cameras, and inertial sensors to detect, classify, and track obstacles, and to distinguish path hazards from vegetation and traversable terrain.
• Develop robust localization and state estimation capabilities, fusing GNSS, INS, and 4D imaging radar to perform reliably in open-field, orchard, and row-crop environments.
• Design and implement diagnostics and safety monitors for anomaly detection across the perception stack.
• Support the prototyping and integration of perception software and sensor kit configurations into a variety of OEM agricultural machines.
• Continuously evaluate and improve perception performance through testing, iteration, and field validation against defined requirements and real-world conditions.
• Support field operations and OEM demonstrations, providing perception expertise to ensure systems are production-ready.
• Communicate technical designs, tradeoffs, and results clearly to both technical and non-technical stakeholders.
• Promote a collaborative, growth-oriented engineering culture through mentorship and knowledge-sharing, contributing actively within an agile Scrum team.

Your Qualifications

You likely have an MSc in Robotics, Computer Science, Computer Engineering, Electrical Engineering, or a related field, with significant experience developing robot perception for real-world systems. A PhD in a related field is also welcome.

Essential qualifications include:

• Minimum 4 years of relevant experience, working with autonomous systems.
• Experience with ROS / ROS 2-based classical and machine-learning systems for sensor fusion, state estimation, and localization (e.g., GNSS/INS, EKF, LIO/RIO/VIO, SLAM).
• Proficiency in modern C++ and Python, applying software architecture and design patterns to build perception systems that scale across multiple platforms and product variants.
• Familiarity with quality-managed software development processes, CI/CD pipelines, regression testing, and benchmarking for perception systems.
• Hands-on experience developing and deploying perception software for real-world robotic systems. Comfortable working close to hardware, including debugging sensor data, sensor calibration, and diagnosing real-world failures.
• Hands-on experience with one or more of 3D LiDAR, imaging radar, or camera-based perception, including point cloud processing.
• Ability to take ownership of complex subsystems, from design through field deployment and maintenance.
• Excellent problem-solving skills and the ability to take ownership and drive features forward in a dynamic, fast-paced, collaborative small-team environment.

It is an advantage to have:

• Experience in agricultural, off-road, or heavy machinery domains.
• Familiarity with functional safety standards (e.g., ISO 25119, ISO 26262, ISO 21448/SOTIF, IEC 61508) and experience with safety-critical functions.
• Experience developing automotive L2+ ADAS or autonomous driving perception, including exposure to open-source AD stacks and platforms (e.g., Autoware).
• Experience integrating and calibrating custom multi-sensor rigs (e.g., multi-LiDAR, radar, and camera setups).
• Experience optimizing high-throughput data processing pipelines in embedded high-performance computers (HPC), including multi-processors and ML inference accelerators.
• Experience with 4D imaging radar and radar point cloud / signal processing.
• Familiarity with PCL, Eigen, Boost.Geometry, or similar libraries for geometric and point cloud processing.

Why This Role

• High-impact, real-world robotics work. You'll build perception and safety systems that ship and run on machines in the field, where robustness, production feasibility, and quality matter more than algorithmic sophistication.
• Breadth across machines and OEMs. You'll integrate perception into a wide range of agricultural machines from different OEMs, spanning varied sensor kits, form factors, tooling, and field conditions.
• Infrastructure built for fast iteration. You'll work in a mature development environment with CI/CD, data logging, and testing frameworks already in place, so your effort goes into improving perception algorithms and closing the loop on real field data.

Join us to help build the perception and safety platforms powering the next generation of autonomous agricultural robots.