Staff Engineer, Robotics Perception

ARX Robotics

Munich, Germany

As a Staff Engineer – Robotics Perception, you will develop the systems that enable robots to interpret complex outdoor environments and make safe, reliable decisions in real time.

You will work closely with autonomy, navigation, hardware, and software teams to integrate perception capabilities into a unified robotics platform. Your work will directly influence how autonomous and teleoperated systems operate in challenging, unstructured environments under real-world constraints.

What You’ll Build

Design and implement perception algorithms for object detection, terrain classification, obstacle detection, and scene understanding in off-road environments
Fuse and process sensor data from LiDAR, cameras, radar, IMU, and GNSS for robust environmental interpretation and free-space understanding
Develop ML and deep learning-based models to improve perception performance in dynamic outdoor conditions
Optimize real-time perception pipelines for embedded systems under strict size, weight, and power constraints
Test and validate perception systems in simulation and real-world environments, including harsh conditions such as mud, dust, and rain
Integrate perception outputs into navigation and path planning systems for safe autonomous traversal
Support system-level calibration, synchronization, and sensor alignment in collaboration with hardware and systems teams
Continuously improve robustness and reliability through field testing and data-driven iteration
Work across teams to ensure perception systems are deployable, scalable, and production-ready

What You Bring

Advanced degree (MSc or higher) in Robotics, Computer Vision, Machine Learning, or a related engineering field
Strong proficiency in Python and C++ for robotics and ML system development
Solid understanding of perception fundamentals including object detection, segmentation, terrain understanding, and obstacle detection in unstructured environments
Experience developing and deploying machine learning-based perception systems in real-world robotics applications
Hands-on experience with frameworks such as PyTorch and TensorRT
Proficiency with OpenCV, PCL, or Open3D for perception and 3D data processing
Experience working with ROS2, simulation environments, and robotics validation workflows
Strong Linux development skills with experience in CI/CD and modern software engineering practices
A structured, self-driven, and proactive approach to solving complex perception challenges
Experience with Rust or CUDA and edge computing platforms are a plus