Ultra Mobility Vehicle (UMV) is focused on advancing robot mobility to make machines more useful in everyday life. The team combines wheels, dynamic legged systems, and whole-body behavior to reach higher levels of movement capability. Using fast terrain perception, our robots adapt to their environment by rolling over or jumping on obstacles. This research is central to developing robotic systems that move with speed, agility, and versatility.
Requirements
- Strong programming skills with Python and C++.
- Experience with ROS 2 and robotic software integration.
- Good grasp of at least a handful of computer vision techniques, for example 3D localization, robot state estimation, depth sensing, object modeling, scene graphs, object pose estimation, Gaussian splatting, or point/object tracking.
- Direct, hands-on experience implementing and debugging factor graph state estimation systems (e.g., using libraries like GTSAM, Ceres, or custom solvers).
- Proven expert-level knowledge of ROS 2 middleware
- Deep expertise with IMUs, including calibration and integration.
- Deep expertise in software optimization, hardware acceleration or parallel computing (CPU, GPU, CUDA).
Responsibilities
- Integrate and extend current state-of-the-art vision algorithms and hardware to work on highly dynamic mobile robots, with high-rate and precision requirements.
- Work with Controls, Hardware, and other researchers on designing, training, developing, and deploying new vision algorithms for dynamic locomotion, navigation, state estimation, etc.
- Collaborate with other teams by designing modular software systems and assisting with software infrastructure like calibration efforts, camera drivers, and containerization.
- Promote and enforce quality and reliability of code through the application of best coding and design practices, mentorship, and regular code reviews.
- Share your innovative work with the community through publications at top-tier conferences (e.g., CORL, ICRA, IROS, RSS, CVPR, NeurIPS, etc.)
- Mentor junior engineers and researchers.
Other
- BS, MS or PhD in computer science, robotics, electrical/mechanical/aerospace engineering, or a related field.
- 7+ years in a related industry
- Experience deploying algorithms on physical systems and solving the system level problems that arise.
- Proven track record of collaboration and mentorship of junior colleagues.
- A proven expert-level knowledge of deep learning in the context of robotics problems like semantic segmentation, 3D reconstruction, point cloud registration, or object detection.
