Max Robotics — Global Service Robot Directory + US Certification

About

Cangsong Pang is a researcher advancing autonomous navigation in challenging environments, with a focus on millimeter-wave radar perception and robotic vision. Their work addresses critical limitations of traditional LiDAR-based systems, which are costly and fail in adverse weather or visually degraded conditions. Pang’s most-cited paper (2023, 16 citations) introduces a relocalization method using millimeter-wave radar point clouds for SLAM in visually degraded environments, enabling robust loop closure and drift correction without reliance on expensive LiDAR. They further extend radar’s potential with RadarMOSEVE (2024, 4 citations), a spatial-temporal transformer network that simultaneously segments moving objects and estimates ego-velocity using radar alone—a breakthrough for all-weather autonomy. In environmental robotics, Pang tackles water surface waste detection with a decoupled diffusion model (2023, 6 citations), improving small-object detection for autonomous cleaning robots. Their contributions bridge the gap between cost-effective sensing and reliable performance under real-world constraints, making autonomous systems more accessible and resilient. Pang’s work is pivotal for students and researchers exploring radar-based perception, SLAM, and environmental monitoring in non-ideal conditions.