Max Robotics — Global Service Robot Directory + US Certification

About

Katherine Liu is a roboticist whose research lies at the intersection of autonomous navigation, 3D perception, and manipulation in unstructured environments. Her most impactful work addresses the formidable challenge of multi-robot search and rescue under forest canopies, where GPS is unavailable and severe perceptual aliasing hinders reliable mapping. Her 2020 paper on this topic, with 122 citations, presents a system that enables multiple UAVs to collaboratively explore and map these visually ambiguous environments. Liu has also advanced manipulation by introducing object-centric neural representations for multi-object interactions and developing ZeroGrasp, a framework that performs zero-shot shape reconstruction to enable robust robotic grasping from partial sensory data. To tackle the challenge of efficient planning in unknown spaces, she proposed learned sampling distributions that integrate geometric and object-level information, enabling more intelligent trajectory generation. Her work on the ROAD framework introduces a recursive octree auto-decoder for compact and accurate 3D shape encoding, addressing scalability limitations of existing methods. Through these contributions, Liu is shaping how robots perceive, navigate, and interact with complex, real-world environments.