Max Robotics — Global Service Robot Directory + US Certification

About

Kai-yuh Hsiao’s research lies at the intersection of robotic manipulation, tactile perception, and human-robot interaction, with a particular focus on enabling robots to grasp unknown objects and engage in fluid, conversational behavior. His most influential work, “Contact-reactive grasping of objects with partial shape information” (2010, 201 citations), introduced a robust approach to grasp selection for unstructured environments, allowing robots to reactively adjust grasps despite sensor noise or calibration errors—a foundational contribution to autonomous manipulation. Hsiao also pioneered the integration of mental imagery and language in robotics; his 2004 paper “Mental Imagery for a Conversational Robot” (117 citations) demonstrated how robots could connect spoken language to visual perspectives, addressing the critical challenge of grounding deictic terms like “my left” versus “your left.” His decision-theoretic framework for task-driven tactile exploration (2010, 69 citations) further advanced precision placement and grasping by using tactile feedback to localize objects. Across his work, Hsiao developed object schemas for responsive language use (2008, 35 citations) and a perceptually-coupled simulation system that gave robots an internal world model (2004, 31 citations). His contributions have shaped how robots perceive, grasp, and converse in dynamic, human-centered environments.