Max Robotics — Global Service Robot Directory + US Certification

About

Li Yang Ku’s research sits at the intersection of robotics, computer vision, and cognitive science, with a central focus on enabling robots to perceive, manipulate, and learn from their environments. His most significant contribution is the development of the Aspect Transition Graph (ATG), an affordance-based model that represents objects as a series of visual “aspects” linked by the actions a robot can perform. This framework allows a robot to predict the consequences of its manipulations and plan sequences of actions, moving beyond static object recognition to dynamic, interaction-driven understanding. Ku’s work on integrating Convolutional Neural Networks (CNNs) into robotic grasping systems has been instrumental in bridging the gap between high-dimensional visual data and low-level motor control, enabling human-like robot hands to pre-shape for grasps based on image input. His 2015 paper on error detection in stochastic robot actions introduced a general framework for identifying and correcting failures early in a task by storing fine-grained event transitions. With over 30 citations across his top papers, Ku’s research has laid important groundwork for building more autonomous, adaptive robots that can learn from their own interactions and recover from unexpected outcomes.