Max Robotics — Global Service Robot Directory + US Certification

About

Yijia Weng is a leading researcher in robotic manipulation, specializing in dexterous grasping and articulated object interaction. Her most impactful contribution is **UniDexGrasp**, a pioneering framework that enables universal robotic dexterous grasping from point cloud observations. This work, with 94 citations, introduces a novel approach combining diverse proposal generation with goal-conditioned policies, allowing robots to grasp and lift objects in high-quality, diverse ways across hundreds of categories—including unseen objects—under table-top settings. Weng further advances the field with **AO-Grasp**, which generates 6-DoF grasps for articulated objects like cabinets and appliances, enabling robots to perform complex interactions such as opening and closing doors. This work includes both a novel grasp proposal model and a dedicated dataset, addressing a critical gap in robotic manipulation. Weng’s research directly tackles key challenges in generalization and dexterity, pushing the boundaries of what robots can handle in unstructured environments. Her work is essential reading for anyone interested in learning-based robotic grasping, point cloud perception, or autonomous manipulation of everyday objects.