We are delighted that Asst Prof Fan Shi and his research team RF, Ziqiu Zeng, and RE, Siyuan Luo, have received a USD16,000 unrestricted research gift funding from Google to support their exploration in AI for robotics. This application was recommended through NUS AI Institute, and we are deeply grateful for both their facilitation and Google’s generous support.
Their proposal name is “Sim-to-Dex: Contact-Rich Graph Neural Models for Robotic Manipulation”. This proposal explores how AI can enhance robotics by developing graph neural network models that efficiently capture the complex dynamics of contact-rich dexterous manipulation on humanoid robots. By integrating physics-based simulation data with AI-driven learning, the project aims to achieve accurate yet real-time decision-making, bridging the gap between simulation and robotic execution.
