Abstract

Abstract: Dexterous manipulation remains one of the grand challenges in robotics, requiring the seamless integration of sensing, mechanics, and control. Recent advances in optical tactile sensing have opened new possibilities by allowing robots to see through touch—using embedded cameras to capture detailed contact geometry, deformation, and force distribution as tactile images. This talk presents our progress in optical tactile sensing and its role in transforming the way robots perceive and interact with the physical world. I will begin by introducing our developments in optical tactile sensors, from GelTip and TouchRoller to ViTacTip and RoTipBot, which innovate omnidirectional touch sensing to achieve high dexterity. Beyond sensing, I will highlight our work on tactile simulation and modelling that supports high-speed and physically grounded tactile perception. Finally, I will demonstrate how integrating vision and touch to enhance perception for dexterous manipulation. Together, these advances point toward a new generation of robots capable of perceiving, reasoning, and acting through both touch and vision. Bio: Dr Shan Luo is a Reader in Robotics and AI at King’s College London, where he leads the Robot Perception Lab and the Centre for Robotics Research at the Department of Engineering. He is an EPSRC Open Fellow and a Turing Open Source AI Fellow. Shan received his PhD in Robotics from King’s College London in 2016 and has held research positions at MIT CSAIL , Harvard University, and the University of Leeds. From 2018 to 2021, he was Lecturer and Director of the smARTLab at the University of Liverpool before returning to King’s. His research aims to endow robots with human-like perception and dexterity, focusing on tactile sensing, multimodal perception, and robot manipulation. His work has appeared in leading robotics journals and conferences including IEEE Transactions on Robotics, IEEE Transactions on Mechatronics, IEEE Transactions on Pattern Analysis and Machine Intelligence, ICRA , IROS, RSS , and NeurIPS. He serves as an Associate Editor for IEEE Transactions on Robotics, and ICRA /IROS, and recently led a Special Issue on Tactile Robotics that received record-breaking submissions. His research has been funded by EPSRC , Innovate UK, AHRC , the Royal Society, and industrial partners including Unilever, with Shan serving as PI on projects exceeding £4M.