On Continuous-Time State Estimation for Mobile Robotic Systems. Seminario prof. Kailai Li
Data dell'evento:
Wednesday, 29 April, 2026 - 14:30
Luogo:
Aula Magna
Contatto:
grisetti@diag.uniroma1.it
Abstract:
Embodied intelligence lies at the intersection where artificial intelligence meets physical bodies, and is vividly exemplified by recent advances in mobile robotics enabled by high-performance perception sensors and actuators. For dynamic state estimation using multimodal sensors, we explore methodologies based on continuous-time motion representation, particularly using B-splines and Gaussian processes. These methods have been validated in a variety of real-world settings for localization, odometry, and mapping, involving sensors such as IMU, LiDAR, UWB, and camera, and have been deployed on wheeled, legged, wearable, and aerial platforms. Furthermore, we introduce sim-to-real reinforcement learning approaches toward achieving motor intelligence in legged locomotion. Building on this foundation, we demonstrate possible ways to incorporate perceptive inputs, enabling robots to navigate through dynamic and unstructured environments. The talk aims to provide balanced insights from both methodological and engineering perspectives, with an emphasis on real-world, real-time challenges.
Embodied intelligence lies at the intersection where artificial intelligence meets physical bodies, and is vividly exemplified by recent advances in mobile robotics enabled by high-performance perception sensors and actuators. For dynamic state estimation using multimodal sensors, we explore methodologies based on continuous-time motion representation, particularly using B-splines and Gaussian processes. These methods have been validated in a variety of real-world settings for localization, odometry, and mapping, involving sensors such as IMU, LiDAR, UWB, and camera, and have been deployed on wheeled, legged, wearable, and aerial platforms. Furthermore, we introduce sim-to-real reinforcement learning approaches toward achieving motor intelligence in legged locomotion. Building on this foundation, we demonstrate possible ways to incorporate perceptive inputs, enabling robots to navigate through dynamic and unstructured environments. The talk aims to provide balanced insights from both methodological and engineering perspectives, with an emphasis on real-world, real-time challenges.
Short bio:
Kailai Li is an Assistant Professor with tenure at the University of Groningen, The Netherlands. He earned his B.E. in Mechanical Engineering and Automation from Tsinghua University, his M.Sc. in Automation Engineering from RWTH Aachen University, including a master’s thesis at ETH Zurich, and his Ph.D. from the Karlsruhe Institute of Technology (KIT), Germany. He now leads the Agile Sensing and Intelligence Group (ASIG) with focuses on building future-proof embodied AI systems for the real world. The group’s research spans sensing to actuation with open-source contributions (GitHub: https://github.com/ASIG-X).
gruppo di ricerca:
keywords: