The neuromuscular system is crucial for animals to achieve locomotion, manipulation, adaptation, and ultimately survival, which inspires the development of next-generation agile and adaptable robots. In this seminar, I will explore the design and development of artificial neuromuscular systems that mimic and offer insight into their biological counterparts. First, I will discuss how powerful, responsive, robust, and adaptable artificial muscles are created by braiding 3D-printed liquid crystal elastomer fibers, along with techniques to integrate proprioceptive sensing capabilities. Next, I will demonstrate the application of these artificial muscles in scalable underwater robots capable of imitating the pulsatile jet swimming of nautilus via buckling of a soft origami chamber. Finally, I will show how we achieve feedback control of an artificial muscle-powered robotic arm using end-to-end neural networks trained through reinforcement learning with sim-to-real transfer. Remarkably, the artificial neural networks exhibit neural activities closely resembling animal motor cortex, thus allowing us to decode the complex neural population dynamics underlying animal movement.
Bio: Dr. Ke Liu is currently an Assistant Professor of the Department of Advanced Manufacturing and Robotics at Peking University. He got his PhD degree from Georgia Tech in 2019. He then worked at Caltech as a postdoc for 3 years before joining Peking University. His research covers soft robots, metamaterials, reconfigurable structures, and AI for design, with focus on the interplay between geometry and mechanical properties. He has published on prestigious journals including Science, Science Robotics, Nature Communications, Advanced Materials, and PRL. He is listed in the Stanford/Elsevier Top 2% Scientists of 2024 yearly impact. He is awarded the Sigma Xi Best Ph.D. Thesis Award by Georgia Tech, the Melville Medal by ASME, the First Prize in Young Faculty Teaching Competition by Peking University, and the 2023 Xiong Youlun Young Talent Award of China.