A novel sim-to-real reinforcement learning algorithm for soft growing robot navigation
Wu H R, Sun F C, Kang Z X, et al
Sci China Inf Sci, 2025, 68(11): 214201
Navigating confined and restricted environments poses significant challenges for conventional rigid robots. In contrast, soft growing robots, which draw inspiration from plant growth mechanisms, offer a biologically inspired solution by elongating through material eversion. This unique growthbased locomotion significantly reduces friction during movement. Furthermore, the robot’s inherent compliance enables it to navigate through environmental interaction, thereby enhancing its capability to confined spaces.