“Brainless“ autonomous soft robotic maze escaper

Researchers at NC State University reported an intelligent soft robot that can self-escape from different challenging mazes without need of human-like brain on the open-access journal of Science Advances () on Sept. 8, 2023. These challenging mazes include multi-channeled mazes, maze on sand, mazes with narrower exits than the soft robot body size, and dynamic mazes with in-situ changing layouts. It harvests thermal energy from environment for self-powered rolling, self-turning, and self-obstacle avoidance. The soft robot is made of liquid crystal elastomers (LCEs), a smart material that can shrink over 50% in response to heat or light, and takes an asymmetric conical twisted shape. The asymmetric shape enables the soft robot to actively self-turn and self-reflect its motion whenever encountering a wall for maze-escaping. This work could have potential robotic applications in self-powered autonomous and intelligent open-task navigation and environmental monitoring in unstructured challenging environments, including desert and high-temperature environments etc. This work is supported by NSF under the awards of CMMI-2005374, CMMI-2126072, CMMI-1944655 and 2026622.