Researchers at the École Polytechnique Fédérale de Lausanne (EPFL) have developed a bioinspired robot capable of changing shape to adapt to different terrains, enhancing both its mobility and energy efficiency. The Good Over All Terrains (GOAT) robot, created by the CREATE Lab in EPFL’s School of Engineering, transitions between a flat “rover” configuration and a sphere, allowing it to switch between driving, rolling, and swimming while conserving energy. The research, led by Josie Hughes, has been published in Science Robotics.
Unlike conventional autonomous robots that follow pre-programmed paths and require significant computational resources to navigate obstacles, GOAT dynamically adjusts its shape to optimize movement. It can roll downhill in spherical mode to save energy, traverse water by swimming, and return to a wheeled rover configuration when necessary. The robot is built from inexpensive materials, featuring a frame made of two elastic fiberglass rods with four motorized rimless wheels. Its transformation is controlled by two winch-driven cables that pull the structure into a ball. A 2 kg payload, containing the battery, onboard computer, and sensors, is suspended at the center of the frame, ensuring protection in sphere mode.
GOAT’s adaptability reduces its reliance on complex sensing systems. It operates with minimal equipment—only a satellite navigation system and an inertial measurement unit—eliminating the need for cameras or detailed environmental mapping. According to CREATE Lab PhD student Max Polzin, the robot’s compliance allows it to navigate extreme terrain with limited perception, relying instead on its ability to adjust to environmental forces.
Future developments for GOAT include refining algorithms to further exploit its morphing capabilities and scaling the design for different payload requirements. Potential applications range from environmental monitoring and disaster response to extraterrestrial exploration, where robots with minimal perception and planning systems could rapidly adapt to unknown landscapes.
Photo Credit: CREATE EPFL
