Drawing inspiration from feather stars, a research team headed by Professor Yin Jie at North Carolina State University has engineered an ultra-lightweight, monostable, dual-drive, 3D-motion soft underwater robot. Weighing a mere 14.5 grams, this robot is propelled by just two air tubes and can seamlessly transition among three swimming modes—jellyfish, fish, and rotor. This capability allows it to execute complex 3D maneuvers, including ascending, descending, hovering, and turning, all without relying on microcomputers, sensors, or real-time control algorithms. Its performance outstrips that of comparable bionic robots in numerous ways, making it an ideal candidate for tasks such as underwater object retrieval, trash collection, and pipeline inspection.
The team has termed the underlying design philosophy “mechanical intelligence.” This approach achieves control functions through the interplay of material properties, geometric configurations, and structural designs, thereby maximizing functionality with minimal actuation. Rather than attempting to directly mimic the intricate physiological structure of feather stars, the team focused on extracting adaptable mechanical principles. By addressing engineering challenges and omitting non-essential details, they created a streamlined, viable engineering unit.
This technology marks the team’s third iteration along the path of mechanical intelligence. Currently, such open-loop control robots still encounter limitations in adapting to diverse scenarios. The team envisions that future integration of mechanical intelligence with electronic control systems could delegate low-level bodily coordination to the former and high-level decision-making to the latter. Moving forward, the team will investigate drive solutions tailored to real ocean environments and refine layered deployment strategies. Their ultimate goal is to establish a scalable design methodology for long-duration, tether-free multi-robot ocean exploration and monitoring. Relevant research findings have been published in Science Advances.
