A six-spoke soft robot developed at Zhejiang University can fold into a smaller package and switch among more than nine ways of moving. The researchers reduced its occupied volume by 79 percent, then demonstrated rolling, crawling, jumping and swimming across obstacles, gels, thick fluids and three-dimensional printed stomach surfaces. The work addresses a practical problem for small robots: one gait rarely works across both narrow openings and uneven, wet surfaces.
The robot is made from an elastomer, a rubber-like material, with channels of liquid metal running through each spoke. When electric current passes through those channels inside a static magnetic field, electromagnetic forces bend or propel different parts of the body. Changing which channels receive current and how they are driven lets the same structure stand, lie flat, roll or use several walking patterns instead of carrying a separate motor for each motion.
The motions depend on a magnetic field and electric current through the robot's channels. Switching between modes took less than 0.35 seconds in the reported tests. In its fastest rolling mode, the robot reached 818 millimetres per second, or 26 body lengths per second. Folding draws the spokes inward so that the robot can pass through a tighter opening, then reopen and select another gait for the terrain beyond it.
The team tested navigation over discrete obstacles, viscoelastic gelatin, viscous liquids and a three-dimensional printed stomach surface. These trials show that one body can keep moving as the mechanical conditions change.
The paper also reports a demonstration of drug transport and release. That illustrates a possible delivery task for the robot, but does not establish safety or a treatment benefit in people.
Sources
- Nature Communications: A foldable small-scale soft electromagnetic robot for multimodal navigation in confined and unstructured environments
- arXiv preprint: A Foldable and Agile Soft Electromagnetic Robot for Multimodal Navigation in Confined and Unstructured Environments
- Zenodo research data record
- Figure 1 by Lv et al., CC BY 4.0; the biomedical route in the figure is a proposed application