Robots use electromagnetic force to create shape-shifting swarm

We have written several stories (see this September 2007 Daily Wire story, for example) about how the news phase in robotics is to get individual robots to coordinate their activities with each other. Swarms of robots can then be used in formation for surveillance or attack. Now, Carnegie Mellon University researchers develop swarms of robots which use electromagnetic forces to cling together and assume different shapes. The goal is to create swarms of microscopic robots capable of morphing into virtually any form by clinging together. Seth Goldstein, who leads the research project at Carnegie Mellon, admits this is still a distant prospect. Hs team, however, is using simulations to develop control strategies for futuristic shape-shifting, or “claytronic,” robots, which they are testing on small groups of more primitive, pocket-sized machines. These prototype robots use electromagnetic forces to maneuver themselves, communicate, and even share power.

One set of claytronic prototypes were cylindrical, wheeled robots with a ring of electromagnets around their edge, which they used to grab hold of one another. By switching these electromagnets on and off, the so-called “claytronic atoms” or “catoms” could securely attach and roll around each other. The robot’s wheels were not powered, so they had to rely entirely on their magnets to maneuver themselves around. “These were the first mobile robots without any moving parts,” says Goldstein. They also used their electromagnets to share power, to communicate, and for simple sensing.

Since using magnetic forces are less efficient at smaller scales, the team has now begun experimenting with electric forces instead. The latest prototypes are box-shaped robots dubbed “cubes” that have six plastic arms with star-shaped appendages at the end of each. These stars have several flat aluminium electrodes and dock together, face on, using static electricity. Electrodes on different stars are given opposing charges, which causes the stars to attract each other.

The New Scientist’s Tom Simonite writes that tests have shown that it is possible to send messages and power to other cubes over the same links. “Our hope is to assemble around 100 cubes to experiment with ideas,” Goldstein says. Rob Reid at the U.S. Air Force Research Lab is collaborating with the Carnegie Mellon team to develop even smaller prototype robots. Reid and colleagues can fold flat silicon shapes into 3D forms as little as a few hundred microns diameter. “We will drive those using electric forces too, by patterning circuits and devices into the silicon design,” Goldstein says. He predicts that by the summer of 2008 they will have prototypes capable of rolling themselves around this way. Modularity is a popular theme with robotics researchers around the world. Other designs include Swarm-bots, Superbot, and M-TRAN.

“The physical mechanism for docking different pieces is really tough to do,” says Alan Winfield, who works on artificially intelligent swarms at the Bristol Robotics Laboratory in the United Kingdom. “Most use mechanical latches with hooks.” Although these physical connections are complex, they do not need power, Winfield points out. “My guess is that electrostatic connectors will come into their own on the micro scale where less power is needed to have a large effect,” he says. Software, however, not hardware, may be the biggest challenge facing researchers working on swarms of robots, he says: “Right now we just don’t know how to design a system that produces a higher overall intelligence from a group of simple agents.” Ultimately, Goldstein believes his claytronic robots may one day achieve this, and much more: “I’ll be done when we produce something that can pass a Turing test face-to-face,” he says. “You won’t know if you’re shaking hands with me or a claytronics copy of me.”