Scientists have developed a prototype for self-replicating robots

Scientists have developed a prototype for self-replicating robots

The image shows a prototype with its innards exposed and its flywheel pulled out.

Robotics researchers have constructed small cubes that with no exterior moving parts are able to propel themselves forward, climb and jump on top of each other and snap together to form arbitrary shapes. The cubes could even move while suspended upside down from metallic surfaces.

The idea first came about in 2011 when a research student at MIT called John Romanishin proposed a new design for modular robots to his robotics professor, Daniela Rus. At first Rus was unsure of the design; however two years later Rus showed a fellow robotics researcher at Cornell University, a video of prototype robots, based on Romanishin’s design, in action. He was also skeptical at first.

This coming November Romanishin, who is now a research scientist in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), as well as Rus and postdoc Kyle Gilpin are going to present a paper describing their new robots at the IEEE/RSJ International Conference on Intelligent Robots and Systems.

The robots are known as M-blocks and are cubes that have no external moving parts. Inside each cube is a spinning mass, a flywheel that can reach speeds of 20,000 revolutions per minute. Angular momentum is transmitted to the cube when the flywheel is braked. Permanent magnets are found along each edge of the M-blocks and along every face, allowing the cubes to attach to each other.

Robotic researchers had previously used an abstraction called the sliding-cube model, where one cube can slide up another and across its top if the two cubes face each other. The robots that implement these sliding-cube models are much more complex than the M-blocks. Existing modular-robot systems are also “statically stable,” which means that the motion can be paused at any time, and the robots stay where they are. The MIT researchers were able to simplify their robots’ design by forgoing the principle of static stability. The magnets on the cube essentially bring the cubes into alignment.

Each edge of a cube has two cylindrical magnets that are mounted like rolling pins. When two cubes approach each other, the magnets rotate automatically so that north poles align with south and vice versa. This means that any cube can attach to the face of any other cube. The edges of the cubes are bevelled, which means there’s a slight gap between the magnets when the two cubes are attached. This strengthens the connection between the cubes and anchors the pivot. Four more pairs of smaller magnets are on each face of the cube.

The team behind the robots aim to make them as miniatures, in effect to have a swarm of microbots that can self-assemble. Even at their current size and with further refinements, the robots could prove useful. The mobile cubes could be used to form different types of equipment, for repairs, or to gain access to environments inaccessible to humans. The cubes could also be used for transportation of specialty equipment.

The researchers are planning to build an army of 100 cubes, and to further refine their robots.