WATCH LIVE:

The Lookout

MIT micro-bots could make a lot of macro-jobs much easier

Ron Recinto
The Lookout

View photo

.

Researchers at Massachusetts Institute of Technology are developing the world's smallest chain robot, less than the size of a dime, designed to link up to others like them and shape-shift into a range of micro-tools.

Put together, four of these machines, with a specialized engine and covered with rings and fittings, look like a tiny, brass mechanical inchworm, but with more versatility and usefulness. "It's a step toward the goal of programmable matter," said Neil Gershenfeld, head of MIT's Center for Bits and Atoms where the micro-bots were created. "The goal is not to just to produce a shape. This is something that can change shape."

Programmable matter is something that can change form based on external commands. Because of this micro-robot's size, a long string of them could be, in theory, programmed to turn into an infinite number of forms. For example, a chain of them could form a wrench. When that tool is no longer needed, the string of robots could be reprogrammed into a coffee cup.

The micro-bot's basic working principle is that each device has a motor that can be programmed to pivot into any angle. Interconnected and individually programmed, a group of a thousand could form any 3-dimensional shape.

"I'm not talking just about just four of these, I'm thinking of a mile-long string of the devices," Gershenfeld said.

Conceptually, a long string of these mini-robots could be programmed to become the building blocks of something larger and more specifically geared to a task. "Imagine a robotic arm and all the electronics and wiring and components you need to manufacture that," Gershenfeld said. "A string of (our) devices would just need to be programmed to essentially into a robotic arm and do the same function."

These micro-bots, if asked, could "become furniture that can walk," Gershenfeld said.

To develop its movable prototype, the MIT team conquered three scientific challenges.

First, they had to develop programs that could turn codes into shapes, meaning transforming binary 1s and 0s into geometric folds.

Then they had to develop a small, efficient motor. To do this, the team opted for a gearless structure using magnets, Gershenfeld said. Changing the polarity on two sets of magnets arranged in a circle drives a steel ring around them. One key innovation is that a just a small amount of power is used when pivoting the ring to a certain angle. Once the unit gets to its position, it will stay in that position even without power, thus making it specifically efficient.

Finally, the team had to design a one-dimensional robot that could be made in a continuous strip and folded into arbitrary three-dimensional shapes,Gershenfeld said. This idea came from nature, as the team mimicked the action of simple proteins that can twist and turn and fold into complex three-dimensional structure that determines its activity.

The MIT device is dubbed "milli-motein" because of its small size, and because it is based on proteins in nature. It's also designed to be inexpensive to mass produce. Spin-offs could be used in the medical device and in the aerospace and airline industries, because of the adaptability of the magnetic motor. Robots in colonies could be used in electrical systems of airplanes, for example, to keep instruments and other important parts of the aircraft in proper positions should power go out.

Kenn Oldham, an assistant professor of mechanical engineering who works on microrobotics at the University of Michigan said the MIT team's work "looks interesting."

"One of the major challenges in miniature robotics is power handling, as power and energy availability is very limited." Oldham said.

The chain robot research is similar to "smart sand" technology, where tiny small robotic blocks can be smartly configured into 3-D objects using complex algorithms. There are also robotic folding origami systems where small flat materials are folded into different shapes, then combined with others to create a robotic network.

But a string of units that could fold itself into any shape would be simpler to control, Gershenfeld's team found, as opposed to something with separate pieces that would have to find each other and assemble in the right order.

View Comments (148)