Meet the spontaneously dancing pint-sized robots

In January 2020, a lab on the second floor at Northwestern University was filled with the moderate clicking of three robots pushing around each other. The trio were in a small circle as they hit each other, although the little bots weren’t the rock ’em, sock’ em type. They were intelligent, active particles – “smarticles” – fitted with two pallet-shaped flaps for the arms, extending less than 6 inches from end to end, and topped with labels to track their position and direction. orientation. The little buggers would go through the unpredictable and unflattering movements of the mess until, every now and then, they gracefully morph into recognizable coordinating movements: a dance.

The smarticles weren’t programmed with any special instructions, or invited to have fun with each other. Bots were prescribed workouts, or movement patterns for their strands, which surprisingly gave way to dance sequences. The models and the physics behind them are described in an article published today in the journal Science. The research was funded by the National Science Foundation, the James S. McDonnell Foundation, and the Army Research Office.

When the smarticles weren’t in sync, there was “chaos of beating and crashing all around the ring which was fascinating to watch, but definitely not orderly,” said Thomas Berstreetta, robotics at Northwestern University and co-author of the article, during a video call. But by teaming up with Pavel Chvykov, a physicist at the Massachusetts Institute of Technology, and Jeremy England, a physicist formerly at MIT and now at Georgia Tech, the research team programmed the smarticles to perform the driving model at the same time.

“Suddenly they were doing this beautiful rotating procession”, BerrEUyou said. “As someone who had smarticles and hadn’t done them before, it was like [Chvykov] came and did a magic trick with my own tools.

Order is in many places in the natural world – the flock of birds, for example, or water crystallizing in ice – but predicting it is a beast in non-equilibrium situations, where external forces are at stake. (And to be clear, the world of non-equilibrium is the big, widest one outside your window – a vast area compared to the feats achievable in a predictable lab environment). In the 1870s, a Swiss physicist by the name of Charles Soret conducted experiments which showed how a saline solution in a tube exposed to heat on one side would cause a larger order of particles on the colder side. Because the molecules move more violently to the hot side of the tube, more of them end up moving to the colder side; cooler molecules, with their delicate movements, don’t end up traveling that far that fast. This means that the particles end up accumulating on the cold side of the tube. The principle, called thermophoresis, was a model for England and Tchvykov seeing the promise of objects in so-called low knocking states.

Clicking is when matter uses the energy flowing through it to move. According to England, the greater the click, the more random or spastic the movement, and the lower the click, the more intentional or incremental the movement. Both could also be true.

“The idea is that if your matter and energy source allow the possibility of a low knocking state, the system will randomly rearrange itself until it finds that state and gets stuck in it,” said England in a Georgia Tech press release. “If you provide energy by forces with a particular pattern, it means that the selected state will discover a way for matter to move that fits that pattern perfectly.”

In this case, the pattern was the prescribed flap movement, and the material moving to match that pattern was the robots slapping each other in rotations and translation around the ring around them. These little flaps were a great testing ground for the idea that low knocking states would lead to stable, self-organizing dances. Unlike other muses, smarticles did not have a molecular source of self-ordering behavior (like how water turns to ice at a certain temperature). The other variables at play in the crystals give way to alternative explanations for the command, obscuring the low-noise idea the research team wanted to test.

Since smarticles only move by contact with each other (they cannot step or roll), there is also less unknown about where objects’ mobility came from, England said. , a problem you would have if all smarticles had little motors propelling them into their dance. When robots can only move by pushing each other, you know the movement you are seeing is the result of collective behavior.

“This article suggests a general principle that complex systems naturally gravitate towards behavior that minimizes ‘clicking’,” said Arvind Murugan, a University of Chicago physicist who is not affiliated with the recent article, in an email. . “The current application to robots shows that the idea survives its first contact with reality. But future work will have to show whether this principle is a good approximation for other complex systems – from molecules to cells to human crowds at a rock concert (after COVID, of course).

Murugan adds that the principle is not always true, “and only approximately true when it is true”. But the idea as interpreted by robots shows that, given this driving force, in a low noise state they will dance.

“As soon as you have a bunch of bots interacting with each other and interacting with people… the idea in this article is that they’re going to be in sync every now and then. And when they sync, there will be some emergent behavior, but you can’t necessarily know what that emergent behavior will be, ”said Todd Murphey, roboticist at Northwestern University and co-author of the article. “If we don’t want to talk about emerging behavior as a fundamental outcome that we should always expect for a sufficiently complex and out of balance system, then we’re going to miss things that can reasonably be expected to happen.”

The implications of robotic movements go beyond refining your DDR technique. Although only three small contraptions spinning, the smarticles display a principle that could be applied to self-driving cars or even humans inside.