Tikalon Blog by Dev Gualtieri

Grey/Gray Goo

April 29, 2019

English is a confusing language. At the time of the American Colonial Period, spelling was often phonetic, and you might even find the same word written in different ways in the same paragraph. Some examples are chuse (choose), musick (music), goal (jail), and chirurgeon (surgeon).[1] William Shakespeare (1564-1616) had his name spelled in a multitude of ways that include Shaksper, Shakspe, Shakspere, Shakspere, and the usual Shakespeare. English spelling became somewhat more uniform after Samuel Johnson (1709-1784) published his 1755 Dictionary of the English Language.[2]

Samuel Johnson (1709-1784), looking the part of the school headmaster you would not wish to cross.

This engraving is from the frontpiece of his Dictionary of the English Language.

As if to confirm the stereotype of the starving academic, Doctor Johnson, as he was called, earned the equivalent of today's $300,000 for his seven year effort.

(Via The Internet Archive.[2])

One thing that confused me when I was a student was the two ways of spelling the color intermediate between white and black; namely, gray, or grey. Both spellings are "correct," but grey is the common usage in the United Kingdom, and gray is the common United States spelling. You can confirm this on Wikipedia, where your search for "gray" redirects you to the page for "grey." The English language Wikipedia is a great resource, but it has an Anglophile bias. I discovered this when searching for "jewelry" many years ago and ended up on a page for "jewellery." It's interesting to note the spelling of "Fifty Shades of Grey" in the novel and film. Google reports twice the number of hits for gray than grey.

Gray is the lukewarm of the color world, and its name has been attached to a supposed nanotechnology threat called "gray goo." The gray goo scenario has runaway self-replicating machines transform the Earth into myriad copies of themselves, an exponential process that creates a "goo" of these machines and waste matter. As can be imagined, gray goo has been the subject of many books and films. One excellent example is the 2008 film, The Day the Earth Stood Still, although I prefer the original 1951 film with Sam Jaffe (1891-1984) as the Albert Einstein stand-in.

Rapid prototyping machines (3D printers) can produced some of the parts to replicate themselves, and nanotechnology has advanced to the point that a smartphone contains many microelectromechanical (MEMS) components that provide timekeeping and the accelerometer that causes your screen image to flip from landscape to portrait mode as your smartphone is rotated. However, we're still quite far from producing a gray goo machine that can analyze and mine the materials in its environment and effectively recreate components that are now created in a $10 billion semiconductor fabrication plant.

Grey goo must have a prime directive to not use others of itself as feedstock; and, perhaps, another directive that allows teams of devices to cooperate in processes to their mutual benefit. computer scientists and robotics experts have harnessed the power of device cooperation to create swarms of useful objects. The venerable IEEE, of which I am a member, has published many articles about swarm robots in IEEE Spectrum. Here's a sampler of articles, all written by Evan Ackerman.

•  Navy Wants Robot Swarm That Can Autonomously Build Stuff, Apocalypse Unlikely, March 11, 2011.

•  Watch SRI's Nimble Microrobots Cooperate to Build Structures, April 16, 2014

•  A Thousand Kilobots Self-Assemble Into Complex Shapes, August 14, 2014

•  NASA Training 'Swarmie' Robots for Space Mining, August 20, 2014

•  World's Largest Swarm of Miniature Robot Submarines, May 5, 2015

•  Swarms of Robots Manage to Not Run Into Each Other, September 8, 2016

Casting call for the next trivial CGI science fiction film.

This is the nymph form of the desert locust (Schistocerca gregaria).

Most people know swarms only through the Biblical plagues of Egypt, one of which was a swarm of locusts. Locusts are actually grasshoppers that transition from a solitary to a gregarious state susceptible to swarming.

(Wikimedia Commons image by Danny Steaven (with background removed). Click for larger image.)

While quite a bit larger than the nanotechnology gray goo of science fiction, a robotic gray goo, sans self-replication, has been developed by a team of computer scientists and engineers from the Massachusetts Institute of Technology (Cambridge, Massachusetts), Columbia University (New York, New York), Cornell University (Ithaca, New York), and Harvard University (Cambridge, Massachusetts).[3-5] They've created a loosely-coupled assemblage of particle robots that act cooperatively. Each robot member is capable of just the simple motions of expansion and contraction. In combination, these robots can perform some simple tasks, and they exhibit some interesting behaviors. Their locomotion is derived from statistical mechanical principles.

The simple expanding/contracting robotic particle on the left can act in concert with other such particles to move within a plane in response to a stimulus. (Left image from a Columbia University YouTube Video by Richa Batra, Shuguang Li, Jane Nisselson, Kyle Parsons/Columbia Engineering. Right image by Shuguang Li/Columbia Engineering,)

Conventional robotic swarms operate with a programmed coordinated motion in which each member is an independently functioning machine that's given explicit movement commands. As a consequence, the failure of a very small number of the swarm members renders the swarm inoperable.[3-4] In contrast, biological organisms are more robust since their high-level behavior is achieved through coordinated action of their components operating stochasticly.[3] The research team showed that this same stochastic, statistical mechanical operating principle can be programmed into particle robots that respond to stimuli without specific programming.[3]

The individual robotic particles perform only uniform volumetric oscillations of expansion and contraction that are phase-modulated by a stimulus signal such as a light source.[3-4] In response to a light source, each robot measures the light intensity, broadcasts its value to the group, and then receives the light intensity sensed by its neighbors to determine the phase delay of its motion. The delays result in a motion towards the light source.[4]

The robot particles closer to the light are designed to start their volume pulsing cycle earlier, and this creates a motion wave throughout the cluster that drives the cluster towards the light. Although the individual robot particles cannot move independently, but only shrink or swell, the light stimulus creates a global motion.[4] This was verified experimentally with a group of up to two dozen robots, and more complex motions were investigated in computer simulations of up to 100,000 robots.[3] Transport of objects was also realized.[3]

A computer simulation of barrier penetration by an assemblage of many robotic particles. The mass of particles streams through a small opening in a barrier. (Still images from a Columbia University YouTube Video by Richa Batra, Shuguang Li, Jane Nisselson, Kyle Parsons/Columbia Engineering.)

Says principal investigator, Hod Lipson, a professor of mechanical engineering at Columbia University,

"You can think of our new robot as the proverbial 'Gray Goo'... "Our robot has no single point of failure and no centralized control. It's still fairly primitive, but now we know that this fundamental robot paradigm is actually possible. We think it may even explain how groups of cells can move together, even though individual cells cannot... We've been trying to fundamentally rethink our approach to robotics, to discover if there is a way to make robots differently... Not just make a robot look like a biological creature but actually construct it like a biological system, to create something that is vast in complexity and abilities yet composed of fundamentally simple parts."[4]

Such a system is robust against the failure of individual components. This was demonstrated also in simulations of obstacle avoidance and object transport with hundreds and thousands of particles.[4] The particle robot collective maintained roughly half its speed when as many as 20% of its members were dormant.[4] Says Lipson, "We think it will be possible one day to make these kinds of robots from millions of tiny particles, like microbeads that respond to sound or light or chemical gradient... Such robots could be used to do things like clean up areas or explore unknown terrains/structures."[4] This research was supported by the Defense Advanced Research Projects Agency and the National Science Foundation.

Grey/Gray Goo

References: