AI News, How to Control Hundreds of Dumb Robots with One Clever Remote
How to Control Hundreds of Dumb Robots with One Clever Remote
Computers have no trouble controlling huge swarms of robots, because the computer can just treat the swarm as a bunch of individually controllable units.
It seems like this would severely limit what can be done with the swarm, but thanks to some sophisticated algorithms and real world randomness, researchers from Rice University have shown that you can get a swarm of robots like this to do absolutely anything you want.
While it's certainly possible to control many robots with a single remote, you can't easily control them relative to each other, making it nearly impossible to (say) surround a target.
Inspired by this comic,Aaron Becker set out to show that it is, in fact, possible to do what Jason Fox has done: use one remote to steer a bunch of similar (but not identical) robots into arbitrary positions and orientations, with just forward, backward, and rotate commands that affect all the robots equally.
Swarm behaviour, or swarming, is a collective behaviour exhibited by entities, particularly animals, of similar size which aggregate together, perhaps milling about the same spot or perhaps moving en masse or migrating in some direction.
The simplest mathematical models of animal swarms generally represent individual animals as following three rules: The boids computer program, created by Craig Reynolds in 1986, simulates swarm behaviour following the above rules. Many subsequent and current models use variations on these rules, often implementing them by means of concentric 'zones' around each animal.
However recent studies of starling flocks have shown that each bird modifies its position, relative to the six or seven animals directly surrounding it, no matter how close or how far away those animals are. Interactions between flocking starlings are thus based on a topological rule rather than a metric rule.
Another recent study, based on an analysis of high speed camera footage of flocks above Rome and assuming minimal behavioural rules, has convincingly simulated a number of aspects of flock behaviour. In order to gain insight into why animals evolve swarming behaviour, scientists have turned to evolutionary models that simulate populations of evolving animals.
These studies have investigated a number of hypotheses explaining why animals evolve swarming behaviour, such as the selfish herd theory the predator confusion effect, the dilution effect, and the many eyes theory. The concept of emergence—that the properties and functions found at a hierarchical level are not present and are irrelevant at the lower levels–is often a basic principle behind self-organizing systems. An example of self-organization in biology leading to emergence in the natural world occurs in ant colonies.
The agents follow very simple rules, and although there is no centralized control structure dictating how individual agents should behave, local, and to a certain degree random, interactions between such agents lead to the emergence of intelligent global behaviour, unknown to the individual agents.
There is also a scientific stream attempting to model the swarm systems themselves and understand their underlying mechanisms, and an engineering stream focused on applying the insights developed by the scientific stream to solve practical problems in other areas. Swarm algorithms follow a Lagrangian approach or an Eulerian approach. The Eulerian approach views the swarm as a field, working with the density of the swarm and deriving mean field properties.
Individual particle models can follow information on heading and spacing that is lost in the Eulerian approach. Ant colony optimization is a widely used algorithm which was inspired by the behaviours of ants, and has been effective solving discrete optimization problems related to swarming. The algorithm was initially proposed by Marco Dorigo in 1992, and has since been diversified to solve a wider class of numerical problems.
Species that have multiple queens may have a queen leaving the nest along with some workers to found a colony at a new site, a process akin to swarming in honeybees. Self-propelled particles (SPP) is a concept introduced in 1995 by Vicsek et al. as a special case of the boids model introduced in 1986 by Reynolds. A swarm is modelled in SPP by a collection of particles that move with a constant speed but respond to a random perturbation by adopting at each time increment the average direction of motion of the other particles in their local neighbourhood. Simulations demonstrate that a suitable 'nearest neighbour rule' eventually results in all the particles swarming together, or moving in the same direction.
This emerges, even though there is no centralized coordination, and even though the neighbours for each particle constantly change over time (see the interactive simulation in the box on the right). SPP models predict that swarming animals share certain properties at the group level, regardless of the type of animals in the swarm. Swarming systems give rise to emergent behaviours which occur at many different scales, some of which are turning out to be both universal and robust.
It has few parameters to adjust, and a version that works well for a specific applications can also work well with minor modifications across a range of related applications. A book by Kennedy and Eberhart describes some philosophical aspects of particle swarm optimization applications and swarm intelligence. An extensive survey of applications is made by Poli. Researchers in Switzerland have developed an algorithm based on Hamilton's rule of kin selection.
The algorithm shows how altruism in a swarm of entities can, over time, evolve and result in more effective swarm behaviour. Examples of biological swarming are found in bird flocks, fish schools, insect swarms, bacteria swarms, molds, molecular motors, quadruped herds and people. The behaviour of insects that live in colonies, such as ants, bees, wasps and termites, has always been a source of fascination for children, naturalists and artists.
The organised behaviour that emerges in this way is sometimes called swarm intelligence. Individual ants do not exhibit complex behaviours, yet a colony of ants collectively achieves complex tasks such as constructing nests, taking care of their young, building bridges and foraging for food.
A swarm may fly for a kilometre or more to the scouted out location, though some species may establish new colonies within as little as 500 meters from the natal nest, such as Apis dorsata. This collective decision making process is remarkably successful in identifying the most suitable new nest site and keeping the swarm intact.
A good nest site has to be large enough to accommodate the swarm (about 15 litres in volume), has to be well protected from the elements, receive a certain amount of warmth from the sun, be some height above the ground, have a small entrance and resist the infestation of ants - hence why trees are often selected. Similar to ants, cockroaches leave chemical trails in their faeces as well as emitting airborne pheromones for swarming and mating.
That means one million locusts can eat about one ton of food each day, and the largest swarms can consume over 100,000 tonnes each day. Swarming in locusts has been found to be associated with increased levels of serotonin which causes the locust to change colour, eat much more, become mutually attracted, and breed much more easily.
The transformation of the locust to the swarming variety can be induced by several contacts per minute over a four-hour period. Notably, an innate predisposition to aggregate has been found in hatchlings of the desert locust, Schistocerca gregaria, independent of their parental phase. An individual locust's response to a loss of alignment in the group appears to increase the randomness of its motion, until an aligned state is again achieved.
the flight patterns appear to be inherited, based on a combination of the position of the sun in the sky and a time-compensated Sun compass that depends upon a circadian clock that is based in their antennae. Approximately 1800 of the world's 10,000 bird species are long-distance migrants. The primary motivation for migration appears to be food;
Studies show that birds in a V formation place themselves roughly at the optimum distance predicted by simple aerodynamic theory. Geese in a V-formation may conserve 12–20% of the energy they would need to fly alone. Red knots and dunlins were found in radar studies to fly 5 km per hour faster in flocks than when they were flying alone. The birds flying at the tips and at the front are rotated in a timely cyclical fashion to spread flight fatigue equally among the flock members.
Fish derive many benefits from shoaling behaviour including defence against predators (through better predator detection and by diluting the chance of capture), enhanced foraging success, and higher success in finding a mate. It is also likely that fish benefit from shoal membership through increased hydrodynamic efficiency. Fish use many traits to choose shoalmates.
In the case of foraging behaviour, captive shoals of golden shiner (a kind of minnow) are led by a small number of experienced individuals who knew when and where food was available. Radakov estimated herring schools in the North Atlantic can occupy up to 4.8 cubic kilometres with fish densities between 0.5 and 1.0 fish/cubic metre.
The largest swarms are visible from space and can be tracked by satellite. One swarm was observed to cover an area of 450 square kilometers (175 square miles) of ocean, to a depth of 200 meters (650 feet) and was estimated to contain over 2 million tons of krill. Recent research suggests that krill do not simply drift passively in these currents but actually modify them. Krill typically follow a diurnal vertical migration.
By moving vertically through the ocean on a 12-hour cycle, the swarms play a major part in mixing deeper, nutrient-rich water with nutrient-poor water at the surface. Until recently it has been assumed that they spend the day at greater depths and rise during the night toward the surface.
Because of their smaller size and relatively faster growth rates, however, and because they are more evenly distributed throughout more of the world's oceans, copepods almost certainly contribute far more to the secondary productivity of the world's oceans, and to the global ocean carbon sink than krill, and perhaps more than all other groups of organisms together.
In his 1800 book, Phytologia: or, The philosophy of agriculture and gardening, Erasmus Darwin wrote that plant growth resembled swarms observed elsewhere in nature. While he was referring to more broad observations of plant morphology, and was focused on both root and shoot behavior, recent research has supported this claim.
Additional inputs that inform swarm growth includes light and gravity, both of which are also monitored in the transition zone of a root's apex. These forces act to inform any number of growing 'main' roots, which exhibit their own independent releases of inhibitory chemicals to establish appropriate spacing, thereby contributing to a swarm behavior pattern.
If one person was designated as a predator and everyone else was to avoid him, the flock behaved very much like a school of fish. Understanding how humans interact in crowds is important if crowd management is to effectively avoid casualties at football grounds, music concerts and subway stations. The mathematical modelling of flocking behaviour is a common technology, and has found uses in animation.
Bidirectional traffic can be observed in ant trails. In recent years this behaviour has been researched for insight into pedestrian and traffic models. Simulations based on pedestrian models have also been applied to crowds which stampede because of panic. Herd behaviour in marketing has been used to explain the dependencies of customers' mutual behaviour.
This could make them attractive for space exploration missions, where failure is normally extremely costly. In addition to ground vehicles, swarm robotics includes also research of swarms of aerial robots and heterogeneous teams of ground and aerial vehicles. Military swarming is a behaviour where autonomous or partially autonomous units of action attack an enemy from several different directions and then regroup.
Military swarming involves the use of a decentralized force against an opponent, in a manner that emphasizes mobility, communication, unit autonomy and coordination or synchronization. Historically military forces used principles of swarming without really examining them explicitly, but now active research consciously examines military doctrines that draw ideas from swarming.
How the Science of Swarms Can Help Us Fight Cancer and Predict the Future
The first thing to hit Iain Couzin when he walked into the Oxford lab where he kept his locusts was the smell, like a stale barn full of old hay.
The room was hot and humid, and the constant commotion of 20,000 bugs produced a miasma of aerosolized insect exoskeleton.
Biologists had already teased apart the anatomy of locusts in detail, describing their transition from wingless green loners at birth to flying black-and-yellow adults.
Couzin would put groups of up to 120 juveniles into a sombrero-shaped arena he called the locust accelerator, letting them walk in circles around the rim for eight hours a day while an overhead camera filmed their movements and software mapped their positions and orientations.
In 1995 a Hungarian physicist named Tamás Vicsek and his colleagues devised a model to explain group behavior with a simple—almost rudimentary—condition: Every individual moving at a constant velocity matches its direction to that of its neighbors within a certain radius.
Couzin figured out an elegant proof for the theory: “You can cut the nerve in their abdomen that lets them feel bites from behind, and you completely remove their capacity to swarm,”
Without obvious leaders or an overarching plan, this collective of the collective-obsessed is finding that the rules that produce majestic cohesion out of local jostling turn up in everything from neurons to human beings.
And the rules may explain everything from how cancer spreads to how the brain works and how armadas of robot-driven cars might someday navigate highways.
But it was only in the computer age—with the ability to iterate simple rule sets millions of times over—that this hazy concept came into sharp focus.
The kinds of collectives that undergo phase transitions, like liquids, contain individual units counted in double-digit powers of 10.
At high speed, with larger game boards, they were able to coax an astonishing array of patterns to evolve across their screens.
Sixteen years later, a computer animator named Craig Reynolds set out to find a way to automate the animated movements of large groups—a more efficient algorithm would save processing time and money.
It included behaviors like obstacle avoidance and the physics of flight, but at the heart of Boids were three simple rules: Move toward the average position of your neighbors, keep some distance from them, and align with their average heading (alignment is a measure of how close an individual’s direction of movement is to that of other individuals).
That would all be miraculous enough, but the flocks created by Boids also suggested that real-world animal swarms might arise the same way—not from top-down orders, mental templates of orderly flocks, or telepathic communication (as some biologists had seriously proposed).
Vicsek, the Hungarian physicist, simulated his flock in 1995, and in the late 1990s a German physicist named Dirk Helbing programmed sims in which digital people spontaneously formed lanes on a crowded street and crushed themselves into fatal jams when fleeing from a threat like a fire—just as real humans do.
All he had to do was tell his virtual humans to walk at a preferred speed toward a destination, keep their distance from walls and one another, and align with the direction of their neighbors.
By controlling only attraction, repulsion, and alignment (how similar a critter’s direction is to that of its neighbors), researcher Iain Couzin induced three different behaviors in a virtual collective, all akin to ones in nature.—Katie M.
When he increased the range over which alignment occurred even more, the doughnut disintegrated and all the elements pointed themselves in one direction and started moving together, like a flock of migrating birds.
When Couzin enters the room where the shiners are kept, they press up against the front of their tanks in their expectation of food, losing any semblance of a collective.
For example, when Couzin flashes light over the shiners, they move, as one, to shadier patches, presumably because darkness equals relative safety for a fish whose main defensive weapon is “run away.”
Each shiner, the theory goes, makes an imperfect estimate about where to go, and the school, by interacting and staying together, averages these many slightly wrong estimations to get the best direction.
When a disorganized group of shiners hits a dark patch, fish on the edge decelerate and the entire group swivels into darkness.
Thomas Seeley, a behavioral biologist at Cornell, used colored paint to mark bees that visited different sites and found that those advocating one location ram their heads against colony-mates that waggle for another.
In your brain, this thinking goes, different sets of neurons fire in favor of different options, exciting some neighbors into firing like the waggling bees, and inhibiting others into silence, like the head-butting ones.
The same dynamics can be seen in starlings: On clear winter evenings, murmurations of the tiny blackish birds gather in Rome’s sunset skies, wheeling about like rustling cloth.
Italian physicist Andrea Cavagna discovered their secret by filming thousands of starlings from a chilly museum rooftop with three cameras and using a computer to reconstruct the birds’
In most systems where information gets transferred from individual to individual, the quality of that information degrades, gets corrupted—like in a game of telephone.
But because the quality of the information the birds perceive about one another decays far more slowly than expected, the perceptions of any individual starling extend to the edges of the murmuration and the entire flock moves.
All these similarities seem to point to a grand unified theory of the swarm—a fundamental ultra-calculus that unites the various strands of group behavior.
In one paper, Vicsek and a colleague wondered whether there might be “some simple underlying laws of nature (such as, e.g., the principles of thermodynamics) that produce the whole variety of the observed phenomena.”
Biologists are used to convergent evolution, like the streamlining of dolphins and sharks or echolocation in bats and whales—animals from separate lineages have similar adaptations.
Either all these collectives came up with different behaviors that produce the same outcomes—head-butting bees, neighbor-watching starlings, light-dodging golden shiners—or some basic rules underlie everything and the behaviors are the bridge from the rules to the collective.
The British mathematician and inventor of the indispensable software Mathematica published a backbreaking 1,200-page book in 2002, A New Kind of Science, positing that emergent properties embodied by collectives came from simple programs that drove the complexity of snowflakes, shells, the brain, even the universe itself.
“I’m very cautious about suggesting that there’ll be an underlying theory that’ll explain the stock market and neural systems and fish schools,”
Tumors spawn populations of rogue, mobile cells that align with and migrate into surrounding tissues, following a subset of trailblazing leader cells.
The cool air has a tang of chlorine, thanks to a 20,000-gallon water tank, 20 feet across and 8 feet deep, home to four sleek, cat-sized robots with dorsal and rear propellers that let them swim in three dimensions.
Her robots can also track moving gradients, avoid each other, and keep far enough apart to avoid collecting redundant data—just enough programming to unlock more complex abilities.
The bots carry out missions with a feedback-controlled algorithm programmed into them, like finding the highest concentrations of oil in a simulated spill or collecting “targets”
Another group of researchers is trying to pilot a flock of unmanned aerial vehicles using fancy network theories—the same kind of rules that govern relationships on Facebook—to communicate, while governing the flocking behavior of the drones with a modified version of Boids, the computer animation software that helped spark the field in the first place.
Yet another team is working on applying flocking behaviors to autonomous cars—one of the fundamental emergent properties of a flock is collision avoidance, and one of the most important things self-driving cars will have to be able to do is not run into people or one another.
Every crow in a murder would carry miniature sensors that record its movements, along with the chemicals in its body, the activity in its brain, and the images on its retina.
It’s a souped-up version of the locust accelerator—combine real-world models with tech to get an unprecedented look at creatures that have been studied intensively as individuals but ignored as groups.
- On Monday, September 16, 2019
The Remote Agent Experiment: Debugging Code from 60 Million Miles Away
Google Tech Talk February 14, 2012 Presented by Ron Garret. ABSTRACT The Remote Agent Experiment: Debugging Code from 60 Million Miles Away The Remote Agent Experiment (RAX) was...
NASA in Silicon Valley Live - Episode 02 - Self-driving Robots, Planes and Automobiles
NASA in Silicon Valley Live is a live show streamed on Twitch.tv that features conversations with the various researchers, scientists, engineers and all around cool people who work at NASA...
AI: perspectives and applications
CMSoft Talk: AI perspectives and applications CMSoft talks perspectives and applications of AI and parallel processing at ITA - Instituto Tecnológico de Aeronáutica for ITABits and ITAndroids...
Doctoral Program Conference: #decoding, Session 1, Unsettling
3/11/16 Power inscribes order on space through codes. Bureaucratic codes measure and normalize dynamic ecologies and constitute the substrate of any infrastructural system, organization, and...