Optical Lattice Clock

Six laser beams create a pattern of standing waves that traps strontium atoms in energy wells. The trapping laser frequency is one that does not interfere with the atoms, which tick at about «9 terahertz, providing unsurpassed timekeeping accuracy

Six laser beams create a pattern of standing waves that traps strontium atoms in energy wells. The trapping laser frequency is one that does not interfere with the atoms, which tick at about «9 terahertz, providing unsurpassed timekeeping accuracy

Robot army will think for itself

Swarms of independent, cooperating robots are heading out to patrol borders and explore space

CELESTE BIEVER, BOSTON

JAMES MCLURKIN has a novel party trick - he can coax 20 small autonomous wheeled robots to form herds, disperse again, wheel in neat circles, sing a harmonic rendition of the theme from Star Wars, and automatically recharge from a power station.

McLurkin, a postgraduate student at the Massachusetts Institute of Technology, is trying to design robots that will work together and make collective decisions. If he succeeds, swarms of robots could one day be put to work in the home, in space and by the military. "A swarm or a team can collaborate to overcome what a single robot might not be able to do," explains Paolo Gaudiano, who works on swarms at Icosystem in Cambridge, Massachusetts.

Soon teams of up to 40 robots could be employed as border security guards and outside airports. Frontline Robotics in Ottawa, Canada, has installed collaborative software on its vacuum cleaner-sized PC-bots and its much larger vehicle-sized "Grunts", which it plans to put to work patrolling a runway at Ottawa airport. The firm has also sold Grunts to a South Korean company called DoDaam Systems, which is hoping to win a contract from the South Korean government to patrol its border with North Korea.

The patrolling robots will use Wi-Fi to share what they see, sniff and hear. They may even be able to triangulate the exact position of an intruder, or the source of a plume of smoke from an explosion, something no single robot could do. "The collaboration software from Frontline Robotics will be key for this application," says Jonathan Lee of DoDaam Systems, based in Daejeon.

But with each PC-bot costing $5000, and Grunts a cool $150,000, these robots are a far cry from the looo-strong swarms of robots many in the field dream of. Aside from being too big and expensive, their software is only semi-autonomous - Frontline robots also need maps and GPS to navigate. While this may be enough to patrol airport runways and borders, these robots probably couldn't cope with uncharted terrain, space or even indoors, where there are no GPS signals.

Furthermore, the size of Frontline's robot team is limited, because each robot has to report back to a single "leader" robot. So far Frontline has only used two robots at a time, but its

"When a pherobot finds an object of interest it em its a series of artificial pheromones"

simulations show that teams of more than 48 will run out of bandwidth with which to communicate.

The ideal is swarms of robots that need no central control. And McLurkin's robots have proved the principle that, equipped with the right algorithms, swarms of hardware can have autonomous control. Last year, for instance,

in a demonstration for the US government, 108 of his creations, 12-centimetre-cube robots that use infrared to detect obstacles and communicate with one another, successfully spread out evenly to fill 280 square metres of floor space in just 25 minutes.

David Payton and colleagues at HRL Research Labs in Malibu, California, avoid the need for centralised control by using "virtual pheromones". Each of Pay ton's "pherobots" is a wheeled cylinder about 7 centimetres tall, 11 centimetres in diameter, and equipped with a transceiver that beams out and receives infrared light. A PDA and several attached chips process the IR signals and steer each pherobot.

When a pherobot finds something of interest, it sends out a series of IR pulses - the artificial pheromones - that code for a number. Its nearest neighbours pick up the signal, increase the number by 1, then re-transmit it to the nearest bots while also saving the number. This continues until a trail of bots has formed with numbers increasing incrementally the further | they are from the | source. It means | other bots can follow this | trail simply by interrogating each botthey meet.

Although McLurkin and Pay ton's teams are funded by the Pentagon, their robots have no obvious military use as yet. And forming patterns and trails on the ground are just party tricks compared to what collaborative robots might do one day. "It looks cool, but ultimately we need to expand into 3D space," says Payton. "We haven't found the killer application yet," he says.

That day may not be far off. NASA and Boeing staff working on the US military's $21 billion Future Combat Systems project say they are very interested in testing the capabilities of swarms. "Buoyed by the success robots are having doing real missions, the need for swarms is being raised to new levels. People are starting to get the concept that these robots are practical and viable," says Colin Angle, co-founder of iRobot in Burlington, Massachusetts. •

Letters

Life and liberty

From Jerry Monk Roger Hicks and Sean Williams welcome the idea of a national DNA database (30 April, p 28). Are we all to be doomed to repeat the lessons of history because the likes of Hicks and Williams will not learn them? As Mark Griffith pointed out in the same issue, contrary to Hicks's assertion that DNA profiles will reliably reveal the identity of a criminal, their use will lead to many gross miscarriages of justice once the courts start to rely on their infallibility and the criminals get wise to the possibilities afforded by used condoms in the street. Faced with DNA evidence, it is most unlikely that the police will even look for the real rapist in such a case.

It is 60 years since my father was a guard at the Belsen concentration camp. The inmates were there in part because the efficient and admirably apolitical German civil service had, with the best of intentions, carefully and conscientiously collected, correlated and filed data on everybody. That enabled the legally elected government to round up gypsies, Jews and other "Untermenschen" with ease. Think what the Gestapo could have done with a DNA database.

It couldn't happen here? During the 20th century governments, most of them legitimate, have slaughtered at least 30 million of their own people. If there is one lesson to be learned it is that governments cannot be trusted. Even if you can trust the present government, what about the next one?

It is not concern for any namby-pamby civil liberties that leads me to oppose DNA databases and ID cards, but for my life. Walton-on-the-Naze, Essex, UK

Public says no

From Douglas Parr, Greenpeace UK William Cullerne Bown takes a rather hackneyed view in seeing debates about science as being "pro" or"anti" (30 April, p 21). Numerous studies of public attitudes have shown that there is no anti-science culture, rather there is considerable public questioning about who is developing what, for what purposes, for whose benefit and with what consequences. When the sums add up, as with the use of mobile phones, people are happy to engage. When they don't - as they most clearly don't with genetically modified organisms - generally people don't want the product. London, UK

Powerful links

From Angela Cropper and Harold Mooney, co-chairs, Millennium Assessment Panel In his letter about the Millennium Ecosystem Assessment (MA), Steve Lonergan says the original article was correct in its assertion that the assessment "had no direct link to the people in power" (14 May, p 28). Nothing could be further from the truth. The MA was first called for in 2000 by the UN secretary general in his Millennium Report to the General Assembly. Governments involved in four international conventions (covering biodiversity, desertification, wetlands and migratory species) took decisions supporting the creation of the assessment, requested specific information from it, and appointed representatives of the conventions to serve on the MA board.

The MA is unique among international assessments in that it has strong and formal ties to governments through these international conventions, but also involves the private sector, NGOs and indigenous groups in the governance of the process. In considering the strengths and weaknesses of this arrangement, the members of the MA board concluded in January 2005 that it was premature to state whether an exclusively intergovernmental process was preferable to a multi-stakeholder process with intergovernmental authorisation.

Based on the experience of the MA, at least one new global assessment - the International Agricultural Science and Technology Assessment- has been designed with a mix of intergovernmental and multi-stakeholder oversight.

Governments are not the only institutions taking decisions on environment and development issues, and indeed are sometimes not even the most important institutions in this regard. The UN can strengthen its own role by embracing processes that seek to involve all the key decision-makers, not just governments.

Stanford, California, US, and Laventille, Trinidad and Tobago

Live long and prosper

From David Friend Mary Midgley suggests that increased longevity will result in overpopulation (30 April, p 28). Not necessarily: long-lived honey bees tend to produce fewer young, whereas short-lived fish breed faster (30 April, p 18). It is the short-lived, fast-breeding "minisleepers" mentioned on the same page who will use up the world's food, mineral and energy reserves. Longdown, Devon, UK

From Mark Bruce Mary Midgley and Chris James show highbrow disdain for us "immortalists". I think I speak for many so-called immortalists when I say that I would be more than happy to be sterilised if this was the requirement for an indefinitely prolonged life. The overpopulation problem would thus be solved. Adelaide, South Australia

Gigantic gas leak

From GeoffreyMentink The strategy of underground carbon dioxide sequestration seems like it is an accident waiting to happen (30 April, p 26). I can see the day coming when, after years of pumping C02 underground, an earthquake liberates the stored gas and suffocates all plants and animals in its vicinity.

Warrandyte, Victoria, Australia

0 0

Post a comment