Trinity College Create Worlds Smallest Shamrock

photo credit: AMBER national material science centre, Trinity College, Dublin, Ireland.

Unless you live under a rock, you probably know that today is St Patrick’s day.

Trinity College in Dublin decided to celebrate the day in their own way, by creating the worlds smallest shamrock. AMBER, Ireland’s national materials science centre, based at Trinity College, etched a nano sized shamrock on to a Trinity College lapel pin, which was later presented to the recipient of the SFI St Patrick’s Day Science Medal in Washington DC on March 13th.

AMBER, Trinity College, Dublin.

The etching is 200,000 times smaller than a grain of salt and 500 could sit side by side on a single strand of human hair. It was created using the AMBER Helium Ion Microscope, which enables extremely high resolution imaging of less than 1 nanometre. Check out the AMBER microscope at work here

What will we find next inside the Large Hadron Collider?

What lies within? Maximilien Brice/CERN, CC BY-NC

The Large Hadron Collider, the world’s largest scientific experiment, is due to restart this month after two years of downtime for maintenance and upgrading. There’s no doubt that having played its role in the discovery of the Higgs boson in 2012, what the media christened the “God particle”, expectations for what the 27km particle accelerator at CERN could achieve this time have certainly been set high.

The Higgs boson is a possible explanation for the origin of mass, something predicted in 1964 by Peter Higgs and several other physicists, and the discovery of which led to the award of a Nobel Prize for physics for Higgs and François Englert in 2013.

So why did it take so long to discover it? As Einstein showed in his mass-energy equivalence (E=MC²), the mass of a particle is a measure of its energy content. If a particle is more massive, it has a greater energy content, and conversely to create a massive particle requires a great deal of energy. So simply put, it wasn’t until the Large Hadron Collider (LHC) was capable of colliding beams of protons with sufficient energy that the Higgs Boson could be created with its mass of 126 billion electron volts (gigaelectronvolts, or GeV). In particle physics it is usual to give masses in terms of energy, and while 126GeV is equivalent to only 2.24x1025kg, this mass is about 127 times larger than a single proton.

So the intention is that following a two-year upgrade the LHC’s new, more powerful electromagnets will be sufficient to accelerate two beams of protons to 6.5 trillion electron volts (teraelectronvolts, or TeV), increasing the potential collision energy from 8TeV in 2012 to 13TeV. And with greater collision energy comes the possibility of creating and detecting new particles of even greater mass. The expectation is that the LHC’s experiments could uncover new particles known as Z particles, new Higgs bosons, and even particles of dark matter.

A map of subatomic particles, known and hypothesised. MissMJ, CC BY-SA

From Higgs to Z

Discovered at CERN in 1983, the Z particle is a force carrier – a particle that carries one of the four fundamental forces of nature: the gravitational, electromagnetic, strong and weak forces. The Z particle carries the weak force, which is implicated in subatomic reactions. A related, theorised particle that could be next to be discovered is the Z prime particle, or Z’. This would help our understanding of gravitons, the carriers of the gravitational force that are theorised but have not yet been detected.

Taking the constituents of the universe as a whole, we have a good understanding of about 5% of it. The remaining 95% is made up of about 68% dark energy and 27% dark matter. With a little over 84% of the universe’s mass being dark – not detectable by any known means – if the LHC can in some way shed some light on the nature of this matter it will move our understanding of the universe forward.

With an upgraded LHC able to provide higher collision energies and the possibility of creating new particles – whether those currently theorised or not – it will have a significant impact on our fundamental understanding of the laws of nature and the accepted model that is used to try and explain them.

Some may point to the cost of the LHC upgrade, estimated at around £70m, as a cost beyond the public purse in these cash-strapped times of austerity. But the possibilities for what it can add to our understanding of the world cannot be ignored either, nor the benefits they might have in other areas, for example medical imaging. Considering how regularly sums far larger than £70m of taxpayers’ money are squandered, CERN’s role as a global educational tool for physicists, mathematicians and engineers must be considered excellent value for money.

This article was originally published on The Conversation. Read the original article.

Apple may have arrived late to the party, but with Watch it’s brought a gun to a swordfight

It’s arrived: Tim Cook’s watch of many colours. Kay Nietfeld/EPA

While all eyes and ears were trained on news of its smartwatch, Apple also used its spring Keynote to introduce changes to Apple TV, revisions to its laptop lineup, and a new service that builds on the health monitoring aspects of smartwatches to perform data collection for medical research.

As one digital TV service after another launches many have been left wondering when HBO, whose television dramas are highly sought and widely watched properties, would play its hand. And here it is: a partnership with Apple that makes the entire HBO back catalogue available through the new HBO Godigital streaming service, available exclusively through Apple TV. So while the Apple TV hardware hasn’t been updated for years, the partnership with HBO (and a price drop to £59) is a nice reminder for those who may have overlooked it.

Apple has extended its reach into car dashboards with CarPlay, into home automation with HomeKit, and into health monitoring with HealthKit. Apple hopes that ResearchKit, a new open-source API and service, will form the foundation for apps that can collect health data from larger numbers of volunteers, increasing sample sizes and frequency of data collection, making the data more useful for researchers. Five apps have been developed so far, to investigate Parkinson’s Disease, asthma, diabetes and cardiovascular disease with research groups in leading hospitals. There is an emphasis on privacy, with the user controlling the degree of information that is being shared.

The new Macbook – neither Air nor Pro – comes with the latest retina display, a faster, more energy efficient processor, and a trackpad that can supply tactile feedback. In a 12″ format that fills out the line between 11″ and 13″, it is lighter and thinner even than the Air, has a re-engineered keyboard and somewhat controversially rolls many ports into just one: the USB-C standard port, which will handle HDMI video, external hard drives and other USB peripherals. Inevitably this is going to mean buying another set of cables.

The new Macbook Air with shiny retina screen. Kay Nietfeld/EPA

Watch my watch

In any other keynote this reveal would have been the main news item. But of course the main event was the watch. Seven months since Tim Cook first revealed the device, it’s been a long wait for more technical details. Opinion is still split on whether it will be a hard sell. With fewer people wearing watches anyway, the market is split between those who want a fitness tracker and those that want a beautiful luxury object. Is there a need for a device which essentially duplicates the functionality of a smartphone? Apple has to convince us that the watch offers more, in clear terms of where glancing at a watch is preferred to pulling out a phone.

Usually reserved to only one or two colours, this time Apple offers 20 different combinations of ways to customise the watch in size, colour, watch and strap material – probably a necessity in order to sell a device that by nature of being frequently visible is more fashion than function.

One watch to rule them (but in many colours). Martyn Landi/PA

The styling of the watch itself is reminiscent of the first iPhone, with three versions in two different sizes, 38mm or 42mm high: the cheapest Apple Sport at £299 with an aluminium body and plastic straps, the middle tier Apple Watch from £479 in stainless steel and wrist bands in leather, steel or plastic, and the gold Apple Watch Edition, which starts at £8,000 – perhaps more expensive even than the Apple Lisa from 1983, which sold at US$15,000 at the time.

All the information, but smaller and nearer. Apple

Most of the functionality of the watch requires an iPhone within a few metres – maps, messages, Siri and other apps are relayed from the phone using WiFi or mobile data. Apple suggests that the battery will last 18 hours in a typical day.

Not first to market, but best?

Apple invests heavily in research and development to create new devices and interfaces that differentiate its products, at least, until competitors release their responses. Apple’s watch uses an Ion-X glass or Sapphire crystal screen which is pressure-sensitive to varying degrees. The side-mounted dial, which Apple terms a digital crown, enables scrolling and clicking, and a button below it jumps to frequent contacts. It has a “Taptic” engine which provides vibration feedback for certain apps, for example suggesting directions in Maps. The sensors on watch’s underside detect heartbeat and combine with the accelerometer to measure physical activity, something Apple is pitching as a major selling point.

Developers are already creating software that will extend their iPhone apps to interact with and be accessible from the watch, as Apple has with its Apple Pay contactless payment system. Miniature messages appear on the device in what Apple calls Glances, giving the impression of dealing with such messages quickly without the hassle of pulling out a phone.

From watches to smartwatches, with only a little relief. XKCD, CC BY-NC

Will it sell? In the past 18 months customers have bought 5m smartwatches or fitness bands, with Samsung flooding the market with many smartwatch devices, but with fitness bands accounting for the majority of sales. Current estimates suggest that Apple could sell more than 8m watches, eight times as many as its largest competitor.

While many of its features will appear in competitor’s smartwatches in the subsequent years, for the moment the eponymous watch is best in class. To sound a note of caution: like the first generation iPhone, the second generation device will probably be half as deep and run twice as long. You may be unfazed about the risks of being an early adopter, but if the idea of paying another few hundred pounds for the latest model next year isn’t appealing, it may be sensible to wait.

This article was originally published on The Conversation. Read the original article.

The best medical images of the year: a mesmerising nano-world where science becomes art

Here’s looking at you. Kevin Mackenzie, University of Aberdeen

Since they were pioneered by Robert Hooke 350 years ago, microscopes have been extending our vision. In the 21st century,scanning electron microscopy (SEM) and confocal microscopy, which uses a pinhole to remove out-of-focus light and allows 3D structures to be built from multiple images, have pushed the boundaries of resolution. Further still, nanoscopes that use fluorescence to circumvent limitations in SEM are already winning Nobel prizes.

The Wellcome Images Awards 2015 are a showcase of the best in science imaging techniques, and this year’s crop of winners gives a rich glimpse into a world normally unseen by the human eye. And as we learn more about the intricacies of our bodies and the world around us, the images transform abstract ideas and the constantly moving machine of our bodies into works of art and visual drama. Four nominees for this year’s prize tell us how they got their image.

---

Robert Marc, Calvin and JeNeal Hatch Presidential Endowed Chair, University of Utah

Distribution of metabolites in a mouse kidney, CMP. Jefferson Brown, Robert Marc, Bryan Jones, Glen Prusky, Nazia Alam

We have been perfecting a technology for imaging the metabolism of cells. The core principle is microscopically visualising arrays of up to a dozen different small molecules and computationally fusing these into readable colour images. The molecules are detected by immune probes developed in my laboratory and each mixture of molecules forms a cellular “signature” colour.

The molecules in this image are mapped into image colour channels: aspartic acid as red, glutathione as green and glutamine as blue. Aspartic acid links amino acid and energy metabolism. Glutathione is our primary defence against oxidative damage. Glutamine is a molecular “battery” that stores and distributes amino groups to other molecules.

The image is a thin slice through the various kinds of filtering cells of the mouse kidney. Surprisingly, the images reveal that all tissues, not just kidney, are rainbows of different cell types, each with a unique metabolism. This contradicts our expectation that metabolism is roughly the same in all cells and opens the door for a new generation of molecular tests of cell state in health and disease. Though the kidney yields a spectacular and fiery display, we have yet to craft a theory that explains it.

---

Gregory Szeto, Koch Institute, MIT

Drug-releasing depots in mouse lungs. Gregory Szeto, Adelaide Tovar, Jeffrey Wyckoff, MIT

This image is one in a series we captured to examine the distribution of engineered, micrometer-sized drug particles (red) targeted to mouse lungs (blue, green). Scale in science and engineering is fascinating, and as technology has advanced, we’ve increased our ability to take both smaller and larger images with detail and range. Using fluorescent imaging and a confocal laser microscope, we imaged whole lungs at different times to determine where drug particles go.

Information at the organ scale is important when designing particles to obtain even, specific and prolonged delivery of drugs throughout the organ. We can now use whole organs (as opposed to traditional thin slices of tissue on glass slides) to image distribution at this level while also acquiring information about smaller details. This same set of images (more than 1,000 total condensed into a single 2D image) can be made into an interactive 3D model where we can zoom in and out of the organ to find out where our particles are aggregating and distributing, potentially down to single cells.

Speaking as both a scientist and an artist (avid photographer and writer), there is a great deal of overlap between the two areas. Science, technology, engineering, and math (STEM) are often set apart from other fields, particularly arts and humanities. In reality, there is so much art in STEM: the visual beauty of an experiment image is not unlike a moving painting or photo; the elegance of manuscript text can rival that of the best prose. Ultimately, innovative scientific research is driven partially by the same intuition that often drives and inspires great art. Striking visuals are just one way for us to stir curiosity in everyone about the work that we love.

---

Michael Hausser, Professor of Neuroscience, University College London

Purkinje cell and dendritic tree, rat cerebellar cortex, SEM Michael Hausser/Sarah Rieubland/Arnd Roth/UCL

This is an image of a single Purkinje cell in the cerebellum captured using a focused ion beam scanning electron microscope. This kind of electron microscope is relatively new to biology, and uses two beams to image neurons within a block of brain tissue: the first beam etches away the top few nanometers of tissue, and the second beam then captures the image. In this way, neurons and circuits can be imaged at extremely high (nanometer) resolution, while keeping the block of tissue intact (without the loss of resolution and distortions resulting from conventional knife-based sectioning of tissues for electron microscopy).

I find this image endlessly fascinating: not only does it reveal the incredible richness and complexity of the dendritic tree of an important cell type (the processing carried out by Purkinje cells is crucial for precisely timed learned movements); but it also exhibits a beauty that is both ethereal yet precise, like a Japanese woodcut.

---

Khuloud Al-Jamal, Senior Lecturer in Nanomedicine, King’s College London

Brain astrocyte cell taking up carbon nano-needles, SEM Khuloud T. al-Jamal, Serene Tay, Michael Cicirko

The blood-brain barrier is a protective layer of cells that regulates entry of molecules to the brain. Despite acting as a protective mechanism, it also acts as a barrier to delivering drug therapies to the brain. Astrocytes, specialised star-shaped glial cells (non-neuronal cells that support and protect), outnumber neurons over fivefold. These cells contiguously tile the entire central nervous system (CNS) and exert many essential complex functions.

Astrocytic tumours are the commonest type of cancer of the brain. Over the past 40 years systems to deliver drugs to treat these cancers have emerged. Carriers that have been investigated have ranged in size from 10 to 1,000 nanometres. One example of such carriers is a carbon nanotube, a hollow cylinder made from a continuous unbroken hexagonal mesh of carbon molecules, first explored by electron microscopist Sumio Iijima in 1991.

Our image is a scanning electron micrograph of an astrocyte cell (coloured in brown), with a diameter of about 20 micrometers, and captured in the process of taking up carbon nanotubes (green colour). New approaches to brain drug delivery will be the key to unlocking the brain and tackling neurological disorders that would otherwise remain incurable.

---

Best of the rest

Albert Cordona, Janelia Group Leader, Howard Hughes Medical Institute

Nervous system in a fruit fly larva, serial section TEM Albert Cordona/HHMI

Mark Bartley, Cambridge University Hospitals NHS Trust

Elderly woman with Kyphosis (curvature of the spine) Mark Bartley

Cutting-edge imaging techniques may delve into the smallest worlds but photographs still powerfully capture our dramatic outer shell as this clinical photograph of an elderly woman’s curved spine shows.

Michael Frank, Photographer, Royal Veterinary College

Reticulum (stomach chamber) Michael Frank

If you’ve never eaten tripe but have wondered what a stomach looks like, here’s an image of one taken from a goat.

Kevin Mackenzie, Microscopy Manager, Institute of Medical Sciences, University of Aberdeen

Kevin Mackenzie, University of Aberdeen

A scanning electron micrograph of a greenfly’s eye.

Nele Dieckermann and Nicola Lawrence, University of Cambridge

Natural killer (NK) cell immune synapse, 3D-SIM Dieckermann and Lawrence, Cambridge University

Natural killer (NK) cells in our immune system are at the frontline of our immune response and are rapidly deployed by the body to fight infections or growing tumours.

Luis de la Torre-Ubieta, Postdoctoral Fellow, Geschwind Lab, UCLA

Mouse brain, coronal view. Luis de la Torre-Ubieta

To see startling Wellcome Images 2014 showing strange and beautiful science up close, click here.

This article was originally published on The Conversation. Read the original article.

Dawn of a new era: the revolutionary ion engine that took spacecraft to Ceres

I’ve seen the future, and the future’s blue. NASA

The NASA spacecraft Dawn has spent more than seven years travelling across the Solar System to intercept the asteroid Vesta and the dwarf planet Ceres. Now in orbit around Ceres, the probe has returned the first images and data from these distant objects. But inside Dawn itself is another first – the spacecraft is the first exploratory space mission to use an electrically-powered ion engine rather than conventional rockets.

The mysterious bright spot of Ceres seen as Dawn approaches. NASA

The ion engine will propel the next generation of spacecraft. Electric power is used to create charged particles of the fuel, usually the gas xenon, and accelerate them to extremely high velocities. The exhaust velocity of conventional rockets is limited by the chemical energy stored in the fuel’s molecular bonds, which limits the thrust to about 5km/s. Ion engines are in principle limited only by the electrical power available on the spacecraft, but typically the exhaust speed of the charged particles range from 15km/s to 35km/s.

What this means in practice is that electrically powered thrusters are much more fuel efficient than chemical ones, so an enormous amount of mass can be saved through the need for less fuel onboard. With the cost to launch a single kilogramme of mass into Earth orbit of around US$20,000, this can make spacecraft significantly cheaper.

This can be of great benefit to commercial manufacturers of geostationary satellites, where electric propulsion can allow them to manoeuvre adding new capabilities to the satellite during its mission. However, for scientific missions such as interplanetary travel to the outer regions of the Solar System, electric propulsion is the only means to carry useful scientific payload quickly across the enormous distances involved.

Inside an ion engine. NASA

Electric space power

There are three broad types of electric propulsion, depending on the method used to accelerate the fuel.

Electrothermal engines use electric power to heat the propellant either by passing a current through a heating element, a configuration known as a resistojets, or by passing a current through the hot ionized gas or plasma itself, an arcjet.

Electromagnetic engines ionise the propellant by turning it into an electrically conductive plasma, which is accelerated via the interaction of a high electrical current and a magnetic field. Known as pulsed plasma thrusters, this technique is in fact quite similar to how an electric motor works.

Electrostatic engines use an electric field generated by applying a high voltage to two grids perforated with many tiny holes to accelerate the propellant, called a gridded ion engine, which is what powers Dawn. Another electrostatic design is the Hall effect thruster, which operates in a similar fashion but instead of high voltage grids generates an electric field at the thruster’s exit plane by trapping electrons in a magnetic field.

Half a century in the making

The concept of electric propulsion has been around for 50 years or more, but was deemed too experimental to commit to major projects. Only now is it beginning to find real applications. For example, keeping geostationary satellites in their correct orbit, to counteract the aerodynamic drag from the very tenuous atmosphere 200km above the Earth. Or interplanetary missions such as Deep Space 1 – the first experimental mission to use ion engines, it was originally intended as a technology demonstrator but performed a successful fly-past of the asteroid 9969 Braille and the comet Borrelly 15 years ago.

Another very successful mission using ion engines was the ESA Gravity field and steady-state Ocean Circulation Explorer (GOCE) satellite which for four years until 2013 was able to map in unprecedented detail the Earth’s gravity field.

The Dawn spacecraft, equipped with large solar panels to power its electrical engine. NASA

Future designs

Now that electric spacecraft engines have entered mainstream use, they look set to reduce the cost of deploying satellites. With compact ion engines onboard, satellites can raise themselves from low Earth orbit to their final geostationary orbit under their own power. This will save enormous amounts of fuel required to lift the satellite through conventional chemical rockets, and allow the use of much smaller launch vehicles which will save a lot of money. Boeing was the first off the blocks in 2012 with an all-electric version of their 702 platform satellite fitted with xenon-powered gridded ion engines, and other satellite manufacturers are following suit.

Currently all electric power designs use xenon gas as the propellant, but the search is on for alternative propellants since xenon is enormously expensive and in limited supply. But electrical power is here to stay, and over the longer term, space tugs and even manned missions to Mars based on nuclear electric propulsion will be the next on the drawing board.

This article was originally published on The Conversation. Read the original article.

Researchers Achieve Record-Smashing Wireless Connection Speeds

Do you know what wireless speeds of one terabit per second gets you? One hundred full-length movies downloaded onto your phone in just three seconds. Researchers in the U.K. are now saying that they’ve achieved just that: 1 Tbps over a 5G connection for the first time ever. That’s the fastest wireless speed to date.

“This is the same capacity as fiber optics but we are doing it wirelessly,” Rahim Tafazolli of the 5G Innovation Centre (5GIC) at the University of Surrey tells V3. Their tests were carried out under lab conditions over a distance of 100 meters (328 feet) using transmitters and receivers built at Surrey.

The speed is more than 65,000 times faster than average 4G download speeds (that’s just 15 megabits per second, or Mbps). And it also smashes the previous 5G (or fifth generation) speed record achieved last October in tests: Samsung’s 7.5 gigabits per second (Gbps), BBC reports, is less than one percent of the Surrey team’s speed.

Tafazolli and colleagues plan to take the technology out of the lab and onto the campus by as early as 2016, before demonstrating it to the public in early 2018. Unfortunately, that doesn’t necessarily mean we’ll be able to use it soon. “An important aspect of 5G is how it will support applications in the future. We don’t know what applications will be in use by 2020, or 2030 or 2040 for that matter, but we know they will be highly sensitive to latency,”he tells V3. Latency is the slowdown that’s introduced by infrastructure, Quartz explains, and it could remain a problem for decades. “We need to bring end-to-end latency down to below one millisecond so that it can enable new technologies and applications that would just not be possible with 4G,” Tafazolli says. Everything from playing 3D holographic games of chess over multiple smartphones to controlling connected cars over 5G may require rapid latency, he adds.

These results were announced at the V3’s mobile conference this week.

Leonard Nimoy Has Passed Away At Age 83

Actor Leonard Nimoy, most famous for his role as the half-human/half-Vulcan Mr. Spock in the Star Trek franchise, passed away in his Los Angeles home on Friday morning at the age of 83. Nimoy had been hospitalized last week due to chest pains that were connected to his chronic obstructive pulmonary disease. The actor attributed the disease to his smoking habit from his youth, though he quit 30 years ago. Complications of the disease ultimately took his life.

Born in Boston, Massachusetts in 1931, Nimoy began acting at an early age. After playing a number of bit roles in television and film, his big break came in 1966 when he was cast as Spock on Star Trek. Over the years, he reprised the role a number of times in the Star Trek: The Next Generation, Star Trek movies, video games, video shorts, and more. His last film credit was Star Trek: Into Darkness.

Leonard Nimoy gave life to a character who has become absolutely iconic in American entertainment and became a beacon of logic and reason. Many of his friends and co-stars took to social media to express their sadness of the news of his passing:

"I loved him like a brother. We will all miss his humor, his talent, and his capacity to love." -William Shatner http://t.co/U8ZN98tVYp

— William Shatner (@WilliamShatner) February 27, 2015

Leonard Nimoy brought us one of the greatest, noblest characters in the history of American storytelling. Someone find the Genesis planet.

— Seth MacFarlane (@SethMacFarlane) February 27, 2015

Befitting of the way Nimoy lived his life and the character he portrayed, his final tweet embodies his way of balancing emotion and logic:

A life is like a garden. Perfect moments can be had, but not preserved, except in memory. LLAP

— Leonard Nimoy (@TheRealNimoy) February 23, 2015

Rest in peace, Leonard Nimoy. Thank you for teaching us to live long and prosper. \\//

Three Men Receive Bionic Hands Controlled With Their Minds

photo credit: Screenshot of YouTube video/ New Scientist

The outlook used to be pretty bleak for those who had lost movement in their limbs due to severe nerve damage, but over the last year or so, some incredible advances have been made that are restoring shattered hope for many.

The amazing breakthroughs include spinal cord stimulation that allowed paralysed men to regain some voluntary control of their legs, a brain implant that enabled a quadriplegic man to move his fingers, and a system that allowed a paralysed woman to control a robotic arm using her thoughts. Science has definitely been on a roll, but this winning streak isn’t showing any signs of slowing down. Now, the world’s first “bionic reconstructions” have been performed on three Austrian men to help them regain hand function. This technique enabled the newly amputated patients to control prosthetic hands using their minds, allowing them to perform various tasks that most people take for granted.

The men that underwent the procedure had all suffered serious nerve damage as a result of car or climbing accidents, which left them with severely impaired hand function. The nerves that suffered injury were those within a network of fibres supplying the skin and muscles of the upper limbs, known as the brachial plexus. As lead researcher Professor Oskar Aszmann explains in a news release, traumatic events that sever these nerves are essentially inner amputations, irreversibly separating the limb from neural control. While it is possible to operate, Aszmann says the techniques are crude and do little to improve hand function. However, his newly developed procedure is quite different, and is proving to be a success.

Before the men could be fitted with their prosthetic hands, the researchers had to do some preliminary surgical work in which leg muscle was grafted into their arms in order to improve signal transmission from the remaining nerves. After a few months, the fibres had successfully innervated the transplanted tissue, meaning it was time to start the next stage: brain training.

Using a series of sensors placed onto the arm, the men slowly began to learn how to activate the muscle. Next, they mastered how to use electrical nerve signals to control a virtual hand, before eventually moving on to a hybrid hand that was affixed to their non-functioning hand. After around nine months of cognitive training, all of the men had their hand amputated and replaced with a robotic prosthesis that, via sensors, responds to electrical impulses in the muscles.

A few months later, the men had significantly improved hand movement control, which was highlighted by a test of function known as the Southampton Hand Assessment Procedure. As reported in The Lancet, before the procedure, the men scored an average of 9 out of 100, which soared to 65 using the prosthetic. Furthermore, the men reported less pain and a higher quality of life. For the first time since their injuries, they were able to perform a variety of tasks such as picking up objects, slicing food and undoing buttons with both hands.

“So far, bionic reconstruction has only been done in our centre in Vienna,” said Aszmann. “However, there are no technical or surgical limitations that would prevent this procedure from being done in centres with similar expertise and resources.”

[Via The Lancet, The Lancet news release and New Scientist]

Mathematicians Work Out Zombie Apocalypse Plan

photo credit: Nemar74/Shutterstock. If zombies strike, get out of the city

Cornell graduate students have adopted disease modeling to find the best response in the event of a zombie epidemic. They haven’t reached the conclusion that the undead are about to rise from their graves and eat us all, but they claim the work could be useful in planning for more likely disease outbreaks.

“Modeling zombies takes you through a lot of the techniques used to model real diseases, albeit in a fun context,” says Alex Alemi, a Ph.D. student in Cornell’s Physics Department. “A lot of modern research can be off-putting for people because the techniques are complicated and the systems or models studied lack a strong connection to everyday experiences. Not that zombies are an everyday occurrence, but most people can wrap their brains around them.”

Alemi and his colleagues are not the first to enlist zombies in disease outbreak planning. The Centers for Disease Control and Prevention provide a zombie preparedness plan that, as we’ve noted before, involves pretty much the same things recommended for all sorts of other emergencies. (Although it leaves out the garlic and sharp sticks handy for vampire outbreaks.) Even the Pentagon got in on the act.

However, Alemi’s team are trying to introduce a little more scientific rigor into their zombie plans. They took the techniques epidemiologists use to forecast outbreaks of infectious diseases and applied them to zombies running loose across the United States. “At their heart, the simulations are akin to modeling chemical reactions taking place between different elements; in this case, we have four states a person can be in—human, infected, zombie, or dead zombie—with approximately 300 million people.”

Modeling what might happen in a small community is hard—a single heroic individual or a particularly lively zombie could change the course of events. But as the numbers of humans and zombies run into the millions, probabilistic techniques come to the fore. “Each possible interaction—zombie bites human, human kills zombie, zombie moves, etc.—is treated like a radioactive decay, with a half-life that depends on some parameters,” says Alemi.

Not surprisingly, the modeling turned up a lot of problems with the way outbreaks are represented in films and books. Instead of a near-simultaneous apocalypse with a few communities holding out, the team found that cities would quickly become zombified, but most rural areas would stay safe for weeks or months.

“I’d love to see a fictional account where most of New York City falls in a day, but upstate New York has a month or so to prepare,” says Alemi, who has history with the undead. He once performed a rap based on the hypothesis that quantum physicist Paul Dirac was a vampire. Sadly, we don’t think Hollywood will be beating at his door so they can describe the next blockbuster as “scientifically endorsed.”

Alemi concludes that the safest place in the U.S. to try to wait things out would be the northern Rockies. Good to know.

Machines master classic video games without being told the rules

Imagine a machine that can learn things from scratch, no pre-programmed rules. What could it do? Flickr/Marco Abis, CC BY-NC-ND

Think you’re good at classic arcade games such as Space Invaders, Breakout and Pong? Think again.

In a ground breaking paper published today in Nature, a team of researchers led by DeepMind co-founder Demis Hassabis reported developing a deep neural network that was able to learn to play such games at an expert level.

What makes this achievement all the more impressive is that the program was not given any background knowledge about the games. It just had access to the score and the pixels on the screen.

It didn’t know about bats, balls, lasers or any of the other things we humans need to know about in order to play the games.

But by playing lots and lots of games many times over, the computer learnt first how to play, and then how to play well.

A machine that learns from scratch

This is the latest in a series of breakthroughs in deep learning, one of the hottest topics today in artificial intelligence (AI).

Actually, DeepMind isn’t the first such success at playing games. Twenty years ago a computer program known as TD-Gammonlearnt to play backgammon at a super-human level also using a neural network.

But TD-Gammon never did so well at similar games such as chess, Go or checkers (draughts).

In a few years time, though, you’re likely to see such deep learning in your Google search results. Early last year, inspired by results like these, Google bought DeepMind for a reported UK£500 million.

Many other technology companies are spending big in this space.

Baidu, the “Chinese Google”, set up the Institute of Deep Learning and hired experts such as Stanford University professor Andrew Ng.

Facebook has set up its Artificial Intelligence Research Labwhich is led by another deep learning expert, Yann LeCun.

And more recently Twitter acquired Madbits, another deep learning startup.

What is the secret sauce behind deep learning?

Geoffrey Hinton is one of the pioneers in this area, and is another recent Google hire. In an inspiring keynote talk at last month’s annual meeting of the Association for the Advancement of Artificial Intelligence, he outlined three main reasons for these recent breakthroughs

First, lots of Central Processing Units (CPUs). These are not the sort of neural networks you can train at home. It takes thousands of CPUs to train the many layers of these networks. This requires some serious computing power.

In fact, a lot of progress is being made using the raw horse power of Graphics Processing Units (GPUs), the super fast chips that power graphics engines in the very same arcade games.

Second, lots of data. The deep neural network plays the arcade game millions of times.

Third, a couple of nifty tricks for speeding up the learning such as training a collection of networks rather than a single one. Think the wisdom of crowds.

What will deep learning be good for?

Despite all the excitement though about deep learning technologies there are some limitations over what it can do.

DeepMind co-founder Demis Hassaabis on the potential of Artificial intelligence to solve some of biggest problems that humanity faces.

Deep learning appears to be good for low level tasks that we do without much thinking. Recognising a cat in a picture, understanding some speech on the phone or playing an arcade game like an expert.

These are all tasks we have “compiled” down into our own marvellous neural networks.

Cutting through the hype, it’s much less clear if deep learning will be so good at high level reasoning. This includes proving difficult mathematical theorems, optimising a complex supply chain or scheduling all the planes in an airline.

Where next for deep learning?

Deep learning is sure to turn up in a browser or smartphone near you before too long. We will see products such as a super smart Siri that simplifies your life by predicting your next desire.

But I suspect there will eventually be a deep learning backlash in a few years time when we run into the limitations of this technology. Especially if more deep learning start-ups sell for hundreds of millions of dollars. It will be hard to meet the expectations that all these dollars entail.

Nevertheless, deep learning looks set to be another piece of the AI jigsaw. Putting these and other pieces together will see much of what we humans do replicated by computers.

If you want to hear more about the future of AI, I invite you to the Next Big Thing Summit in Melbourne on April 21, 2015. This is part of the two-day CONNECT conference taking place in the Victorian capital.

Along with AI experts such as Sebastian Thrun and Rodney Brooks, I will be trying to predict where all of this is taking us.

And if you’re feeling nostalgic and want to try your hand out at one of these games, go to Google Images and search for “atari breakout” (or follow this link). You’ll get a browser version of the Atari classic to play.

A web browser version of Atari’s breakout found in Google images search. Google Images

And once you’re an expert at Breakout, you might want to head to Atari’s arcade website.