by Tom Standage
*Three-month moving average
What is augmented reality?
Most people, by now, have heard of virtual reality (VR). Giant technology companies, from Google to Samsung to Sony, are hoping that it will be the next big hit in consumer electronics, though it has failed to break out of a specialist niche. Its close cousin, augmented reality (AR), is less well known. Yet many people think that AR, when it comes, could have a much bigger impact than VR ever will. What exactly is it?
The first thing to realise is that “reality” means two almost entirely different things depending on which technology you are talking about. VR aims to generate an immersive artificial reality: a convincing computer simulation of the world for its users to explore. AR, on the other hand, sticks with “real” reality, and uses computers to layer useful or interesting information on top of it. That is not a new idea. AR’s early ancestors include the heads-up displays that were fitted to jet fighters starting in the 1950s, projecting information about airspeed, heading and the like directly onto the cockpit glass. Many people with smartphones will have had experience of more advanced versions. Snapchat, a messaging app, is famous for its ability to doctor photos of faces to give people rabbit ears, baseball caps, improbable moustaches and so on. Pokemon Go, a popular smartphone game, uses AR to superimpose virtual creatures onto the real world. Users of Google’s Translate app can point their phones at street signs and menus written in foreign languages, and see the text magically translated into their native tongue.
But AR’s proponents want to go much further than that. Their goal is to develop “smart glasses” that can project three-dimensional images in the user’s field of vision that appear to blend perfectly into the real world. For now, the firm that has made the most progress is Microsoft. Its HoloLens headset is a self-contained computer that uses a suite of sensors to build a 3D model of the world around it. It can then do everything from placing a set of virtual “Minecraft” blocks onto a kitchen table to generating virtual cadavers for anatomy students to study. Other companies are interested, too. Magic Leap, a startup based in Florida, has attracted $2.3bn in investment to develop a similar technology. Facebook, which bought Oculus, a VR company, for $2bn in 2014, says its ultimate goal is to produce a set of glasses that can do both VR and AR at the same time.
For now, that is a long way off. The HoloLens is impressive, but it is just an early incarnation of the technology, which can be expected to improve rapidly in the coming years, just as the brick-like mobile phones of the 1980s evolved into modern smartphones. And for AR to take off as a consumer technology, inventors will need to solve more than just technical problems. Social factors matter, too. What is the right etiquette if you are talking to someone and a text message pops into your field of vision? Will wearing smart glasses at the dining table, or in a meeting, be considered impolite? Most technology analysts think AR will make its first inroads in the workplace, where social mores are less important. Smart glasses can help a technician identify a component that needs to be replaced, for example, or give a surgeon the illusion of being able to see inside a patient during an operation. VR lets you escape into a different reality. But because AR is used in the real world, it could have many more benefits – and is also likely to have unexpected social consequences.
Why we’re still waiting for the space elevator
For decades, engineers and science-fiction writers have dreamed of lifts capable of carrying things into orbit from the Earth’s surface. Konstantin Tsiolkovsky, a Russian scientist, suggested the idea in 1895, inspired by the Eiffel Tower. And in 1979 Arthur C. Clarke wrote an entire novel, The Fountains of Paradise, about the construction of such a space elevator. Thanks to SpaceX and other private spaceflight companies, rocket launches have fallen in price in recent years. Each launch of the Falcon Heavy, the most powerful rocket in operation today, costs around $90m. But whisking satellites, space probes and even people into orbit on a giant elevator might be cheaper, more reliable and more civilised than using giant fireworks – if one could be built. Unfortunately, the technical challenges are formidable.
The basic idea of a space elevator is to run a fixed cable from a point on the Earth’s equator to a space station directly overhead, in geostationary orbit (that is, at an altitude of 36,000km). Objects at that altitude circle the planet once a day, so they have the useful characteristic of appearing to hover over a fixed spot on the Earth’s surface. Cargo-carrying vehicles can then be run up and down the cable. They need to be powered on the way up, but can reclaim energy as gravity helps them on the way down. These vehicles would have to be quite large to carry people: even if they moved at 500kph, the trip in each direction would take three days. And building a 36,000km-long high-speed railway on Earth would be hard enough. Building a vertical one into space would be much more difficult.
The chief obstacle is that no known material has the necessary combination of lightness and strength needed for the cable, which has to be able to support its own weight. Carbon nanotubes are often touted as a possibility, but they have only about a tenth of the necessary strength-to-weight ratio and cannot be made into filaments more than a few centimetres long, let alone thousands of kilometres. Diamond nanothreads, another exotic form of carbon, might be stronger, but their properties are still poorly understood. Even if a suitable material could be found, the part of the cable within the atmosphere would be subject to weather disturbances, and the vehicles running up and down it could also cause dangerous oscillations. Anchoring it to a moveable, seagoing platform might help, but keeping the cable steady would still be a tall order. A further worry is collisions: there are thousands of satellites and other items in orbit around the Earth, from an altitude of around 2,000km upwards. Any impact with the cable could cause disaster.
True believers in space elevators continue to look for ways around these problems, but they may be insurmountable. The idea refuses to die, however, possibly because of its elegance and simplicity. Perhaps the dream will be realised, just not on Earth. Building a space elevator between the moon’s surface and lunar orbit (to transport things such as visiting tourists or material mined on the moon) would be far easier, because of the weaker gravity and lack of atmosphere. Anyone hoping to take a space elevator into orbit from Earth, however, faces a long wait.
How astronomers spotted the first interstellar asteroid
On October 19th 2017 Rob Weryk of the University of Hawaii saw something rather strange. In pictures produced by Pan-STARRS 1, a telescope on Haleakala, he identified an unusually fast-moving, faint object that he concluded could not have originated in Earth’s solar system. It was travelling at more than 25km per second. That is too fast for it to have a closed, elliptical orbit around the Sun. Nor could its velocity have been the result of the extra gravitational kick provided by an encounter with a planet, since the object arrived from well above the ecliptic plane near which all the Sun’s planets orbit. Indeed, after swinging around the Sun, it passed about 25m km below Earth, before speeding back above the ecliptic plane. Observations from other telescopes confirmed that Dr Weryk’s object was the first extrasolar object to be spied by astronomers within our own solar system.
The object was originally classified as a comet and thus named C/2017 U1 (the “C” stands for comet). But it lacked the tail of gas and dust produced when such icy rocks fly close to the Sun. Furthermore, an analysis of the sunlight it reflected suggested that its surface was reddish, and was mostly rock. So it was first reclassified as an asteroid, A/2017 U1. Then, once its interstellar origin had been confirmed, it was renamed 1I/2017 U1. It was also given a proper name: ‘Oumuamua, from a Hawaiian word meaning “scout”. Measurements of its brightness suggest that it is a cigar-shaped object, about 230 metres long and 35 metres wide, tumbling end over end. Its rocky nature is puzzling. Comets are formed on the cold periphery of distant solar systems. Asteroids reside within such systems’ interiors, where any comet-like volatiles will have been driven off by the heat of
their parent stars. Models of planet formation suggest that interstellar objects are more likely to be comets, as they can be more easily dislodged from their orbits than asteroids.
One possible explanation is that over many millennia, as 1I/2017 U1 travelled between the stars, cosmic rays might have transformed the icy, volatile chemicals that would be expected to stream off a comet into more stable compounds. Another is that the Sun is not the first star 1I/2017 U1 has chanced upon, and its volatile materials have been boiled off by previous stellar encounters. Or it could indeed be that the object was rocky to begin with – perhaps once orbiting its parent star in an equivalent of our solar system’s asteroid belt, before its ejection by an encounter with a Jupiter-like planet.
Why, then, has nothing like 1I/2017 U1 been seen before? Those planet-formation theories suggest such objects should be a reasonably common sight. Perhaps the theories are wrong. Or perhaps these interstellar visitors have been overlooked in the past, and 1I/2017 U1 will now inspire a spate of such sightings in future. Sadly for astronomers, it may not be visible long enough for these questions to be resolved decisively. It is now charging out of the solar system towards the constellation of Pegasus – at 44km per second. Small uncertainties in the calculation of its trajectory may mean that where exactly it came from and where it is heading will remain a mystery. But of its interstellar origin there is no doubt.
Why drones could pose a greater risk to aircraft than birds
The “Miracle on the Hudson” – the successful ditching of a US Airways jetliner into New York’s Hudson River in 2009 after it hit a flock of geese – taught frequent flyers two things. First, it really is possible to land an aircraft on water, just as is shown on seat-back safety cards (at least for a small, narrow-body jet). Second, and more worryingly, the incident showed how dangerous birds can be to aircraft, particularly when they get sucked into engines. Aircraft engines are supposed to be designed to withstand an impact by the feathered creatures. Using big guns, chickens have been fired at aircraft engines in safety tests since the 1950s. More recently, that has prompted another question. If birds can be so dangerous, what about drones?
New research suggests that small unmanned aerial vehicles (UAVs) might actually do more damage than birds at the same impact speed, even if they are a similar weight. The study, published by the Alliance for System Safety of Unmanned Aerial Systems in conjunction with Research Excellence, a think-tank, used computer simulations to examine the impact of bird and drone collisions with planes in more than 180 scenarios. The researchers found that the drones’ rigid and dense materials – such as metal, plastic and lithium batteries – can put planes at much greater risk than the relatively squishy body of a bird. Kiran D’Souza, one of the authors, says that in every collision scenario with a drone there was at least minor damage to the plane – and sometimes it was much more severe. In one case, the researchers discovered that if a drone were to hit an aircraft engine’s fan blades when it is operating at its highest speed, the blades could shatter, causing the engine to fail.
These findings paint a grim picture, given that in the past two years the number of drone sightings by pilots has surged. According to the Federal Aviation Administration (FAA), there are around 100 cases each month of drones potentially endangering an aircraft, and two collisions have already happened in North America. In September 2017 a drone collided with a helicopter near Staten Island in New York, and the following month an aircraft was struck by a drone in Québec City. Both aircraft landed safely, but given the regulations in place, neither UAV should have been flying in the first place. In America drones are required to stay in sight of their pilot, which was not the case with the helicopter collision. They also must fly at or below 400 feet and yield to manned aircraft. In Canada UAVs are not allowed to operate above 300 feet, and airports, helipads and seaplane bases are considered “No Drone Zones”. Europe has also seen at least three aircraft collisions with drones since 2010, and has been conducting research to examine the dangers.
With so many drone-owners flouting the rules, some experts think that it is just luck that a serious collision has not yet happened. America’s FAA says it is working to create new regulations to reduce the risks drones pose to planes. New safety standards for engines may be needed to make them more resistant to impacts with the flying contraptions. In addition, existing regulations on where drones can and cannot fly should be more strictly policed. The FAA is also looking to UAV-makers and users to develop detect-and-avoid technology to prevent collisions with each other, and with manned aircraft. Given the millions of travellers who take to the sky each day, efforts to keep them safe will surely have to come from both sides.
What is the point of spam e-mail?
According to internet folklore, the very first spam e-mail was sent in 1978, to around 400 recipients. The sender was given a ticking-off, and told not to do it again. Alas for that golden age. These days, a torrent of e-mails littered with misspellings promising to cure wrinkles, enlarge penises, banish fat or wire millions in unclaimed offshore wealth is the fate of almost everyone with an e-mail address. Other e-mails aim to harvest usernames and passwords, or contain deceptive links to malicious software designed to take over a user’s computer. According to one estimate from SecureList, a cyber-security firm, roughly 60% of all e-mail is spam. But why? What is the point of the avalanche of spam?
In a word, money. Spam is the digital cousin of ordinary, paper-based junk mail. Firms send this out because they think it will drum up business. By reducing the cost of communication, the internet turbocharges that business model. Real-world junk mail might be profitable if only one recipient in a thousand decides she needs double-glazed windows or a greasy pizza. But sending an e-mail is far cheaper than sending a piece of paper. With a list of address and the right software, thousands of messages can be sent every hour. And because internet users do not pay by the message, the marginal cost per message is essentially zero. All this means that even if only one recipient in a million is conned into buying some dubious pills or clicking a link that reveals their credit-card details, the revenues far outweigh the costs.
The relative anonymity offered by the internet also enables spammers to hide their identities, which allows more obviously criminal uses of e-mail. Phishing e-mails, which try to persuade users to enter sensitive details such as banking passwords into fake (but convincing-looking) websites, can be very profitable, because the data they harvest can allow their controllers to loot bank accounts or go on buying sprees with stolen credit-card information. Malicious attachments can subvert a user’s machine, perhaps recruiting it into a “botnet”, a horde of compromised machines that can be rented out to attackers to knock websites offline. And then there is “ransomware”, in which a malicious program encrypts all the files on the victim’s computer, then displays instructions demanding payment to unscramble them. All this is made possible by giant lists of e-mail addresses that are bought, sold and swapped between spammers. Those, in turn, are generated from leaks, hacks, guesswork and addresses collected from users of shady websites and subsequently sold on.
Busts are not unheard of (a big Nigerian spammer, believed to be behind thousands of online scams earning more than $60m, was arrested in August 2016). But they are not common enough to put a meaningful dent in the business. Instead, computer firms such as Microsoft and Google have become locked in an arms race with the spammers. Spam filters began appearing in the 1990s, as the internet gained mainstream popularity. Spammers altered their tactics to work around them (which is why spam is full of deliberate misspellings such as “v1agr*”). For now, tech firms have the advantage: artificial-intelligence filters can be trained to recognise the characteristics of spam messages, and reroute them to spam folders. Training those filters requires them to have plenty of recent examples to practise on. With spam, at least, that is not a problem.
Why the police should wear body cameras
Grainy footage of police officers shooting me
mbers of the public has become unhappily familiar in recent years. Smartphones, which have proliferated, enable anyone to record police actions. The footage of the death of Keith Lamont Scott, which prompted violent protests in North Carolina in September 2016, was striking for another reason – it came from the police. It is increasingly common for police officers to sport a camera on their uniforms. A growing body of evidence suggests that the gadgets improve the behaviour both of cops and those they deal with.
A study published in 2016 by researchers at the University of Cambridge and RAND Europe, a think-tank, suggested that body cameras can slash the number of complaints made about the police. Over the course of a year, around 2,000 officers in two forces in America and four in Britain were randomly assigned cameras according to their shift. Compared with the previous year, the number of complaints brought against them dropped by 93%. The number of complaints also fell when officers were not wearing cameras during the trial, an effect the authors call “contagious accountability”. According to Barak Ariel, one of the researchers on the Cambridge study, officers who wore cameras but only started recording in the middle of their interactions with the public were more likely to use force than those not using them. So for the best results, police officers should have little or no discretion in when to turn the cameras on or off.
Civil-liberty campaigners welcome the chance to keep an eye on the police. Many police forces are enthusiastic too. Dealing with complaints is expensive. Cameras also improve the behaviour of members of the public and reduce the number of bogus complaints brought against the police. They are an efficient way to collect evidence. They can be used in training; officers can learn from their colleagues’ actions. Mr Ariel reckons British cops are more open to the devices than their American counterparts. Police unions in Boston and Cincinnati say they should not be rolled out until their contracts are changed to reflect the new work that cameras will demand.
