TechnologyQuarterly

Robots take up sewing

Making betterbatteries

Pioneering small satellites

May 30th 2015

Flying into the futureFlying into the futureInside tomorrow’s airliners

20150530_TQ_MAY.indd 1 19/05/2015 12:29

HUMAN hands are extremely good atmaking clothes. While many manu-

facturing processes have been automated,stitching together garments remains a jobfor millions ofpeople around the world.As with most labour-intensive tasks, muchof the workhas migrated to low-wagecountries, especially in Asia. Factoryconditions can be gruelling. As nationsdevelop and wages rise, the trade moveson to the next cheapest location: fromChina, to Bangladesh and, now that it isopening up, Myanmar. Could that migra-tion be about to end with the develop-ment ofa robotic sewing machine?

There have been many attempts toautomate sewing. Some processes cannow be carried out autonomously: thecutting offabric, for instance, and some-times sewing buttons or pockets. But it isdevilishly difficult to make a machine inwhich fabric goes in one end and finishedgarments, such as jeans and T-shirts, comeout the other. The particularly tricky bit isstitching two pieces ofmaterial together.This involves aligning the material cor-rectly to the sewing head, feeding itthrough and constantly adjusting thefabric to prevent it slipping and buckling,while all the time keeping the stitches neatand the thread at the right tension. Nimblefingers invariably prove better at this thancogs, wheels and servo motors.

“The distortion of the fabric is no lon-ger an issue. That’s what prevented auto-matic sewing in the past,” says Steve Dick-

erson, the founder ofSoftWearAutomation, a textile-equipment manu-facturer based in Atlanta, where Dr Dick-erson was a professor at the Georgia In-stitute ofTechnology.

The company is developing machineswhich tackle the problems ofautomatedsewing in a number ofways. They usecameras linked to a computer to track thestitching. Researchers have tried usingmachine vision before, for instance byhaving cameras detect the edge ofa pieceoffabric to workout where to stitch.

The Atlanta team, however, have great-ly increased accuracy by using high-speedphotography to capture up to 1,000 framesper second. These images are then manip-ulated by software to produce a higherlevel ofcontrast. This more vivid imageallows the computer to pickout individualthreads in the fabric. Instead ofmeasuringthe fabric the robotic sewing machinecounts the number of threads to deter-mine the stitching position. As a conse-quence, any distortion to the fabric madeby each punch of the needle can be mea-sured extremely accurately. These mea-surements also allow the “feed dog”,which gently pulls fabric through themachine, to make constant tiny adjust-ments to keep things smooth and even.

Dr Dickerson patented the idea in 2012and won a $1.3m research contract fromDARPA, a US Department ofDefenceresearch-and-development agency. Themilitary interest in sewing arises from a

Made to measure

Factory automation: A robotic sewing machine could throw garment workersin low-cost countries out of a job

Monitor

1 Robotic sewing machines thatcould replace garment workers,security systems that replacepasswords, preventing invasivespecies entering theMediterranean through theSuez Canal and generatingelectricity from glass

Difference engine

5 Your phone on steroidsThe coming 5G networks forultra-fast connections

Air travel

6 Flying into the futureThe passenger cabin is changingin tomorrow’s airliners

Air travel

9 Watching the world go bySupersonic flight is returning—in reduced form

Rational consumer

10 Ready to runcibleA guide to smartphones that aresomething different

Better batteries

11 Charge of the lithium brigadeHow energy storage is improving,steadily but surely

Brain scan

13 Britain’s spacemanSir Martin Sweeting, a pioneerof small satellites

The Economist Technology Quarterly May 30th 2015 Monitor 1

1

Contents

On the coverFrom seating to lighting andentertainment, things will bedifferent for whatever classyou fly. The cover image, byFactorydesign, a Londondesign agency, presents oneidea of the interior of a futureairliner, page 6

2 Monitor The Economist Technology Quarterly May 30th 2015

2

1

1941requirement that the departmentgives preference to American supplierswhen buying uniforms. Bill Lockhart, aSoftWear Automation executive, says lastyear the project began successfully stitch-ing together pieces of fabric roboticallyand that one machine being developed isnow able to stitch a perfect circle—some-thing that only a highly skilled humanoperator would dare to attempt.

But neat stitching is only part of thesewing process. Most garments are as-sembled from various different pieces offabric: 20 or more sections typically for apair of jeans. This means a robotic sewingmachine also needs to be able to pick upmaterial, feed the appropriate sections tothe sewing head and remove them whencomplete. To do that, the company hasdeveloped a materials-handling systemwhich it calls LOWRY. This uses a vacuumgrip to pickup pieces of fabric and movethem to another machine, which mightcut, stitch, add buttons or carry out otherfinishing tasks. LOWRY is programmable,so can switch easily from working withone size ofmaterial to another.

A stitch in timeAlthough the first LOWRY will be deliv-ered to an American factory later this year,commercial versions of the firm’s roboticsewing machines capable ofautomatingthe more difficult tasks in making gar-ments will not be ready until next year.

FrankHenderson, the boss ofHen-derson Sewing, an Alabama-based textile-equipment firm, and an investor in Soft-Wear Automation, reckons that roboticsewing will be attractive to Americanfashion brands wanting to bring produc-tion closer to home and produce garmentsrapidly to catch new trends. Inditex, a “fastfashion” Spanish company, whose brandsinclude Zara, famously does this by mak-ing many of its garments manually inEurope to speed up its time-to-market.With designs and samples shuttling to andfrom Asian factories, it can take monthsbefore new clothes finally turn up in otherAmerican and European stores—by whichtime they can be out of fashion.

JackPlunkett, ofPlunkett Research, amarket-research company, says pressureon Asian clothing manufacturers to keepwages low while improving workingconditions is leading many to lookatautomation, too. On that point, SoftWearAutomation’s Mr Lockhart says a Bangla-deshi company has already expressedinterest in the firm’s technology.

But it is not only the production ofgarments which could be increasinglyautomated. Nike is now using a tech-nology called Flyknit to make some of itstrainers. Flyknit uses a computer-con-trolled knitting machine to automaticallyweave strands ofpolyester yarn into theshape of the upper part ofa shoe, instead

ofhaving it manually stitched togetherfrom individual panels, the way mosttrainers are made in Asian factories. Nikehas not said how far it intends to take thetechnology, but it has the potential toproduce customised trainers for individ-uals and to do so locally—perhaps evenwithin stores.

Shoemakers are already using 3D

printers, which build up material additive-ly, to make prototypes ofshoes. Exoticclothing and shoes made with 3D printersare becoming regulars on the catwalks atmany of the world’s leading fashionshows, although the materials they areprinted from tend to be various sorts ofplastic, which can make the garmentssomewhat clunky and shoes a bit clog-like. However, researchers are working onways to print more flexible materials. Onesuch project involves a collaborationbetween Disney, Cornell University andCarnegie Mellon University. Their 3D

printer uses layers ofoff-the-shelf fabric tomake soft objects, such as cuddly toys.

The real test ofhow successful robotswill be at making clothing and shoes willdepend on how efficient and reliable theywill be, and how fully they can automatethe process. If time-to-market and custo-misation are priorities, then the robotsmight win—even if some manual inter-vention in production is required. But formass-produced lines, where every centmatters, any human involvement couldkeep manufacturing offshore. The lessonfrom industrial automation in other sortsof factories, though, shows that robotskeep getting better and cheaper. It may bea while coming, but the writing seems tobe on the wall for sweatshops. 7

PASSWORDS are the bane ofa lifeenveloped by information technology.

And they are proliferating wildly as moreand more devices and online servicesdemand them. But passwords are easilylost, forgotten and stolen. Though cheerfulreports of the password’s imminent de-mise have been made before (perhapsmost famously in 2004 by Bill Gates whenhe was the boss ofMicrosoft), all haveturned out to be premature. However,new ways are emerging to remove, or atleast reduce, the reliance on tapping outawkward combinations ofnumbers,

letters and symbols.The idea is to make greater use of a

process called multi-factor authentication.The password is, as security folkput it,“something you know”. It can be mademore robust by being paired with “some-thing you have”, which could be a deviceor app which receives or generates a un-ique code, known as a token, with a shortperiod ofvalidity. Such gadgets are alreadywidely available in online banking forusers to generate a code when accessingtheir account. A code can also be texted toa user’s mobile phone when logging intotheir e-mail on a computer.

The token (which might be exchangedwirelessly between, say, a phone and apayment terminal) could be paired with“something you are”, such as an analysisofa user’s voice, his iris or a fingerprint. Asmore smartphones and computers nowcome with biometric systems which cantake such measurements, it is becomingeasier to link the something-you-havewith the something-you-are, thus leavingthe password out of the process.

To enable the wireless transfer of asecurity token, devices are being fittedwith near-field communication (NFC)chips. As these chips use less power thanBluetooth, NFC works over a shorterrange: just a few centimetres. Devices needto be close or touching to communicate,which is ideal for contactless smartphonepayments, such as Google Wallet andApple Pay. During the transaction the usercan confirm his identity by holding afinger over the phone’s fingerprint reader.

Other sensors in the phone can beenlisted to help confirm identity duringmobile payments. One idea is to tap into aphone’s GPS system, a bit like the waybanks use location data to spot paymentsand cash withdrawals being made inplaces that fall outside a customer’s typ-

Passé words

Cyber-security: Technology is finallymaking strides in reducing the needto remember lots of complicatedp@s$w0rd$

The Economist Technology Quarterly May 30th 2015 Monitor 3

2

1

ical spending behaviour. And a profile canbe built up of the phone itself to ensure itis the device it purports to be and not onethat has been cloned.

A consortium offirms called the FastIdentity Online Alliance (FIDO) is lookingat various approaches to strengthen multi-factor authentication. It has built a rosterof impressive members, including hard-ware firms such as Lenovo, Samsung andARM, which makes many of the chipsused in mobile devices; payments firmsPayPal, Visa and MasterCard; and e-com-merce giants such as Alibaba.

FIDO has two approaches, one ofwhich offers the ability to do away withpasswords altogether. Both rely on public-key cryptography, which is a bit ofappliedmathematics that already underpins a lotof the web’s security infrastructure. Thisform ofencryption creates two inter-twined cryptographic keys—vast strings ofnumbers—one public and safe to shareand one that is private.

Who goes there?Any person or website in possession ofthe public key can verify that a digitalsignature has come from someone inpossession of the private key, but cannotguess what that key is. Similarly, usingonly the public key, a person or service canencrypt a document in a way that only thepossessor of the private key, which isassociated to a specific device, can use.

The idea is that a device would, onrequest, create a pair ofkeys, public andprivate. It would pass the public key to, forexample, an online retailer, which thenassociates the key with a user’s account.On the next login, the identity checkworks both ways: a faked version of theretailer’s website will not have the publickey the FIDO software is looking for, so theuser would be alerted that something isamiss. And the retailer can check that thelogin has been performed by the holder ofthe associated private key.

Permitting access to FIDO via a device’sbuilt-in security features, such as a finger-print reader or other biometric scan, addsthe something-you-are to the FIDO key’ssomething-you-have. So a password is nolonger required. Nevertheless, securitypeople are cautious, so provision is beingmade for an optional PIN number ifneed-ed, but as an auxiliary measure.

FIDO’s other standard, Universal 2nd-Factor Authentication (U2F), demotes thepassword rather than replaces it. U2F

would be incorporated into a small physi-cal token, like a dongle which could beplugged into a computer’s USB socket orlinked to it with NFC. Yubico, a Califor-nian company, already makes versions ofthese. On sites that support it, only a sim-ple PIN might be needed as a first step toentry. While that PIN might be guessed ordisclosed, security is beefed up by a sec-

ond step of tapping a button on the device,or tapping it against a smartphone, totrigger cryptographic keys.

This sort of thing has been tried before,but unseating the password only to re-place it with devices, apps and otherprocedures risks making things just astroublesome. And FIDO is not the onlydog barking up the tree ofsafer solutions.Yahoo recently introduced another way toavoid remembering passwords: Americanusers can now set their accounts so thatwhen they go to log in a one-time on-demand password is texted to their phone.

Some passwords and PINS may not,then, disappear completely. They mightstill be required on occasions to, say, up-date a person’s account details. But multi-factor authentication will at least helpdeter hackers and thieves from breakinginto phones, computers and other devicesby making that more difficult, says MarkusJakobsson, an American researcher whostudies how computer systems are used.Having to get hold of—or trying to clone—apiece ofhardware and copying someone’sfingerprints are complications that shouldkeep most users safer.7

INVASIVE species are a menace. In theMediterranean, poisonous pufferfish

frequently turn up in fishermen’s nets.Summer swarms ofstinging jellyfish closebeaches as far west as Italy. And in the east,rabbitfish are eating most of the algae thatnative species rely on. Along with some350 other non-indigenous species ofma-rine life, all three migrants arrived fromthe Red Sea via the Suez Canal.

Marine experts are concerned thateven more invasive creatures will turn upin the Mediterranean now that Egypt hasstarted workon doubling the capacity ofthe Suez Canal with enlarged, deepenedand new waterways. Egyptian officialshave said the workwill not add to theproblems and are due to present an envi-ronmental-impact assessment to theEuropean Commission. But Europe’sadministrators have been lackadaisicalabout seeking controls to stem the in-vasion, says Argyro Zenetos of the Greekgovernment’s Hellenic Centre for Marine

Research in Athens. What, in any case,could Egypt do about it?

There are a number of technologiesemployed to deter aquatic migration.America’s Mississippi River basin is infest-ed with bighead and silver carp importedfrom China in the 1970s to stockponds.�

oracious eaters, they deprive nativespecies of food. To keep them from gettinginto the Great Lakes through the ChicagoSanitary and Shipping Canal, stretches ofit are electrified. The voltage trails offfroma peakofabout 2,000 volts in the centre ofsome electrified sections to zero about 22metres either side. This gradient creates azone of increasing discomfort sufficient tocause fish to turn around.

Electrified systems are used in otherparts of the world and to keep fish out ofthe intake pipes at power plants. Smith-Root, an American firm, has electrified aspot in the Puntledge River near Courte-nay, British Columbia, to discourage sealsfrom decimating the juvenile salmon that

No way for fish

Marine controls: How to prevent Egypt’s enlargement of the Suez Canalmaking it easier for Red Sea creatures to colonise the Mediterranean

A charged view

Solar power: Using glass to makeelectricity out of sunlight could turnbuildings and cars into their ownpower stations

solar panel.Last year a team at Michigan State

University led by Richard Lunt, formerlyat MIT, displayed a variation of the ap-proach using extremely small organicmolecules, which Dr Lunt describes as“exceptionally transparent to the humaneye”. These molecules absorb specificnon-visible wavelengths of light and thenglow at a different IR wavelength. Thisglowing light, which is also invisible to theeye, is guided to the edge of the glasswhere it is converted to electrical energyby thin strips ofphotovoltaic cells. Thearrangement, known as a transparentluminescent solar concentrator, allowsmost of the glass to be kept clear of solarcomponents. The first version had a pow-er efficiency ofonly about1%. But it is earlydays and the researchers hope to boostthat considerably.

Rolling it outIt should be possible by using materialsthat absorb non-visible wavelengths oflight to produce thin-films ofsolar cellscheaply using industrial processes thatmake large rolls, says Rutger Schlatmann,director of the Competence Centre forThin-Film and Nanotechnology for Photo-voltaics, a Berlin-based industry researchgroup. But as he points out: “It is visiblelight that carries by far most of the energy.”That means, however good they are, trans-parent solar cells may never rival solarpanels designed to capture the maximumamount of light. Nevertheless, what theycan trap could still be useful.

Semi-transparent solar cells can beused to produce coloured or tinted glass,which helps when shading is required.Heliatek, a company based in Dresden,Germany, uses organic materials to makesolar-cell films which are up to 40% trans-

parent. With a solar efficiency ofover 7% they can produce elec-tricity-generating tinted glassin buildings and car sunroofs.

One development that isattracting a lot of interest is the

use ofa family ofcrystalline mate-rials called perovskites, which could

allow semi-transparent solar cells tobe made relatively cheaply in large

rolls. A group at Brown University inProvidence, Rhode Island, recently re-

ported they had made ultra-thin filmswith perovskite crystals that are capableofa solar efficiency ofover15%.

Oxford Photovoltaics, spun out of theUniversity ofOxford in 2010 to commer-cialise thin-film solar cells, reckons perov-skites are good for over 20%. The firmcalculates that ifa 35-storey office block inLondon was clad with perovskite cellsthey could generate almost 60% of thebuilding’s energy consumption. Whenelectricity bills are high or batteries arerunning low, every bit of juice counts.7

WITH so much glass in buildings, carsand the screens ofmobile devices, it

is understandable why researchers wouldlike to come up with transparent solarcells which could generate electricity andtop up batteries.

Solar cells workby absorbing the pho-tons in sunlight and converting them intoelectrons, which are gathered by elec-trodes to flow into a circuit. Most solarcells are opaque to absorb all the light theycan to maximise their efficiency. So, tolookout ofa window or use the screen ofa smartphone, a layer ofsolar cells has tolet some light throug���et the more trans-

parent the cells, the less energy they pro-duce—or at least that is how it works withtraditional solar technology based onsemiconducting materials such as silicon.

An alternative is to make solar cellsfrom substances that absorb light only atwavelengths which are invisible to thehuman eye, such as those in the infrared(IR) and ultraviolet (UV) spectrum. Thatwould allow visible light to pass through.One company working on this is Ubiqui-tous Energy, a spin-offfrom the Massachu-setts Institute ofTechnology (MIT) in 2011.It is developing solar cells using transpar-ent organic materials that absorb IR andUV wavelengths.

Taking light from only part of the spec-trum would reduce the percentage ofsunlight’s energy that can be convertedinto electricity. Ubiquitous Energy ishoping, some thinkoptimistically, toexceed 10%. That compares with 20-25%efficiency for a typical non-transparent

4 Monitor The Economist Technology Quarterly May 30th 2015

2 congregate there (bigger animals feel thejolts more than small fish).

Another method is to create a curtainofbubbles. Produits Étang.ca, a Quebecmanufacturer, pumps high-pressure airthrough submerged plastic pipes that havebeen drilled with microscopic holes. Thisthrows up a dense, turbulent curtain ofbubbles. Most fish are spooked and jel-lyfish avoid bubbles lest they fill theirumbrella. A three-pipe bubble curtain foran enlarged Suez Canal would cost lessthan $1m, air compressors included, reck-ons Mario Paris, the firm’s chiefexecutive.

But bubbles do not stop all species. Tomake them scarier, Ovivo, another Que-bec firm, illuminates bubbles with flash-ing bright lights and installs underwaterspeakers to produce loud noises. Althoughit seems a bit like an aquatic disco theconstantly changing lighting and soundsequences are scientifically calculated tobe as obnoxious as possible to variousaquatic species. The combo works wellenough to mostly keep Chinooksalmon, aspecies not typically afraid ofbubbles, outofa pumping station that draws waterfrom the Sacramento–San Joaquin RiverDelta in Tracy, California.

The Panama Canal, which was built inthe early1900s, uses locks to lift vessels toGatun Lake, an artificial expanse ofwaterthat was created to reduce the amount ofdigging. Being a freshwater lake it providesa barrier which prevents marine speciesfrom moving between the Pacific Oceanand the Caribbean Sea. Ifa competingcanal is ever completed across Nicaragua,as the government intends, a natural lakewill also be used as a freshwater barrieragainst migrating species.

The Suez Canal does not use locks asthe sea levels at either end are similar.However, few organisms survived passagefor about 80 years after it opened in 1869.This was due to a natural barrier, the so-called Bitter Lakes, a former salt pan. Thelakes, though, no longer draw enoughextra salinity from the ground to hinderthe passage ofsea creatures.

This leads some scientists to proposerecreating a highly saline area. Noa Shen-kar, a zoologist at Tel Aviv University, whoorganised an international workshop lastyear to discuss such ideas, says pumpingbrine from a desalination plant couldcreate such a stretch. The typical salinityof the Red Sea is around 40 parts-per-thousand (ie, every kilogram ofwatercontains 40 grams ofsalt). If some of theSuez Canal was raised to just 44 parts,marine organisms would be repelled orkilled, she adds. Chad Hewitt, a marineecologist at the University ofWaikato inNew Zealand, suggests digging a section ofthe canal through another salt pan. Ifexperts conclude Egypt should installsome sort ofbarrier, then turning the clockbackmay be the answer.7

The Economist Technology Quarterly May 30th 2015 Difference engine 5

NEW MOBILE networks come alongonce every decade or so. Starting

around 1980, the first generation of cellularphones relied on analogue technology.When the second generation arrived in1991, wireless networks began to go digital.By 2001, the third generation dumped cir-cuit-switching, which uses direct connec-tions, for more efficient packet-switching,which moves data around in blocks.Around 2010, fourth-generation networksembraced IP (internet protocol) technol-ogy wholeheartedly, providing mobile de-vices with broadband access to the inter-net. Given past experience, the nextgeneration is due in 2020.

Mobile-phone companies are now dis-cussing what to include in their proposedfifth-generation (5G) networks. There is afeeling of urgency as heavyweights from outside the traditionalphone business, such as Google and Facebook, are moving intothis area. One thing 5G networks must do is meet the growing de-mand for ubiquitous and instantaneous connectivity. That re-quires future networks to have a “latency” (ie, response time) ofno more than a few milliseconds. The speed at which two devicescan begin to communicate with one another over today’s 4G net-works is about 50 milliseconds, and around 500 milliseconds forthe still widely used 3G services.

Another requirement is a data rate of at least one gigabit persecond (1Gbps) to start with, and multiple gigabits thereafter. To-day’s 4G networks, based on a standard called long-term evolu-tion (LTE) technology, can manage 10-100 megabit per second(mbps), depending on conditions. Most mobile carriers are stillrolling out their LTE services; only a few have started to install thelatest LTE-advanced equipment (true 4G as opposed to the half-baked versions some have been marketing as the real thing). Thepeak bit rate of LTE–advanced is claimed to be 1Gbps. In the realworld, however, it is more like 250mbps. Given the tenfold im-provement between generations in the past, an average 5G down-load speed of 1Gbps seems realistic—with the chance of up to10Gbpsas the technologyripens. This is likely to increase even fur-ther the diversity ofmobile devices on offer (see page 10).

Two are better than oneTwo technical features—carrier aggregation and MIMO (multipleinput/multiple output) antennae—are responsible for giving LTE-advanced its big boost over earlier 4G iterations. Neither tech-nique is particularly new, but both are expected to play a centralrole in helping 5G fulfil its promise.

First, carrier aggregation. This is a way of boosting downloadspeeds by plucking signals from a number of local base stations,instead of simply the most powerful one in the vicinity. These dif-ferent channels—often with different frequencies—are combinedinto what is effectively a single fat pipe capable of delivering dataat a far higher rate than would otherwise be possible.

Given the global shortage of radio spectrum, most telecomsfirms have snapped up frequencies wherever they can. As a result,few oftheir chunks ofspectrum are contiguous. Fortunately, carri-

er aggregation also allows mobile opera-tors to patch together theirdisparate blocksof spectrum. This is going to be even moreimportantwhen 5G arrivesaround 2020 ina still more crowded wireless world.

Much the same goes for MIMO. Thisworks by transmitting two or more datastreams via two or more antennae, andhaving the receiving antennae process allthe incoming signals instead of just thestrongest one. Today’s MIMO devices tendto have three or four antennae on both thetransmitting and the receiving ends. Butwhat if each end had tens of antennae oreven hundreds? That would translate intoa significant increase in download speed,and a far more efficient use of the availableradio spectrum.

Which spectrum that will be, though,has still to be decided. Today’s wireless devices operate in thecrowded 700MHz to 2.6GHz part of the radio-frequency compass.It is not as though once 5G hits the airwaves chunks of spectrumused today by 4G and even 3G networks will suddenly becomevacant. Mobile carriers will still have to continue their older ser-vices for the millions of subscribers who do not immediately up-grade to the latest devices—and may not do so for years to come.

The obvious answer is for5G to migrate from today’s very-highfrequencies (VHF) to either the super-high (SHF) band between3GHz and 30GHz, or even to the extremely high (EXF) one at 30-300GHz. Current occupants of these rarefied frequencies (alsoknown as “millimetre waves” because of the wavelength) includesatellite television, microwave relay links, air-traffic control radar,radio astronomy and amateur radio.

In most regions of the world, a chunk of spectrum around60GHz has been designated for use by the public without a li-cence. With their latest standard (802.11ad), the Wi-Fi communityplans to exploit this band for streaming ultra-high-definition vid-eo around the home. In typical configurations, it should beammore than 6Gbps overmodest distances. At least in the laboratory,Nokia, a Finnish networking firm, has achieved speeds of 115Gbpsusing such frequencies.

As always, there are drawbacks. One is that such extreme fre-quencies are easily blocked by walls and even people movingaround. They also get absorbed by the atmosphere, although theabsorption effect only becomes significant at distances greaterthan 100 metres or so. By going to 70GHz and above, atmosphericabsorption disappears, but rain can still cause the signal to fade.

All of which suggests that 5G will need base stations closer tousers than current cellular towers. As it so happens, that is alreadya trend. So far, microcells—no bigger than a Wi-Fi modem—havebeen used mainly inside buildings to overcome poor reception. Tohandle 5G’s needs, hundreds of them will be required to fill thegaps between existing base stations. With the tiny antenna boxesattached to lampposts and the sides of buildings, few people willever notice them, let alone object to their presence—as is so oftenthe case when new cellular towers are planned. So it is not justfasternetworkingspeeds thatare promised, but lessangst over thetowers that phones connect to.7

Your phone on steroids

5G networks: Proposed new networks promise ultra high-speed connections for mobile users

6 Air travel The Economist Technology Quarterly May 30th 2015

1

THE twinkling stars above the passen-gers gradually fade away as the night

sky lightens and the sun begins to rise. It isan illusion, as it has long been daylight out-side. But the projected image has a pur-pose: gently to awaken those on board andhelp their body clocks adjust to a new timezone. As the airliner begins its descent toNew York, the ceiling and walls turn trans-parent to provide a panoramic view of theManhattan skyline. And on lining up to arunway, the aircraft’s seats automaticallychange shape, becoming more upright andfirm to provide additional support for thelanding. Welcome to JFK airport, some-time in 2050.

This vision of what it will be like insidean airliner of the future comes from Air-bus. The European aerospace giant got itsengineers to look at how flying mightevolve from the passengers’ perspective.The fuselage has a “bionic” structure con-structed from composite materials whichmimic the bones of those masters of flight:birds. The composition of bone consists offibres, which are light but also immenselystrong when arranged to carry tensionwhere it is needed. The structure savesspace and also helps to reduce the aircraft’sweight, so it burns less fuel.

The upper part of the bionic cabin is

covered with what Airbus describes as a“biopolymer membrane”, a sort of toughplastic coating which can be electronicallycontrolled to turn opaque or transparenton command, thus eliminating the needfor conventional windows. This too helpsto make the fuselage light and strong.

The traditional rigid divisions into first,business and economy classes have gone.This is thanks to the transforming seats.Made from “memory” materials whichcan morph into a different shape and thenreturn to their initial form, they adapt tothe size of an individual’s body—and theirtravel budget. The more you pay, the morespace and comfort the seat will provide.This would enable airlines to configureseating according to demand. And thatmeans there will no longer be any need toupgrade passengers from cattle class if therear is overbooked.

The technological elements that couldmake this concept become reality can al-ready be seen in new aircraft and in cabindesigns that are much closer to production.How the experience of flying will changedepends, however, as much on the unfor-giving economics of air travel as on theimagination of the designers.

Producing the interior of an aircraft is acostly business—not least because in order

to provide a distinctive product most air-lines want something different, so the in-sides are custom-built. As a result carrierswill spend some $10 billion this year, up by5% from 2014, on cabin interiors for newand refurbished aircraft, estimates ICF In-ternational, a consultancy ba�����irgin-ia. Making that kind ofexpenditure pay de-pends on what flyers you have in mind.

In the battle for passengers, the price ofthe ticket is usually the most importantthing. Given a choice, air travellers alwayssay they want more room and extra com-fort, but are usually not prepared to payany more money for them—or at least not alot more. The emergence of various formsof a new airline class, often called “pre-mium economy”, is an answer. But there isa wideninggulfbetween the luxuryof firstand business classes, and the austerity en-dured by hoi polloi at the back.

The differences are most apparent inthe amount of space a passenger gets. Theseat pitch in standard economy (measuredas a point on one seat to the same point onthe seat in front) is typically between 78cm(31 inches) and 82cm. Spirit Airlines, a bud-get American carrier, has trimmed thepitch of its standard-economy seats to just71cm. At the other end of the spectrum AbuDhabi’s Etihad Airways is offering a nearly12-square-metre three-room first-class suitewith a shower called “The Residence” onits Airbus A380s. It comes complete with aprivate chef and a butler trained at the Sa-voy Hotel in London.

Turn left…It is in the business-class cabin that airlinesare spending most heavily, however, be-cause it is more profitable for them. Oneseat manufacturer reckons that this yearairlines will install in new and refurbished

Flying into the future

Airline interiors: How technology is changing the passenger cabin forwhatever class you fly

aircraft roughly 2,000 new seats in firstclass, but 50 times as many in business.

Business class has already changed a lotin recent years. It can provide the same or abetter level of comfort than that availablein the first-class cabin less than a decadeago. For a start, if a business-class seat doesnot extend into a fully flat bed, the airline’soffering is no longer “even in the game”,says Blake Emery ofBoeing, America’s big-gest aircraft manufacturer.

The result is thatbusinessseats are turn-ing into areas ofpersonal accommodation.These spaces will get more sophisticatedand specialised. Teague, a Seattle companywhich has designed interiors for all mod-els ofBoeing’s airliners, worked with Nike,sleep experts and professional coaches todesign a concept called the “athlete’splane”. It includes a training room and asleeper pod (pictured below) to enhancerelaxation. The effects of air travel on thebody’s natural rhythms means that ath-letes travelling over multiple time zonesare statistically more likely to lose against ahome team, says Teague’s Devin Liddell.

Powerful computer-aided design andsimulation provide designers with moreways to explore the clever engineering re-quired to create as much room as possiblein a confined space. Paperclip Design, aHong Kong company, has come up withthe closest thing yet to a morphing seat.Called Butterfly (also pictured below), it iscomposed ofunitsofdouble seatswith theaisle seat offset backwards. In premiumeconomy both seats would be used, butthe cabin could be upgraded quickly tobusiness by having the inboard seat leftempty to flip over as a sleeping platform.

Anotherapproach involves positioning

seats at various heights to gain space. Ja-cob-Innovations, a Massachusetts firm,has a design called StepSeat that lifts everyother seat about18cm allowing extra roomfor them to recline. A more radical arrange-ment is a series of stacked cocoons calledAir Lair (illustrated on the next page). It of-fers a third more passenger capacity, saysAdam White, head of Factorydesign, itsLondon creator. So far, airlineshave tendedto shun these so-called 3D-seating set-upsbecause they think that some passengers,for reasons of status, will not want to sit ondifferent levels.

Developing and designing a new busi-ness-class sleeper-seat or pod can take twoyears or more, and top-of-the-line modelsmight cost up to $350,000 each, once theyare stuffed with electronics. The price ispartly explained by the standard requiredto protect passengers from a seat breakingin a crash. This was raised in 2009 byAmerica’s Federal Aviation Administra-tion from withstanding a deceleration ofnine times the force of gravity (9g) to 16g.Although this mean that seats and their fit-tings have to be made stronger, the use oflightweight composite materials, such ascarbon fibre, allows thinner seats to meetwhat has now become a global standard.

Carbon fibre is already widely used tomake aircraft fuselages and wings. It isstronger than steel but lighter even thanaluminium. That strength comes from thepowerful links between carbon atoms—similar to the toughness imparted to a dia-mond. The fibres are woven together andarranged in position using detailed com-puterised stress-analysis for maximumstrength. The fibresare then embedded in ahard resin. Carbon technology is advanc-

ing rapidly, with even greater performancecharacteristics claimed for single-atom-thick layers of graphene. Such work maywell lead to the advanced compositesneeded to make bionic fuselage structures.

… turn rightNew lightweight technologies are alsohavingan impact in the backofthe aircraft.Last year Air Mediterranée, a French carri-er, removed the 220 economy seats in anAirbus A321—it bought them for about$300,000 in 2006—and replaced themwith a new, lighterversion. The skinny seat(pictured on the next page) is made by aParis startup called Expliseat and weighsjust 4.2kg (9.3lb) compared with the 12kgseat it replaced, says Air Mediterranée’sChristophe Costes.

Expliseat’s seat-frame is constructed oftitanium and carbon fibre. The “titaniumseat”, as it is named, cost Air Mediterranéenearly three times as much as each alumi-nium one it replaced, but this will be morethan compensated for by fuel savings. Ex-pliseat’s lightest seat weighs just 3.9kg. AirTahiti, which has also bought the com-pany’s seats, reckons the weight savingswill allow some of its turboprop aircraftoperating off short runways to carry 55rather than 50 passengers, says BenjaminSaada, Expliseat’s co-founder.

That hints at how the space savingsfrom thinner seats are likely to be used ineconomy: not to provide more legroombut instead to pack in an extra row or twoof seats. Some new aircraft will also havesmaller lavatory cubicles as a result ofmore petite plumbing. All this helps seat-ing“efficiency”, the industry’s euphemismfor density. Shrinking the seatback pocket

Teague’s bed for weary athletes and Paperclip Design’s transformer seats

The Economist Technology Quarterly May 30th 2015 Air travel 7

2

1

8 Air travel The Economist Technology Quarterly May 30th 2015

2

1

and placing tray stowage higheron its seatshas already helped Germany’s Lufthansaincrease seating on its Airbus A320 fleetfrom 150 to 168, saysSamuel Engel ofICF In-ternational. Tallying the additional seatspacked into all aircraft in itsfleet, Lufthansagained the equivalent capacity of 12 newA320s, he adds.

It could be worse. Some carriers haveflirted with standing room. Spring Air, aChinese budget carrier, has proposed in-stalling such a scheme to increase capacityon its aircraft by 40%. In 2012 Michael O’Le-ary, never one to miss making headlines,promoted the idea ofa padded backrest forstanding passengers on Ryanair, the Irishdiscount carrier which he runs. Aviointe-riors, an Italian seatmaker, produced adownward-sloped perch called Skyriderbut got no takers. Public reaction, says Er-manno De Vecchi, the firm’s boss, was, inessence, “sitting on a saddle crammed inlike a sardine? Forget it.” Safety certifica-tion for “standing” seats would, in anyevent, be highly improbable.

If more space in the economy cabin isunlikely, passengers on all budgets will seesome benefits from the changes to cabins.First, technologies that reduce weight andtherefore fuel consumption should help tolowerfares further. Even the padding in theseat cushions is changing with improvedmaterials. New types of foam and fire-re-sistant coverings have been developed toproduce lighter cushioning. This has result-ed in a fall in the weight of a typical seat’scushioning from 1.8kg a decade ago to lessthan halfa kilo now, according to Aviointe-riors. Mark Hiller, head of the aircraft-seat-ing division ofRecaro, a German firm, saysthat in 20 years the company has cut theoverall weight of its seats by 20% on threeseparate occasions. Each kilo of weightshed from an aircraft reduces its annualfuel bill by at least $100, he adds.

Second, new designs may help addresssome of the most common irritations ofeconomy-class travel. Having a passenger

shocks of impacting knees. The material islighter and softer on the kneecaps than therigid plastic shields which are sometimesused, adds Mr Saada. Teague’s Mr Liddellsays one area where more innovation isneeded is to make the dreaded middle seatmore inviting—especially in economy. Atthe moment, he adds, for passengers themiddle seat is “your sentence for waitingtoo long to book”.

The third improvement is likely to bethe ambience of the cabin. BAE Systems, aBritish firm, recently made its first sale of aset-up called IntelliCabin that features LED

lighting designed to induce calm, improvesleep and charm flyers with illuminationschemes like an artificial sunset for dinner,says Jared Shoemaker, head of cabin tech-nology. A Boeing team of engineers, psy-chologists and marketing people are alsoconcocting LED-lighting schemes to makecabins appear roomier—“to essentially dis-tract people from the discomfort” of a tightsqueeze, says Boeing’s Mr Emery. Accentu-atingceilingcurveswith sky-blue light gen-erates a “sense of space expanding above”

Two can be company sharing an armrest and seats that are slimming down

drop his seat backwards into another pas-senger’s space is a great cause of angst.(Hence the brisksales ofa $22 gizmo calledthe Knee Defender, now banned by someairlines, which attaches to a tray strut andprevents the seat in front from reclining.)James Lee, the boss of Hong Kong’s Paper-clip Design, has come up with a fixed-posi-tion seat in which the back cushion can belifted forward from the bottom and kept in-clined by stuffing a briefcase or jacket be-hind it. This “anti-technology” seat, addsMr Lee, has the benefit of doing away witha reclining mechanism. Expliseat’s titani-um seat also remains upright.

Elbows at warAnother annoyance is the battle over ashared armrest. But Mr Lee has an answerfor that, too. He has patented a design withtwo flat surfaces at different heights, pro-viding room for two elbows (picturedabove). Having knees pushed into passen-gers’ backs is a further frustration. Expli-seat embeds a taut sheet of a secret po-lymer into the rear cushion to absorb the

Airborne pod life

and bathing vertical surfaces in white lightcreates an illusion ofgreater width.

When Boeing’s 787 was introduced in2011 it pioneered a number of enhance-ments. The aircraft is fitted with large “elec-trochromic” windows that trap a thin gelbetween two panes of glass. An electricalcurrent is applied to darken the gel, allow-ing passengers to choose from five transpa-rency settings. The 787 also improves theair in the cabin. Airlines have long kept airdrybecause humiditycorrodesmetals. Butthat is less of a problem with the 787’slargely carbon-fibre fuselage. This meansthe air can be more moist. And becausecarbon fibre is stronger than aluminium,cabin pressure can be maintained at ahigher level, closer to what it is on theground. The 787 also pumps air into thecabin electrically rather than having it bledfrom the compressors in the jet engines,which risks fumes entering.

The goggle boxAugmented reality might help ease thejourney, too. Epson, a Japanese electronicsfirm, has created goggles, called MoverioBT-200, which project images onto a lensthat appears to be a screen five metresaway. Some of the first versions are beingtested by a South Korean airline. The gog-gles could be used to watch films, which inthe future are likely to be beamed wireless-ly around the aircraft’s cabin, eliminatingthe need for lots of bulky in-flight enter-tainment equipment.

Video screens made from thin filmscould be used on the back of skinny seatsand placed on cabin walls to provide anenhanced view of the outside and, eventu-ally, as a replacement for windows. Theview would be relayed to the screen by ex-terior cameras.

The Centre for Process Innovation, aBritish technology group, is exploring theuse of organic light-emitting diodes(OLEDs) to make the thin-film screens.OLEDS use a luminescent layer of organiccompounds to emit light in response to anelectric current. Matthew Herbert, a man-agerat the centre, says that in less than a de-cade production technology will be inex-pensive enough to mass-manufactureflexible OLED displays for use in aircraft.Using such screens instead of windowswould reduce weight and improve thestrength of the fuselage. One company,Spike Aerospace of Boston, is proposing toreplace windows with screens in a super-sonic executive jet (see box). Step by step,the vision of that New York-bound flight in2050 is moving closer to reality. 7

The Economist Technology Quarterly May 30th 2015 Air travel 9

2

SUPERSONIC travel for airline pas-sengers came to an end on October

24th 2003 when a British Airways Con-corde completed the last scheduledflight from New York to London. Withthe ability to cruise at 2,160kph(1,350mph, or around Mach 2—twice thespeed ofsound) the fastest a Concordemade it across the Atlantic was a shadeunder three hours, compared with theseven or eight hours it takes in a subson-ic airliner. A number of things did forConcorde, including heavy fuel con-sumption, its sonic boom restrictingspeed over land and a fall in passengersafter an Air France Concorde crashed inParis in 2000, killing all 109 people onboard. No replacement aircraft has everemerged. But a supersonic executive jetmay be a different matter.

One contender is the S-512, which isbeing developed by Spike Aerospace, aBoston company. Instead ofwindows itwill have a “multiplex digital cabin”—athin-screen video display either side ofthe passenger compartment that wouldbe fed with a live view taken by sixexterior cameras (a pair ofconventionalwindows would be retained). The inte-

rior effect would be dramatic although,as the company points out, for much ofthe time there might be little to see apartfrom clouds or the stars in a night-timesky. On those occasions films could beshown instead—or even a PowerPointpresentation for workaholics.

Using video windows would makethe jet lighter, quieter inside and lessexpensive to build, saysikKachoria,Spike’s chiefexecutive. The S-512 couldcarry12-18 passengers at Mach 1.6 and,unlike Concorde, would have the rangeto fly from Los Angeles to Tokyo in justsix hours. Although at an early stage inits development, Mr Kachoria says thecompany has received $5m deposits forseveral of the roughly $100m jets, whichit hopes to start delivering in 2020.

Advanced aerodynamics and itsrelatively small size should reduce thesonic boom ofan executive jet. Aerion, acompany based in Nevada, is hoping tomake a Mach 1.6 executive jet called theAS2. Throttled back to around Mach 1.2,the company reckons it could fly overland without the boom ever reachingthe ground. In 2014 Airbus agreed tocollaborate with Aerion on the project,which aims to have a prototype readyfor test flights in 2019. If supersonictravel does return, it looks like beingmore exclusive than Concorde ever was.

Watching the world go by

Executive jets faster than thespeed of sound are ready to fly offthe drawing board

The view from the executive suite

10 Rational consumer The Economist Technology Quarterly May 30th 2015

IS IT a clock? Is it a rock? It’s a Runcible. That is the name of whatmight be the strangest-looking smartphone ever designed.

Styled after an old-fashioned fob-watch, it takes its name from theworks of Edward Lear, a 19th-century British poet who inventednonsense words. Its Californian (where else) creator, Monohm,bills it as an “anti-smartphone” that does not dare to distract itsuser with alerts and beeps and other forms of information over-load. Instead of lots ofhyperactive apps, for instance, the Runcibleprovides only a simple summary of its user’s social networking.

Runcible is not the only unusual smartphone destined to ap-pear on the market. Google will begin a trial in Puerto Rico laterthis year of a “modular” device called Ara. Users will be able toupdate the phone, saywith a newbatteryora bettercamera, byre-placing one of its plug-in modules. LG sells smartphones with acurved body and a pair of digital eyes on the back that change ex-pression (supposedly to establish “emotional” contact with theuser). And the latest model from Yota, a small Russian maker, has asecond black-and-white display on the back, which is meant to al-low a user to read for hours without draining the battery.

So should consumers expect smartphones to evolve into allsorts of different shapes and sizes? Before the arrival of the touch-screen phones, afterall, there wasa greatervarietyof shapes, fromclamshell designs to slide-out keyboards. Yet unusual designs,such as the Runcible, are likely to remain rarities for now—activi-ties like typing are suited to palm-sized rectangular devices. If con-sumers really want an unusual-looking smartphone then theyhad best wait a few years, says Geoff Blaber of CSS Insight, a mar-ket-research firm. To provide really novel shapes, componentssuch as batteries and screens will first have to evolve, perhapswith both becoming flexible. Even then, because of their clout inthe market and their economies of scale, it is likely to be Apple orSamsung which pioneer them.

The choice of brands available to consumers is multiplyingfast, however. Most of the newcomers are regional brands, whichcome by the hundreds and not just in China, where the phenome-non is most prevalent; Coolpad, OnePlus and Oppo are amongthe better-known names. France has seen the rise of Wiko, whichis now one of the country’s most popular smartphone brands. Af-

rica, too, has its brands, such as VMK, which hails from Brazzaville,the capital of the Republic ofCongo.

Although these firms make much of their regional roots (“Afri-can technology for Africa” is VMK’s tagline), most are combina-tions of Chinese manufacturing and local marketing. The produc-ers may add nice features (one of Wiko’s latest models, calledHighway Pure, for instance, is only 5.1 millimetres thick), but theiraim ismainly to serve buyerswho want to upgrade to a real smart-phone, but cannot afford the latest Apple iPhone or Samsung Gal-axy. The Highway Pure, with a respectable list of specifications,costs less than $300; VMK’s Elikia L costsabout$110. At those pricesmost of these phones represent good value, and they work per-fectly well. (The odd-looking phones are a lot dearer: when theRuncible is launched later this year it is expected to be priced to-wards the top end; the two-screen YotaPhone 2 retails for $680.)

The question is whether, and forhow long, the regional brandswill be available to consumers elsewhere. The competition forcheap smartphones is cut-throat and some of the smaller regionalbrands will not survive, predicts Francisco Jeronimo of IDC, amarket-research firm. Xiaomi, a fast-growing Chinese firm, is oneof the big brands, but problems with intellectual-property rightsmight prevent it from moving as quickly as it would like into mar-kets outside China. That means India’s Micromax may becomethe first big regional Asian smartphone brand to be widely offeredto Western consumers. The company is already India’s biggestsmartphone vendor, according to some analysts, with a marketshare of 22% compared with 20% for Samsung, its main rival. Mi-cromax boasts an impressive line-up of dozens of smartphonemodels, ranging in price from less than $30 to more than $200.

Call the plumberNovelty-seekers are not just interested in design and price; theyalso care about the services that handset-makers offer. Given that,Mr Jeronimo says, buyers should look at brands which view theirdevices as more ofa loss leaderand try to make money by offeringattractive services. This is Xiaomi’s territory. The Chinese com-pany is often compared to Apple and its boss, Lei Jun, feted as asecond Steve Jobs, but the better comparison may be to Googleand its co-founders. Xiaomi’s ambition is to be an online power-house, not justa handset-maker. In April MrLei predicted the com-pany’s services revenue will triple to nearly $1 billion this year,about 6% ofexpected total sales of$16 billion.

Micromaxhasambitions in thisarea aswell. One ofits most in-teresting handsets is Yureka, a $140 budget smartphone poweredby Cyanogen, a variant of Android. It allows users to customiseand extend Google’s operating system and even exchange the on-line giant’s mobile services for other offerings.

“With Android, every phone became the same. And all thedata goes to Google,” says Rahul Sharma, a co-founder of Micro-max. Cyanogen, he says, will allow the firm to become more likeXiaomi by adding differentiating features and services. “If I’mlooking for a plumber, my phone should be able to tell me wheth-er a friend already knows one,” explains Mr Sharma. And smart-phones should come with a selection of services pre-installedwhich users can pick and choose from, he adds, rather than beingoffered only a default set from Google. And if that all sounds a bittoo sensible, there is always the Runcible. 7

Ready to runcible

A guide to alternative smartphones: Want to be different? There are now many handset brands to choose from,besides the ubiquitous iPhone or Galaxy

The Economist Technology Quarterly May 30th 2015 Better batteries 11

1

IT IS hard to imagine modern life withoutbatteries. These storehouses of power

open up new vistas, whether connectingpeople with the world through portabledevices or travelling in electric cars. Yet,like many freedoms, the price is vigilance;the constant fretting over the charge meter.

Anyone who has spent time in an air-port in recent years can attest that one ofthe most popular places to wait for theplane is by the rare wall socket or speciallybuilt tables festooned with electrical out-lets. And for all the promise of the electriccar, the distance it can travel on a singlecharge is limited, adding a new phrase tothe lexicon ofmotoring: “range anxiety”.

The inability of power storage to keepup with new technology frustrates many,especially entrepreneurs in Silicon Valleywho bemoan the lack of a Moore’s law forbatteries. This is the name given to a 1965prediction by Gordon Moore, a co-founderof Intel, that the cost of microchips wouldcontinue to fall as the number of transis-tors crammed onto a given area of siliconwould double every 18 months or so. Forchipmakers like Intel this turned into a self-propelling prophecy that is—just about—still delivering cheaper computing power.

Batteries have improved, but nowherenear the pace of Moore’s law. Most mobiledevices and electric cars are now poweredby lithium-ion batteries (see box 1). Theywere commercialised by Sony in the early1990s and have got steadily better. The bat-teries are lighter and their capacities haveincreased several times over the years, aswitnessed by ever-thinner laptops andsmartphones.

The lithium-ion battery “is almost anideal battery,” says Vincent Battaglia, thehead of the Electrochemical TechnologiesGroup at Lawrence Berkeley National Lab-oratory in California. Lithium is the light-est metal on the planet and it can hold acharge extremely well compared to heavi-er alternatives, such as lead, zinc and nick-el-cadmium. Unlike the latter it does notsuffer from a “memory effect”, whichmeans lithium batteries do not need to be

run down before recharging.There are some drawbacks. Lithium is

highly reactive—overcharging and manu-facturing faults can result in an internalshort-circuit causing the battery to heat upand sometimes burst into flames. Millionsof laptop batteries have been recalled andsome of Boeing’s 787s were grounded in2013 because of fires. Engineers are nowbetter at managing this hazard.

An important measure of a battery’sability is its “energy density”; the amountof energy that can be stored for a givenweightorvolume. Dependingon construc-tion, a lithium battery can store 100-250watt-hours per kilogram—more than twiceas much as a nickel-cadmium one. An elec-tric carwith a 24 kilowatt-hour lithium bat-tery has a range of175km (109 miles) or so.

Even though China is now applying itsmanufacturing muscle to their production,lithium batteries remain relatively expen-sive: typically around $500 per kilowatt-hour of capacity. Hence a battery pack foreven a small electric car can cost around$10,000. Many in the car industry believethe range needs to be close to 500km andthe cost around $100 per kilowatt-hour be-fore all-electric vehicles will move into themassmarket. Thatwould also allowsmart-phones and laptops to run for days.

Such a battery would require a step-change in technology. Many researchersare trying, but often run into difficultiesscaling up promising experiments into aproduct that can be mass-produced. Somescientists are not sure if the energy densityof lithium-ion batteries can be improvedmuch beyond present levels without sig-nificant changes in the materials used tocreate the electrodes.

Dr Battaglia’s team are working onwhat he describes as “transition metals”.These are combinations of manganese,nickel, cobalt and graphite which can beadded to a lithium battery’s electrodes.Once the right recipe is determined, theidea is that it will increase energy densitywithout having to develop a whole newtype of battery. It would, though, be more

Charge of the lithiumbrigade

Energy storage: The lithium-ion battery is steadily improving, but newresearch aims to turbocharge the technology

12 Better batteries The Economist Technology Quarterly May 30th 2015

2 ofan incremental improvement.Other researchers are trying for more.

Yi Cui and his colleagues at Stanford Uni-versity are developing thin films—someonly atoms thick—to enclose the positiveelectrode. This would allow it to safelycontain more lithium, which coupled witha sulphur negative electrode (sulphur, likelithium, also hasa veryhigh energycapaci-ty) would enable a battery to hold aboutfive times as much energy by weight as to-day’s lithium batteries do. A similarly hugeincrease in capacity is promised with workby Chengdu Liang and his team at OakRidge National Laboratory. They are devel-oping a lithium-sulphur battery that has asolid rather than liquid or gel-like electro-lyte. This would also make the batterymore stable. But both projects face severalmore years of work even if these batteriescan be made commercially.

Tiny solid-state batteries, as those witha solid electrolyte are known, are alreadyfound in small devices and sensors, oftenproviding backup power to a microchip.They can be made by depositing materialsonto a substrate, rather like the way semi-conductors are made. Despite an extreme-ly high energy density, making large solid-state batteries has been too expensive forphones and cars. Nevertheless, some com-panies hope to change that. Sakti3, a Michi-gan firm, aims to make big lithium-basedones at the $100 per kilowatt-hour scale—although it does not say when. Dyson, aBritish maker of vacuum cleaners, hasbeen sufficiently impressed by the tech-nology to invest $15m in the company re-cently. Volkswagen has put money intoQuantumScape, a Silicon Valley companyalso working on solid-state batteries.

In theory, lithium-air batteries wouldprovide the highest energy densities—air,after all, is extremely light. Researchershave been experimenting for years how tomake such batteries, but no commercialbreakthrough appears in sight.

In some applications, though, weight isless ofa problem and here lithium will facecompetitors. Giant batteries are being de-veloped to store electricity on the grid,which could transform the market for in-termittent renewable sources such aswindand solar. The way utilities deal withspikes in demand is to add generating ca-pacity—so called “peaker” stations. If sur-plus power could be stored fewer such sta-tions would be needed and supplies couldbe balanced more easily and cheaply.

Less than 0.01% of electricity is present-ly stored, says Philippe Bouchard, vice-president of business development for

EOS Energy Storage, a New York startup.“Every other commodity supply chain hassome form of storage at the point of gener-ation and through delivery,” he adds. MrBouchard is pitching his large container-sized zinc-based batteries for storage toNew York and Californian utilities. Thesecost $160 a kilowatt-hour to store electric-ity, which the company says makes batterygrid-storage financially worthwhile.

Other firms making big batteries in-clude giants like GE in America, South Ko-rea’s LG, Japan’s NEC and startups such asAquion Energy, which was spun out ofCarnegie Mellon University in Pittsburghand is backed by Microsoft’s co-founder,Bill Gates, among others.

Powering homeSuch batteries could also be used by busi-nesses and households. If it were possibleto accumulate power overnight, when it ischeap, in a battery and discharge it duringthe day, when it is expensive, it would saveusers money. Solar and wind power couldbe stored this way too, instead of beingpumped back into the grid in exchange fora discount on the bill. Such systems mightallow some businesses and homes tomove, at least in part, “offthe grid”.

Smaller versions of grid-scale energy-storage systems would be required for do-mestic use. One possibility is flow batter-ies, which generate electricity when a

charged liquid-electrolyte is pumpedthrough them (see box 2). In theory, the ca-pacity of a flow battery is as big as the con-tainers in which the electrolytes arestored—which fora stationary flow batterymay not be a problem. Michael Aziz at Har-vard University has been working on flowbatteries for the grid, but his team thinksthey are close to being able to make a safe,relatively inexpensive system, about thesize of a domestic heating-oil tank, to fit inthe basement ofa home.

Many flow batteries rely on pumpingpricey solutions that contain metals, butthe Harvard team says their system couldhave a greater commercial potential be-cause it could be made more cheaply thanthose using metal-based solutions. This isbecause the battery relies on naturally oc-curring organic materials known as qui-nones, which bear similarities to the ener-gy-storage molecules used by plants andanimals. At present the researchers need touse bromine, a particularly toxic material,on one of the electrodes to boost perfor-mance. “It’s not something I would put inmy basement,” Dr Aziz admits. But histeam are optimistic they will find a way touse organic materials on both electrodes.Such a battery would be allowed in DrAziz’s basement.

It could take several years before flowbatteries fordomestic use come to the mar-ket. But lithium batteries, if they could bemade cheaply enough, could compete as alighter and more compact alternative tobatteries made from heavier and cheapermaterials for storing energy in buildings.

This is the thinking of Elon Musk, thebillionaire founder of Tesla, a Californianproducerofelectric cars (such as the ModelS pictured on the previous page). Tesla isbuilding a $5 billion “gigafactory” in Neva-da in partnership with Panasonic, its Japa-nese battery supplier. Production shouldstartnextyearand by2020 reach the equiv-alent of enough lithium battery-packs topower 500,000 cars.

Rather than banking on a break-through, Tesla reckons steady improve-ments and the manufacturing efficienciesof its giant factory will reduce battery costsby some 30% and help it sell more electriccars. Some of the batteries, though, willalso be offered to businesses and house-holds for energy-storage. In May, Tesla an-nounced the Powerwall battery for thismarket. A 10-kilowatt-hour version willcost $3,500 (excluding controls and instal-lation). Even if big innovations in lithiumtechnology fail to materialise, lithium bat-teries seem to have a bright future. 7

The Economist Technology Quarterly May 30th 2015 Brain scan 13

1

SOMETIMES a childhood fascinationcan shape a lifetime. For Sir Martin

Sweeting it began with forging a commu-nications device out of two tin cans con-nected with a taut piece ofstring. This ledto a schoolboy interest in amateur radio,which turned to a passion in the late 1960sat the time of the Apollo Moon landingsand intoxicating space imagery from filmslike “2001: A Space Odyssey”.

While studying for his PhD in the late1970s at the University ofSurrey in Guild-ford, a town south-west ofLondon, SirMartin made equipment from whateverhe could lay his hands on to pickup sig-nals from American and Soviet satellites.In 1985, with £100 ($150) borrowed fromthe vice-chancellor, he founded SurreySatellite Technology (SST), a firm whichhas pioneered the use ofsmall satellitesbuilt from cheap off-the-shelfequipment.

Space used to be the province ofsu-perpowers, but the arrival of lower-costsmall satellites brought down the price ofentering the space business to a levelwhich companies, research groups, devel-oping nations and even some schoolscould afford. Now Sir Martin is preparingSST for a future in which he plans to pro-vide essential services to a new generationofspace travellers.

Before SST, space was not a particularlypromising field in flared-trousered 1970sBritain. The country had quit the spacerace as post-war enthusiasm for rocketshad dwindled. Government spending cutsled to ambitious launch systems andmanned-flight programmes being aban-doned. Deciding that he did not havemuch chance of joining America’s spaceagency, NASA, Sir Martin started to trysome things for himself. With the firstmicrocomputers arriving on the scene hewondered if it was possible to use one tobuild a small satellite that was reasonably“intelligent” compared with the large andexpensive hard-wired satellites that werebeing put into orbit at the time.

Sir Martin persuaded the university tolet him try. “I didn’t have any money sohad to go around scrounging from in-dustry and anybody else I could think of,”he says. With a team ofvolunteers thesatellite was built. It was about the size of acouple ofmicrowave ovens and weighed

72kg. Some of the big satellites that govern-ment space agencies were putting intoorbit were as large as a London double-decker bus. Through the amateur radionetworkhe managed to persuade NASA toprovide a free launch, and in 1981 themicrosatellite piggybacked into space aspart ofa mission to put a large scientificsatellite into orbit.

The Surrey satellite was a novel pieceofengineering. Besides containing anon-board computer, allowing the satelliteto be reprogrammed from the ground, itwas also fitted with an early charge-cou-pled device, a type ofsensor which wouldbecome the heart ofmodern digital cam-eras, and a digital speech-synthesiserextracted from a toy. This meant that be-sides being able to take pictures ofEarththe satellite could “speak’’ to the worldbelow. The idea was that radio amateursand schools could use basic equipment totune in to hear data, such as the tempera-ture, being read out. “Amazingly,” says SirMartin, “to everyone’s surprise it worked.”In 1984 the team built a second satellitecontaining a new computer and an updat-ed camera. It too was given a free launchby NASA and, although now not in thebest ofhealth with its solar-poweredbatteries deteriorating, it is still transmit-ting back to Earth.

The spin-out powers upAfter years ofbegging and borrowing, SirMartin and a team offour others beganSST to see if they could turn the produc-tion ofsmall satellites into a business.They were encouraged by interest from anumber ofcountries that wanted to get afoothold in space, including Pakistan,South Korea, Malaysia and Chile. “Andbecause we were linked to a university, wewere prepared to train their people andshare information with them,” adds SirMartin. Many of those tutored during theconstruction of the firm’s first satelliteswent on to form the nucleus ofothercountries’ space agencies—and produceoverseas competitors that would come toemulate SST.

SST has, nevertheless, maintainedwhat it says is a 40% share of the globalexport market for small satellites. It hasbuilt and launched more than 40 of them,ranging in weight from 3.5kg to 660kg, andhas over 20 others in production alongwith various payloads for other spacemissions. The company now employs 500people. In 2009 Airbus bought 99% of thefirm’s shares from the University ofSurrey,which retained 1%. As part of the deal with

Britain’s spaceman

Sir Martin Sweeting: The pioneer ofsmall satellites is laying plans for theinfrastructure and services neededfor travel to other planets

14 Brain scan The Economist Technology Quarterly May 30th 2015

Offer to readersReprints of Technology Quarterly are availablefrom the Rights and Syndication Department.A minimum order of five copies is required.

Corporate offerCustomisation options on corporate orders of100 or more are available. Please contact us todiscuss your requirements.For more information on how to order specialreports, reprints or any queries you may haveplease contact:

The Rights and Syndication DepartmentThe Economist20 Cabot SquareLondon E14 4QWTel +44 (0)20 7576 8148Fax +44 (0)20 7576 8492e-mail: rights@economist.comwww.economist.com/rights

2 the European aerospace giant, SST contin-ues to be run as an independent companyand Sir Martin, who was knighted in 2002,serves as its executive chairman.

Building small satellites from industry-standard parts means reaping the benefitsofhuge investments by car companies,consumer-electronics firms and others indeveloping sophisticated components formass-produced goods. This can limit theabilities ofa satellite but it provides econo-mies ofscale which are impossible toachieve ifmaking bespoke components.“We were being parasitic, ifyou like,” saysSir Martin.

Besides lower costs, commodity com-ponents have become increasing reliable—essential in building a satellite which,once in orbit, may not be repairable. In theearly days, this could mean testing individ-ual widgets, such as transistors or capac-itors. Hundreds of them might have to beshaken, cooked and abused in variousways to see which ones were best. Nowa-days, with modern manufacturing meth-ods, components have been miniaturisedand integrated into units, like microchips.So testing can be done on more completesystems. And because tolerances havebecome much tighter, fewer parts nowcontain defects. However, even some ofthe whizziest components made for terres-trial use might not be up to the rigours ofspace. One of the biggest problems isradiation. Some bits ofa smartphone, acommon source ofminiaturised electron-ic components and sensors, are not asresilient to radiation as other parts. Soknowing which bits to use and which toavoid has become part of the satellite-making art.

In 1998 the company’s engineers want-ed to see how small they could make afunctioning satellite. The result was aboutthe size ofa beachball (the 3.5kg one inSST’s inventory). It was one of the first ofwhat came to be called a nanosat. Overthe next five years or so some 1,000 nano-sats are expected to be launched by com-panies, startups and researchers.

By 2000 SST had reached a turningpoint. “Before then these small satelliteswere useful, interesting, but could carryout only a few niche tasks,” says Sir Mar-tin. After 2000 technology improvedrapidly, providing greater capabilities.“Now our satellites can rival the majorityofconventional satellites,” he adds. Takingpictures ofEarth, for instance, can now bedone in high resolution with a small satel-lite fitted with the latest high-power imagesensors and, maintains Sir Martin, at

about one-fifth of the cost ofmost tradi-tional large satellites.

Some ofSST’s latest satellites have aone-metre resolution (ie, one pixel on thespacecraft’s sensor represents one squaremetre on the ground). This means it mightbe possible to identify a shape smallerthan one metre, but not in detail. So thesatellite could, for instance, easily spot acar. Some big high-end satellites can takepictures down to a resolution of50cm (thelegal limit in America for commercialsatellites), although military ones arethought to be able to zoom closer.

The snag is, at some point the laws ofphysics dictate a more powerful telescopemeans a bigger and more expensive satel-lite to transport it. And as the resolutionincreases the field ofview diminishes, sothe image produced is a bit like looking atthe world through a drinking straw.

As a result there are plenty ofuses forsatellites with resolutions of20 metres orso, including the ability to take a picture ofan area several hundred kilometres wide.This can, for instance, help in geographicsurveys, monitoring deforestation orobserving crops. In partnership withgroups in other countries SST also oper-ates a constellation ofsatellites which canrapidly be deployed in the event ofadisaster. The lower-resolution ones can,say, reveal the extent ofa flooded area, andthe higher-resolution cameras take a lookat specific problems, such as a bridge thathas been washed away. These satelliteswere used following the Asian tsunami in2004, Hurricane Katrina in 2005 and afterthe recent earthquake in Nepal.

To a new wavelengthThe company is now developing observa-tion systems that go beyond the visiblespectrum. These satellites use synthetic-aperture radar, which can penetrateclouds and workat night. It relies on usingalgorithms to process radar pulsesbounced backfrom the ground to build ahigh-resolution image.

SST is also looking at ways to addressworries about space debris. There aresome 500,000 bits ofspace junk in orbit,but Sir Martin does not think that puttingup more satellites—even 1,000 nanosats—will make much difference. “The real riskis when a satellite fragments,” he says.One satellite exploding because ofa faultor when it reaches the end of its workinglife can produce thousands ofpieces ofdebris. Hence there are plans to developsystems that can tip old satellites out oforbit for them to burn up safely on re-

entry. TechDemoSat-1, a one-cubic-metresatellite which SST launched in 2014 tocarry out a number ofstudies, is fittedwith a “de-orbit sail” developed by Cran-field University in Britain. The ultra-thinsail will be deployed at the end of themission to turn it towards Earth.

Other satellites give another hint ofwhere SST is going. The company is build-ing its first small geostationary platform,which instead oforbiting Earth will re-main in a fixed position relative to thesurface, as telecommunications satellitestend to be. This is for Eutelsat, a Paris-based company, which provides satelliteservices to telecoms firms and broadcast-ers. In addition SST is building equipmentfor Galileo, a new constellation ofsatel-lites that will operate a European GPS

navigation system. “The reason we are doing that is we

want to learn about navigation payloads,”says Sir Martin. The discovery ofwaterdeposits, he believes, means that in thecoming decades there will be sustainedhabitation on the Moon and Mars. “Weneed to play a part in supporting this,” headds. In likening space travel to the Cali-fornian gold rush that began in 1848, SirMartin points out that it was not necessar-ily those who went to find gold that mademoney, but rather those who provided theinfrastructure: railways, hotels, picks,shovels and the like. So, SST aims to be aservice provider supplying, say, a GPS

system around the Moon, or communica-tions satellites to relay data from otherplanets. “We need to be building the ex-pertise up now ifwe are going to be readyto do such things,” he says. It is a far cryfrom the days of tin-can telephones.7

Some of whizziest components made for terrestrialuse might not be up to the rigours of space