SPACE: A sudden light
New capabilities, new entrepreneurialism and rekindled dreams are making space exciting again, says Oliver Morton
SHORTLY after sunset there had been juddering green stabs of lightning to the south, but by a quarter to one in the morning there is nothing in the warm, wet July air over Cape Canaveral but a thin patchwork of moonlit cloud. And then, precisely at the time it was meant to happen, there is a sudden light on the horizon, some 18km away. A light that rises.
A 550-tonne machine taller than a 20-storey building is throwing itself into the sky. Its initially unhurried ascent burnishes the clouds bronze. As the rocket climbs above them, its pace quickening, the roar of its engines speaks to its power as the sharp line of its exhaust illustrates its precision.
Two minutes and 21 seconds after launch, 61km above the Atlantic and still well in sight, the first-stage engines shut down. The two stages of the rocket—a Falcon 9, built by a company called SpaceX that is seeking to reinvent space travel—separate; the single engine of the second stage ignites, driving its cargo ever faster to the east where it is quickly lost to sight. When, nine minutes after departure, that engine is turned off, the rocket will be in orbit 240km above the Earth, travelling at 7.5km a second. Two days later its cargo of provisions and scientific experiments will be delivered to the International Space Station (ISS).
Will ye no come back again?
The spectacle, though, is not over. Shortly after separation, at an altitude of more than 100km and while still climbing, the rocket’s first stage slews around before bringing three of its engines back to life to change its course. It reaches the highest point of its trajectory and starts to fall back to Earth, engine-end first. A bit more than halfway there, the engines fire again to cushion the shock of its re-entry.
When it is about 10km up, and still falling at well over the speed of sound, the engines fire for the last time. The rocket’s return has none of the stateliness of its departure; the bright light plummets to the horizon with swift purpose. As it reaches the ground, a flat, fiery flower spreads out from its base. Four landing legs the size of oaks smack into the concrete pad. A second later the double whipcrack of its sonic boom signals the end of the eight-minute journey.
Next year it will be 60 years since people first witnessed the majesty of a satellite being launched into orbit: Sputnik 1, hurled into the night sky in Kazakhstan early on October 5th 1957. Such knocking on heaven’s door remains a thrilling experience, and probably always will. The heavens themselves, though, have over that time become significantly more pedestrian.
Just 15 years separated the launch of the first satellite and the return of the last man from the moon, years in which anything seemed possible. But having won the space race, America saw no benefit in carrying on. Instead it developed a space shuttle meant to make getting to orbit cheap, reliable and routine. More than 100 shuttle flights between 1981 to 2011 went some way to realising the last of those goals, despite two terrible accidents. The first two were never met. Getting into space remained a risky and hideously expensive proposition, taken up only by governments and communications companies, each for their own reasons.
Now SpaceX, founded in 2002 by Elon Musk, an entrepreneur who around the same time also set up Tesla, a car company, is trying to provide the cheap, reliable, routine route to orbit that the shuttle could not. It is the first company since the 1980s to enter the launch business with newly designed rockets, and has developed some completely new tricks. Though the July lift-off of its ISS-resupply mission was the sort of thing Cape Canaveral has seen hundreds of times, the successful return of a rocket’s first stage had been witnessed there only once before, late last year.
Mr Musk intends to keep up the pace of innovation. Late this year or early next his company will launch a rocket with three reusable stages, the Falcon Heavy, capable of handling bigger payloads than any other launcher working today. SpaceX is also developing a rocket engine more powerful than any previously developed for a commercial programme. More striking still, before the end of 2017 it is due to start delivering human space travellers to the ISS—the first private company to do so, unless Boeing, which has a similar contract with NASA, pips it to the post.
Impressive as its prospects are, SpaceX is only one contender. Blue Origin, a company backed by Jeff Bezos, the boss of Amazon, has a sub-orbital rocket, the New Shepard, that can come back from the edge of space to land under power in the same way the first stage of a Falcon 9 does. It may well take a capsule with people on board into space next year. A number of other new companies with unmanned rockets are entering the launch business, too.
New rockets, though, are not the only exciting development. The expense of getting into space during the 1980s and 1990s led some manufacturers to start shrinking the satellites used for some sorts of mission, creating “smallsats”. Since then the amount a given size of satellite can do has been boosted by developments in computing and electronics. This has opened up both new ways of doing old jobs and completely novel opportunities.
The 24 satellites of the American government’s Global Positioning System probably represent the world’s single most important space-based asset, vital to the American armed forces and tremendously useful to a couple of billion smartphone users. Building up the constellation to its current size took two decades. Now smallsat companies talk about launching constellations several times that size in just a year or two. Those satellites will be in low orbits, but not all small spacecraft will stay close to home. By the end of 2017 Moon Express, based in Florida, hopes to deliver a 10kg payload including a rover to the Moon, making it the first commercial company to land anywhere beyond Earth. Other companies are looking at smallsats as a way of prospecting asteroids for mining.
No single technology ties together this splendid gaggle of ambitions. But there is a common technological approach that goes a long way to explaining it; that of Silicon Valley. Even if for now most of the money being spent in space remains with old government programmes and incumbent telecom providers, space travel is moving from the world of government procurement and aerospace engineering giants to the world of venture-capital-funded startups and business plans that rely on ever cheaper services provided to ever more customers.
As they prove that they can make money, they will grow further, and fast. But in many cases money is not the only aim. Their founders are people who think that going into space can benefit the human race more broadly. Mr Musk wants to set up a permanent colony on Mars in a couple of decades. Mr Bezos hopes that millions of people will one day work in orbit. Neither of these aspirations seems likely to be realised. But the effort to get there may well re-establish space as a realm of possibility and inspiration.
Flocks of cheap little satellites could transform the space business
ROCKETS are the thrilling, spectacular bit of space flight. But without something useful to carry they are basically just fireworks. To get a sense of the new entrepreneurial approach to unearthly enterprise, start instead with the radical changes in what it takes to make a spacecraft.
In Palo Alto, California, there is a factory that has been making spacecraft since the year Sputnik was launched, and before anyone in Palo Alto had heard of Silicon Valley. SSL, previously known as Space Systems/Loral, has built more than 100 communications satellites, of which 81 are still in operation today. The dozen or so currently spread through this warren of clean rooms the height of cathedral naves represent more than a year of the company’s order book.
They are all based on the same structure: a cylinder 1.2 metres across enclosed in a square box. The more the satellite has to do, the taller the box it is built on, the longer its solar panels and the larger and more complex the array of antennae and reflectors through which it sends data to its earthbound clients. Sky Muster II, nearing completion, is among the biggest. Designed to provide broadband communications across the less densely populated parts of Australia, it stands nine metres tall, with a complex array of reflectors tailored to serve the outback.
The communications-satellite business is dominated by four operators, Eutelsat, Inmarsat, Intelsat and SES. They make most of their money from companies that want to send television signals to people’s homes, but also serve markets for data transfer and mobile communications. They demand ever more of the handful of aerospace companies like SSL that have the expertise to compete for their custom, says Paul Estey, head of engineering and operations at the factory.
The industry is innovative but also very loss-averse. The smallest of the SSL communications satellites may sell for $100m or so, the biggest for perhaps three times that. Add on $100m for the launcher, and the satellite may not start showing a profit for a decade. Because of the need for a long lifetime in a hostile environment with no chance of any repair, a new technology that carries any significant risk will simply not be flown.
An hour’s drive up Route 101 you will find a very different spacecraft factory. Planet, until recently known as Planet Labs, occupies a shabby-chic building in the South of Market area of San Francisco. A room the size of a largish Starbucks on the ground floor houses the desks and tools needed to build 30cm-long satellites each weighing about five kilos. If you know what to look for, you will recognise many of the components as coming from other sorts of device, most notably smartphones. Making one of these “Doves” (pictured), as Planet calls them, takes about a week. At the back of the room there are dozens packed up ready to be shipped off. This is the new face of space: small objects, large numbers.
Doves are part of an extended family of very small satellites known as cubesats. In the late 1990s researchers at Stanford University and California Polytechnic State University in San Luis Obispo realised that a certain amount of standardisation would make very small satellites much easier to launch. They came up with a standard called the “1U” cubesat: a box 10cm by 10cm by 11.5cm with electronic and physical interfaces that would allow it to fit alongside others of its ilk in a dispenser that could fly as a “secondary payload” (launchers often have more capacity than they need for their main cargo). The standard caught on. By early 2013 some 100 cubesats had flown, and the tools required to design and build one were so well developed that a class of schoolchildren with an inspired teacher could take on the task.
Planet’s founders, Chris Boshuizen, Will Marshall and Robbie Schingler, thought cubesats might be the basis of a business. While working at NASA’s Ames Research Centre in the early 2010s, they looked at what could be done with the largest telescope that would fit into a “3U” cubesat, three 1Us stuck end to end. Pointed towards Earth from a low orbit like that of the ISS, such a telescope could take pictures with a resolution of five metres or a bit better. That was nothing like as good as the images being sold by companies using bigger telescopes in much larger satellites. But 3U cubesats could be deployed by the dozen or the hundred. For some markets, such as agricultural monitoring, the sheer quantity of the information gathered by such flocks might make up for the low resolution.
The first 28 Doves were sent up from where they were deployed to the ISS in 2014. The launch was celebrated at Planet’s headquarters with a pancake breakfast, as has been each of the 13 launches since. Planet currently operates 63 spacecraft. Their capabilities may be limited by their size, but the company claims that the sophistication of their technology is a match for any satellite anywhere. And they support a promising business model. Mr Marshall says Planet now has over 100 customers for the data that the Doves send back. It looks poised for significant growth.
Planet’s success stems partly from the continuously falling cost and rising capability of consumer electronics—especially components for smartphones, which sell by the billion and where size and low power usage are crucial. But that would be of no use without a willingness to improve the satellites frequently—indeed, incessantly. By June this year the Doves had been through 14 upgrades. Today’s spacecraft have a different camera from their predecessors, new antennae, rebuilt electronics and a power system based on the lithium-ion battery packs used in Tesla cars, rather than the original AA battery format. The satellites now “see” in four colour bands rather than the original three. They have become much better at telling where they are and which way they are pointing. According to Mr Marshall, in terms of performance per kilo the Doves are now 100 times better than the state of the art five years ago. Such agile innovation is normal in Silicon Valley, but it is not something the satellite world has seen before.
Fly, my pretties
To do things this way requires an attitude to risk alien to the world of big, expensive satellites: Planet expects some of its innovations to fail. It knows that Doves launched from the ISS have only a short life anyway, re-entering the atmosphere after nine to 18 months aloft. This attitude speeds up progress and provides resilience for the company as a whole. A big communications satellite can carry the fate of a whole company with it. When Astra1A, the first dedicated direct-broadcast television satellite, was sitting on top of Europe’s first Ariane 4 rocket in 1988, Rupert Murdoch knew that if it blew up, his nascent Sky broadcasting business would blow up with it—quite possibly taking the rest of his media empire down in flames too. Planet has twice had the bad luck to see a flock of Doves fall to Earth from the fiery wreck of a failed launch, and lived to tell the tale.
A company can welcome risk only if its investors take the same view. Planet’s do. This is another consequence of building a business on small, cheap satellites; the amount of capital needed is relatively modest. Planet has raised almost all its capital from Silicon Valley angel investors and venture funds. Just as technological improvement can be accelerated when your satellites weigh just a few kilograms and have parts lists in the 1,000s, so getting funding is a lot easier when their cost is a few hundred thousand dollars or less. The total invested in Planet to date, after three rounds, is $158m; at SSL that would buy a single satellite.
In 2001-05, venture investments in space businesses worldwide totalled just $186m. In 2011-15 they had risen to $2.3 billion, according to a study by the Tauri group. Half of those investors were based in California, and most of this money has gone either into small satellites or into new launchers tailored to those satellites’ requirements. Venture capitalists feel increasingly at ease about the technology involved.
The business aspirations of companies like Planet are familiar, too. As the Tauri report puts it, the new wave of space companies has been able to sell itself to VCs as a way to “follow the path terrestrial tech has profitably travelled: dropping system costs and massively increasing user bases for new products, especially new data products”. Fashion is another factor. Like Doves, Silicon Valley investors flock; the past few years of success for SpaceX, founded by one of their own, has made space a particularly appealing place for the flock to settle. This new source of capital looks like producing a great many satellites. In July Euroconsult, a consultancy, estimated that in the period from 2016 to 2025 some 3,600 commercial small satellites might be launched, including over 2,000 flown by VC-funded Earth-observing companies.
Others, including some with deeper pockets, want to take the smallsat revolution further. Today’s big communications satellites are almost all to be found in an orbit 36,000km above the Earth. This is because, at that height, it takes them 24 hours to go round once—which means that, seen from the ground, they seem to sit stationary in the sky. In businesses that depend on a single antenna pointed in a single direction, that is a huge advantage. But it has costs. The amount of data you can handle with a given antenna and amplifier drops off according to the square of its distance from the surface. This means that closer to Earth you can do more with less. You can do it faster, too: going 36,000km up to “geostationary” orbit and back again delays a radio signal by a quarter of a second, a problem for some applications.
All the same, communications satellites have mostly forgone the advantages of lower orbits, for two reasons. The lower the orbit, the more satellites you need to make sure one can always be seen from the ground. And satellites that move across the sky require receivers that can track them. This does not mean moving dishes; today’s receivers can track electronically. But such technology is demanding.
The more the merrier
OneWeb, a project being put together by, among others, Intelsat, the Virgin Group and Airbus Industries, is based on the idea that modern antennae can surmount this communication problem, and that the smallsat approach can sort out the coverage problem. It plans to use some 648 satellites in orbits just 1,200km up to offer seamless communications to any spot on Earth. Its business plan turns the need to cover everywhere to cover anywhere into a feature by focusing on developing countries; nowhere will be too remote for it to serve. The first satellites are to be launched next year.
This is not something you can do with cubesats, or on a startup budget. OneWeb is a multi-billion-dollar proposal. Its prototypes are being made at an Airbus plant in Toulouse. In Florida OneWeb and Airbus Space and Defence are building a factory where they hope to produce up to four 150kg spacecraft every day, using highly automated systems; that is more by an order of magnitude than anything the satellite world has seen before.
Not only is the project technologically very ambitious; it also faces a lot of competition. Google, where OneWeb’s innovators were working at one point, is looking at stratospheric balloons as an alternative way of providing connectivity in the developing world. Facebook is eyeing high-flying solar-powered drones.
The incumbent communications-satellite industry is paying attention, too. At Google the OneWeb founders worked on a system called O3B, named for the “other three billion” people not yet getting data services. After they left, the system went forward without them. When it is finished, it will consist of 20 satellites orbiting at about 8,000km. This summer SES, one of the big four comsat operators, took complete control of the project, buying out Google and its other original partners. Meanwhile SpaceX, which until now has operated purely as a launch provider, is talking about a low-orbit communications system of its own, with perhaps 4,000 small satellites. That one project would use three times as many spacecraft as there are in the skies today.
Earth-observation satellites are changing the world—yet again
IN TERMS of engineering ambition, operational complexity and capital requirements, big communications-satellite constellations outstrip the small-satellite revolution in Earth observation. In terms of world-changing potential, though, things may well be the other way round.
Satellites are only a marginal part of the communications business; they matter in some niches, such as multichannel television, but they represent only a small fraction of the $2 trillion telecoms business. The marginal can still matter. The as-yet-unconnected “other three billion” that projects like O3B and OneWeb aim to serve are on the margins of the world economy, and systems that connect them up affordably would be a great boon. But it would be an expansion of the remarkable transformation in computing and communications already being wrought by smartphones connected in all sorts of other ways. What is now happening in Earth observation, on the other hand, is a whole new story. For the fourth time in 60 years, space is revolutionising the way people think about the planet.
The first revolution might be called an anywhere revolution. From the early 1960s on, spy satellites were able to look wherever their handlers wanted them to, even deep into enemy territory. They allowed cold-war adversaries to assess each other’s nuclear and other capabilities and provided a way of monitoring arms-control agreements. That helped to keep the cold war cold.
The second revolution was an everywhere revolution. The pictures of the Earth taken by the Apollo astronauts gave the planet’s inhabitants their first sight of their common home seen from afar. Contrasted with the dead husk of the moon and the infinite emptiness of space it seemed small, beautiful, intensely precious. Those pictures accelerated the advent of modern green politics.
The third revolution was another anywhere revolution. This time, though, the novelty was to know your position anywhere that you happened to be. The GPS satellites launched by America’s Department of Defence allow billions of devices to pinpoint their precise positions. That smartphones, cars, goods containers and girl guides know exactly where they are is now central to everything from orienteering to Uber.
The current, fourth revolution is both an anywhere and an everywhere revolution. It is the transformation of the Earth into a gigantic set of data that can be both interrogated and extrapolated.
The number of Doves Planet is able to fly allows it to provide images of every point on the planet fairly frequently; its ambition, likely to be realised fairly soon, is to use “sun- synchronous” orbits (see graph) to image everywhere on Earth at the same time every day. Spire, another cubesat startup based in San Francisco, does not look at the Earth’s surface but listens to its radio signals. Every ship on the planet is required to have a transmitter that continuously broadcasts its location, and before long Spire expects to have data on every ship on the planet every hour.
All-seeing
BlackSky, a startup based in Seattle, is at the anywhere end of the market. Its satellites are larger than Doves, and their bigger optics give them better resolution (one metre or so, meaning that they can pick up cars, which matters for a lot of applications). They can also be made to take pictures of targets off their orbital track, rather than seeing only straight below them. With 60 of these satellites in a range of orbits, the company aims to be able to produce a picture of any point on the Earth’s surface between 55ºN and 55ºS within 90 minutes of being asked. Other new outfits offer different combinations of resolution and repeat visits.
Cloud storage and processing play a big part in this new revolution. Planet has invested heavily in the pipeline that takes raw data from its 12 ground stations around the world and turns them into a usable product, but it buys storage and processing power as needed from cloud-computing companies. Without such services, startups could never cope with the terabytes of data that their satellites produce every day.
New markets matter too. The Earth-observation companies that started up in America in the 1990s all had a single dominant and expert customer for their high-resolution images: the little-known National Geospatial-Intelligence Agency in Virginia. Serving its requirements made money for the companies involved but hardly encouraged diversity. The industry eventually coalesced into a single company, DigitalGlobe. It is thriving; this September it will launch another of the big, capable high-resolution satellites it puts into orbit every few years. But the government still accounts for well over half its sales.
The new companies will also sell to the government, but few if any of them are relying on it. Instead, their hopes of rapid growth rest on customers who have not previously used satellite data but have questions they want answered. Both the satellite companies and the third parties that use their data have invested heavily in machine-learning technologies that can extract those answers from the huge amounts of data stored in the cloud by understanding what they see and recognising when things change.
They can tell a shipping line—or, soon, an airline—exactly where all its vessels are. They can chart economic growth by recognising the spread of cities and the traffic within them, or the amount of light that they give off at night. They can provide a reinsurance company with daily updates on any changes relevant to its risk portfolio. They can inform futures traders about the state of crops across an entire continent, or individual farmers about the state of crops in a particular field. They can combine their data with other georeferenced data, such as Twitter feeds, to produce images of disasters, demonstrations, conflagrations and celebrations as they happen.
If you think the best way to look for some truth about America is to count the cars on the New Jersey Turnpike, it is easily done. The same applies to any equally obscure metric in any other country. The potential of immense sets of data that cover the world in growing detail, are refreshed more or less in real time and can learn to pick up all sorts of objects and phenomena autonomously seems inexhaustible.
In among all the novelty, old sorts of forecasts will be overhauled, too. As well as hearing radio signals from the Earth below, Spire’s satellites can listen to the transmissions from America’s 24 GPS satellites, and from similar systems being fielded by Europe, Russia and China. Given their different orbits, the Earth will sometimes come between the two satellites, and its radio signal will have to pass through some of the Earth’s atmosphere before the planet blocks it out completely. The way that the signal fades in the atmosphere can be used to calculate the temperature and pressure along the line connecting the two spacecraft, providing a valuable new source of raw data for weather forecasting. Spire has 12 satellites today and hopes to have 44 before the year is out. By the time it has 100, it could be producing 100,000 atmospheric cross-sections every day: terabytes of valuable data from thin air.
Being cheap is not the be-all and end-all of a launcher
FOR decades, lower launch costs seemed to be the sine qua non of progress in space travel. Enthusiasts saw reducing them by orders of magnitude as the key to being able to do much more in space. That is one reason why there is so much excitement around SpaceX, which has undercut the competition enough to take a significant share of the launch market. Its potentially reusable spacecraft seem to promise continuing reductions in launch costs in the future.
The smallsat revolution shows that this stress on dollars per kilogram was too simplistic. If you can get much more capability out of each kilo, then the cost of doing things in space will drop even if the cost of launches does not. In the smallsat world innovation comes first and new launch services follow. The key factor is not necessarily a very low cost per kilo, but new standards and speed of service.
This is the market that Peter Beck, CEO of an American-owned, New Zealand-based company called Rocket Lab, wants to serve. His company’s Electron rocket is due to make its first flight from New Zealand’s North Island later this year. Backed by Silicon Valley money, the Electron is designed to deliver a 150kg payload to a sun-synchronous orbit for just under $5m—the same price as that currently charged by Spaceflight, a Seattle company that brokers “ride-share” opportunities for smallsats to fly as secondary payloads on big launchers.
Rocket Lab’s $33,000 per kilo sounds dear when a Falcon 9 can deliver a kilo to low orbit for a tenth of that price or less. But you have to buy in bulk, paying $62m or more for launching 20 tonnes on a whole Falcon 9. And you may have to wait for a couple of years because there is a queue. For little agile companies currently shopping around for shared rides to often suboptimal orbits, like that of the ISS, 30 3U cubesats in just the right orbit within months of signing a $5m contract sounds a lot more appealing.
Building rockets with the low unit costs that smallsats require is challenging, even if the payloads are modest. Mr Beck’s response combines mass production, new manufacturing techniques and materials and new ideas. Rather than have different engine designs for both the first and the second stage, he has gone for just one type, nine of which are used for the first stage and one for the second. Not coincidentally, this is the same cost-saving approach as that taken by the Falcon 9: SpaceX has shown it is cheaper to build lots of engines to the same design than smaller numbers to a range of them. Rocket Lab also uses 3D printing to produce the engines, and makes its fuel tanks out of carbon composites, which being lighter give the engines less to lift. And it has some tricks all of its own, notably the use of battery-powered pumps to push fuel and oxidiser into the engines.
Alpha, a smallsat launcher being developed by Firefly, a company set up by SpaceX veterans, uses similar materials, but has a different new idea for getting the fuel into the engines, and is also using a novel clustered-engine design called an aerospike on its first stage. Richard Branson’s Virgin Galactic is in the market too. The company’s original purpose was to give tourists joyrides in sub-orbital spacecraft, and that is still on the cards, but the company is also planning to launch smallsats using LauncherOne, a rocket that will be carried under the wing of a converted Boeing 747. Again, the engines are largely 3D printed and the tanks made of carbon composite. The first flight is expected next year.
A crowded space
In its early days SpaceX, too, was aiming for the small-launcher market; the Falcon 1, which first flew in 2007, was much the same size as the Electron. But it was ahead of its time. There was a need for it but not a viable market, the company’s COO, Gwynne Shotwell, has since said. Luckily for Mr Musk, a NASA contract for resupplying the ISS made possible the development of the Falcon 9. Once it was making big launchers and space capsules, SpaceX did not return to the smallsat market; instead it branched out into the market for launching multi-tonne communications satellites to geostationary orbit, which had been dominated by Arianespace, a European consortium.
Mass-produced engines and other innovative approaches have made SpaceX very competitive on cost, but there are limits to how useful that is in this market. Russia, China, Japan and India, as well as Europe, all have their own launch industries, and will keep them for national-security reasons. That might not matter if SpaceX were able to increase the overall size of the market, but rockets typically cost less than the satellites they launch, and it is the total cost that sets demand. Making rockets $10m or $20m cheaper is neither here nor there.
So SpaceX (which declined to comment on the record for this article) has little commercial incentive to slash its prices, and at present is has no obvious new markets. Modest smallsat constellations do not make sense for it; the manifest that Rocket Lab hopes to spread over 50 launches in a year would fit on a single Falcon 9. And the only really big smallsat constellation, the OneWeb communications system, has signed launch contracts with Arianespace and Virgin Galactic (both companies in which OneWeb’s owners have stakes). This may be why SpaceX is talking about building its own constellation of 4,000 communications satellites. A venture on that scale might get real benefits from very low-cost Falcon 9 launches.
It’s not what you launch, it’s what you do with it
ON EARTH, valuable assets are serviced, upgraded and recycled; they are also protected and, now and then, attacked. None of this is yet happening in space. But all of it could.
Some satellites break down; all of them, eventually, run out of fuel. Robotic service spacecraft that can identify a satellite, grab it and manipulate it could get around those problems. Orbital ATK, an aerospace company based in Virginia, has developed a spacecraft along these lines that works a bit like a mobility scooter; it provides somewhere with an engine for an elderly spacecraft to settle down in when it can no longer get around on its own. Orbital hopes to use such spacecraft to extend the lives of communications satellites that are out of fuel but still making money, and has signed a contract with Intelsat to this end.
An alternative to assisted mobility is refuelling. The Naval Research Laboratory, NASA and DARPA—the Pentagon’s advanced-technology arm—are all working on various projects for spacecraft that could refuel satellites and even repair them in orbit, using a range of tools and complex robotic arms.
A more radical approach is to use orbiting robots to make new spacecraft rather than service old ones. Tethers Unlimited, based in Washington state, is working on a “SpiderFab” that would combine robotic arms with a form of 3D printing to create structures much larger and more delicate than anything that can fit into the fairing of a launcher. Satellite-makers have become adept at folding up solar panels and antennae so as to fit a lot of spacecraft into those small spaces. But the complicated unfurlings, articulations and poppings-into-place required tend to increase both expense and risk, and make some approaches impossible.
Structures built with technologies such as the SpiderFab could change the way the communications-satellite industry works. Platforms with big solar panels and engines would take care of the housekeeping for a whole range of communications packages that could be smaller and launched more frequently, thus keeping the technology much more up-to-date and reducing risk.
All this might have an even bigger effect on science. Space is a great place for telescopes, but it is very difficult to get a big one up there. One reason why the James Webb Space Telescope that NASA and the European Space Agency are due to launch in 2018 has a budget of $8 billion is the need to fold a sunshield the size of a tennis court and a polished mirror 6.5 metres across into the 5.4-metre fairing of an Ariane 5. A combination of big structures and techniques that let a number of small mirrors spread over a large area do the work of a single much bigger mirror would allow remarkable new instruments to be built. Such instruments might be much better than those on the ground at observing, for example, the fascinating planets being discovered around other stars.
Another use for robots in space is asteroid mining. Some asteroids have orbits similar enough to the Earth’s to allow a spacecraft in orbit to get to them with a relatively small amount of fuel—much less than what is needed to get it into orbit in the first place. Like many other staples of science fiction, mining these flying boulders and mountains is now on the Silicon Valley startup agenda. The commodity of greatest interest is not a precious metal (though some of those are to be found on asteroids) but something that in space is much more valuable: water.
Human space travellers need water, as well as oxygen, which can be made from water. They, and their spacecraft, also need fuel: hydrogen made from water fits that bill. Once a certain amount of activity is taking place in orbit, especially if it involves a human presence, getting water from asteroids, in some of which it can be found either as ice or as hydrated minerals, could start to make more sense than hauling it up from Earth.
This will take time, but Deep Space Industries and another asteroid-mining startup, Planetary Resources, recently found a patient investor. Luxembourg knows a bit about space; two of the big four satellite operators, Intelsat and SES, in which its government is a shareholder, are based there. It is also well able to afford a flutter, and tightly knit enough to give a far-out idea with a few enthusiastic supporters a hearing. This summer Luxembourg announced that both companies would benefit from the 200m it will be spending on asteroid-mining initiatives.
What goes up must come down
Fuel from beyond could keep some satellites in orbit indefinitely. Others, though, need to be got rid of. In the lowest orbits—those of the ISS and below—the problem has a natural solution: drag in the outer reaches of the atmosphere will bring anything down in a matter of decades (the ISS has to be regularly boosted). But in other orbits space debris—consisting of dead satellites and their fragments, as well as the leftovers of discarded launchers—builds up. Even in the most debris-ridden orbits, between 700km and 900km, the risk of hitting something is pretty low; the chance of a close call is perhaps 1% per satellite-year, according to Brian Weeden of the Secure World Foundation, an American NGO. But satellites have been lost to such collisions; more satellites mean more such collisions; and collisions create yet more junk. Left to itself, the problem can only get worse. “It’s like climate change,” says Mr Weeden. “By the time it becomes a really big problem it may be too late to do anything about it.”
One answer is to make satellites more responsive and ensure that their operators are better informed. America’s armed forces use radar and telescopes to keep track of everything bigger than about 10cm across, and provide warnings when a bit of junk is going to come close to a functioning satellite. Analytic Graphics, a company that sells orbit-planning software, is moving towards offering a similar service that is better tailored to the needs of commercial-satellite operators. Like Earth-observation startups, but in reverse, it is using ever cheaper commercial technology to do what only governments did before. It is currently tracking about half as many objects as the US Air Force provides data on, but its capacities are fast increasing. It may soon offer its customers a better service.
The other answer is to clean things up—yet another job for robots. In 2017 an ESA spacecraft built by Surrey Satellite Technology, a pioneering smallsat company now owned by Airbus, will test its ability to ensnare a nearby cubesat in a net, reel it in and attach a “dragsail” to consign it to death by re-entry. It will also look at a technology for harpooning bits of junk. In the same year Astroscale, a Singapore-based startup, plans to launch a satellite to get a better measure of space junk. (Ground-based radar can see little that is less than 10cm across; the size of a cubesat was chosen in part because anything smaller would risk being invisible from Earth.) In 2018 Astroscale plans to try out a satellite with an adhesive patch to which any piece of junk can stick.
As it happens, very similar technologies might also be used for removing satellites that some people want in position and others do not. Anti-satellite (ASAT) weapons that target satellites have been developed by America, China and Russia, but if they were used it would be fairly obvious who the attacker was. A stealthy little satellite that could take them out from close by might work better; the victim might never know for sure if its satellite had been attacked or just broken down.
Here, too, better local information would help. America recently sent a pair of small satellites up to geostationary orbit to keep a much closer eye on both its own satellites and those of other countries. A second pair was launched on August 19th. But there are limits to this approach. As Doug Loverro, America’s Deputy Assistant Secretary of Defence for Space Policy, puts it, space is an environment which, at the moment, favours attack over defence.
Star wars writ small
That said, Mr Loverro identifies three ways of ensuring a continued satellite capability, all of which America is pursuing. One is active defence: measures that would make an ASAT attack more difficult. The second is resilience. Commercial service providers and America’s allies have more assets in space than ever before. The use of those capabilities, both on a routine basis—relying on commercial satellite links rather than bespoke military ones for many operations, for example—and as required in an emergency would make it harder for an adversary to deal a crippling blow with an ASAT strike. There would be just too many targets.
The third response is replenishment, which means having some back-up satellites safely on the ground, ready to be sent into space at very short notice. Another of DARPA’s space projects, XS-1, is challenging commercial teams to develop partially reusable launch systems that could get a couple of tonnes into orbit every day for ten days in a row. That could improve the economics of small satellites yet further, and allow the armed forces to feel much more certain of their ability to keep using space.
But there is another side to that use. Some systems designed to hit satellites might also be able to take out an intercontinental ballistic missile (ICBM) during the part of its trajectory when it is above the atmosphere. In the 1980s America’s “Star Wars” programme came up with the idea of so-called brilliant pebbles—thousands of small satellites that could spot and intercept rising ICBMs. The viability of such a system was hotly debated until the first Clinton administration pulled back from all space-based missile defences, on the argument that they would prove destabilising. That rendered the question moot.
The geopolitics of missile defence remain tricky. However, it is a sure bet that in a world where a smartphone has as much processing power as a Cray supercomputer had in the 1980s, and startups are launching satellites by the hundreds, a brilliant-pebble constellation is technologically a lot more plausible than it used to be, even though it might still prove politically unacceptable. Moreover, these capabilities, though at their most developed in America, are not unique to it. Warfare, both defensive and offensive, may yet prove to be an application where, as with communications, navigation and observation, space-based assets offer a regional or worldwide service that provides a distinct edge over surface-based alternatives.
There is no compelling need for people in space, but they will keep going anyway
SPACE need not necessarily become a battlefield, but the possibility is not without precedent. Military strategists have long known the value of taking the high ground, and the remarkable kinetic energy that comes with orbital velocities is a gift to weapon designers. The space race was a way to pursue the cold war by other means; its rockets were the children of V2s and the cousins of ICBMs. And its heroes were warriors, representing their nations in a strange new form of single combat.
The early astronauts had no real technical or operational purpose; their presence, like their combat, was symbolic. But the symbolism was central to the whole enterprise. Well before Sputnik, science fiction had established space travel as one of the fundamental metaphors for future transcendence, a rising above and beyond the limits of the human which would be meaningless if humans were not involved. Superpower competition made the same demand. If space was a race it had to have winners, and those winners had to be people, singular people whose achievements, made possible by the work of hundreds of thousands, would inspire not just their fellow citizens but the whole world looking up at them.
And inspire they did. A generation of children watched the Apollo landings and wondered what was coming next. That wondering went on for the next 40 years—wistful, fitful, sometimes angry, hungry for more. The “orphans of Apollo”, to borrow the title of a documentary film, watched the flying and failing of shuttles; the growth of the comsat industry; the Star Wars programme; the ISS; the roverisation of Mars: and none of it satisfied them. It was not simply that they wanted more astronauts. Astronauts kept flying almost all the time, because the powers that put them in space felt that giving up would entail a loss of prestige, a small but real diminishment. The orphans wanted those astronauts to do something, to take things further. Human space programmes stuck in low orbit with no higher purpose than self- perpetuation could not make good their loss.
Now some of those who wondered “what next?” are answering their own question. Many involved in the new generation of space ventures are motivated by more than profit. They want to extend humankind’s grasp and its sense of what it is. Some of this can be satisfied through the technologies of the small, the many and the robotic. These can do more than make money and serve humanity; they can inspire a wonder of their own (see Brain scan below). But for some the promise of space cannot be fulfilled just by hardware and imagery. And the radical improvements in earthly technology that have made ever more capable spacecraft possible have also made some of the technologically attuned entrepreneurs interested in space travel rich enough to direct their efforts beyond the near-term dictates of commerce.
Elon Musk is the foremost of these superpowered orphans. He has shown that he can drive the costs of space travel lower, possibly far lower, than a government bureaucracy can. For more than a decade he has talked about the need for a self-sufficient colony of people on Mars to ensure that the human race could survive an Earth-wrecking cataclysm. He has made it clear that his company, SpaceX, which recent investments have valued at $12 billionCK, will not become a publicly traded company before it is well on the way to getting that colony started. His purpose is not to maximise shareholder value but to make history.
At the end of September Mr Musk will reveal his road map for Mars colonisation at the International Astronautical Congress in Guadalajara, Mexico. A key part of the scheme is likely to be a new engine, the Raptor, far more powerful than the Falcon 9’s Merlin. Reusable rockets powered by clusters of Raptors will lift both Mars-bound spacecraft and their fuel to orbit. That fuel, unusually, will be liquid methane, which yields more energy per kilogram than the kerosene that Merlins use—and can quite easily be made from ice and carbon dioxide, both of which are available on Mars. Thus methane-powered spacecraft could not only get to the planet; they could also get back.
Billionaire boys’ club
Mr Musk is only one of a number of billionaires with a yen for space. In the 1996 Peter Diamandis, who is now co-chairman of an asteroid-mining startup, Planetary Resources, set up the Ansari X prize, a $10m reward for anyone who could build a vehicle able to lift three people higher than 100km—and thus, technically, into space—twice in two weeks. It was won in 2004 by SpaceShipOne, an experimental aircraft built by Scaled Composites, an outfit that excels at such things, and financed by Paul Allen, who founded Microsoft with Bill Gates. Mr Allen is now funding Stratolaunch, again in partnership with Scaled Composites. It aims to build the world’s biggest aircraft as a platform from which to launch satellites and, conceivably, people into space.
Richard Branson, a British businessman, founded Virgin Galactic to build a space-tourism business out of Scaled Composite’s X-prize-winning know-how. His SpaceShipTwo should let six paying customers fly into the blackness of space and experience zero gravity; about 700 people have paid deposits for tickets. Its development has been far slower than expected, and in 2014 an accident claimed one of its aircraft as well as the life of one of its pilots. But SpaceShipTwo should be back in the air soon. And Mr Branson can call on more than just his own wealth to cushion the blows of fate: Arbor Investments, an Abu Dhabi sovereign-wealth fund, has invested $380m in the venture.
Mr Branson’s main rival in the space-tourism business is Jeff Bezos of Amazon. New Shepard, the small reusable rocket built by Mr Bezos’s private company, Blue Origin, is capable of taking a reasonably roomy space capsule to the same sort of height as SpaceShipTwo. But Mr Bezos’s ambitions go far further. Blue Origin is building a new engine much larger than that used by New Shepard, similar in size to SpaceX’s Raptor. He is talking of selling this rocket to others, but doubtless also has plans for using it himself.
When Mr Bezos outlines his long-term vision for space, he conjures up dreams strongly influenced by ideas championed in the 1970s by Gerard K. O’Neill, a professor at Princeton. O’Neill imagined a future in which all the heavy industry on Earth would be transferred to orbit, there to be powered by unlimited and uninterrupted sunshine, some of which would be beamed down to Earth by huge solar-power satellites. Industry’s workers would live in vast space settlements; its raw materials would come from the Moon and the asteroids; its effluent would be swept away by the solar wind. Mr Bezos talks of a similar “great inversion” in which orbital space becomes a swarm of industrial satellites employing millions of people while the billions below restore Earth to a pristine patchwork of cities, parks and wilderness, receiving much of the hardware they need as industrial manna from heaven.
Blue Origin’s motto is “Gradatim Ferociter!” (Gradually ferocious!); it could be the tortoise to SpaceX’s hare. Rather than racing off to Mars, Mr Bezos is building up a sub-orbital space-tourism business first, then, presumably, a high-capacity, low-cost reusable launcher. From there, maybe, the assembly of an orbiting destination (another of Apollo’s wealthy orphans, Robert Bigelow, made his money in Las Vegas hotels, and longs to expand into orbit). Later perhaps some installations on the Moon, or on asteroids, to provision the guests? If Mr Bezos is willing to devote a significant fraction of what he has earned from Amazon to such things over the coming decades, his slow and steady approach might achieve a lot. Among other things, satisfied space tourists—well off, by definition—may swell the ranks of future space investors.
Such undertakings could outstrip, or absorb, national human space-flight programmes. China and Russia both aspire to putting people on the Moon. Europe’s space agency has similar plans, though it lacks a crewed spacecraft. America talks of Mars as its next destination, but seems in no hurry; and if Mr Musk’s big rockets head there, NASA may pivot back to the Moon. It is good for doing interesting science, and there are resources, too: bits of the Moon, like some asteroids, have ice. A largely scientific moonbase may become America’s default destination.
Such a moonbase might turn out significantly cheaper than the ISS—which is, at a cost of some $100 billion, the most expensive object humans have ever built. Just as post-shuttle NASA now uses contracts with private launchers like SpaceX to resupply the ISS, and will soon rely on them to get crew members there and back too, it will surely take a similar attitude to a future moonbase, contracting with Blue Origin, SpaceX, Boeing or some other company for the delivery of supplies and other services. That should keep costs down. So should the provision of robot assistance and the adaptation of other new space technologies to human needs. Companies such as Moon Express that are planning private missions to the Moon are not driven solely by the Google Lunar X prize, which promises $20m to a mission that meets certain objectives. They see themselves making money providing infrastructure for lunar science and, eventually, settlement.
Stretching the magic
None of this is yet certain. Mr Musk’s record is impressive, but he is trying to change the world not once but twice, both re-energising Earth with the solar-panel, battery and car business built around Tesla and providing an alternative to it with SpaceX. The magic could be spread too thin. So, perhaps, could the cash; both Tesla and SpaceX have in the past come within hours of bankruptcy, just as both have repeatedly failed to meet ambitious timetables. Even if Mr Musk can make spacecraft that get to Mars much more cheaply than previously thought, it is hard to see how they can be paid for with just part of the $5.5 billion launch business.
The powerful Raptor could be a risky beast, at least early on. Taking hundreds of people to Mars is a task of a different order from taking a handful to the ISS. And some aspects of Mars itself could scupper the plans. The planet’s surface is laced with poisonous perchlorates; its gravity may be too low for women to carry babies to term, or children to grow up healthy. Mr Bezos’s Earth-centric ideas may look more reasonable. But they require manufacturing industries that greatly benefit from being in space. And those industries have to consider people who need air, food and places to live as more desirable workers than tireless solar-powered robots specifically designed for vacuum and microgravity—unless people to want to do the work so much that they will pay for the privilege, or contrive to receive subsidies.
As far as a human presence goes, perhaps the most that space can hope for is to become a new Antarctic, protected from military expansion by treaty, suited only to research and tourism. It would not be a new Earth, or a greatly inverted one. But it would be an addition to the human realm well worth having; Antarctica, after all, is a wonderful thing. And the efforts of the orphans to create a yet greater future will, as long as there is no terrible loss of life, provide insights into what visionary drive, technological acumen and capital can achieve.
Humankind’s expansion into space may never meet with crowning glory on other planets or pass far beyond Earth’s orbit. But the years and decades to come will see something bolder and more inspirational than the staid circlings of those just past. And whatever the fate of the most ambitious ventures, the navigable, networked and knowable world that today’s satellites are creating, reinforcing and enriching will endure.
As well as the people mentioned in the text, the author would like to thank many others for their time and insight, including: Carolyn Belle, Grant Bonin, Mark Brender, Carissa Bryce Christensen, Dwayne Day, Jay Falker, Chester Gillmore, Scott Herman, Robert Hoyt, Stéphane Israël, Steve Jurczyk, John Karcz, Kevin Lausten, Creon Levit, John Logsdon, Pavel Machalek, Jonathan McDowell, Chris McKay, Brian O’Toole, Scott Pace, Andrew Petro, Peter Platzer, Will Pomerantz, Will Porteous, Bob Richards, Gordon Roesler, Rand Simberg, Sir Martin Sweeting, Rachel Villain, Micah Walter-Range, George Whitesides and Pete Worden