How does "laboratory X": factory crazy projects where you create a "new Google"

At 6:40 am sounds the horn three times. "Gas!" cries the man in a helmet and fluorescent vest. You hear a hissing noise, and the helium begins to flow. Of the cylinders, stacked logs near the truck, the gas flows through a series of hoses until you find yourself at a height of 14 meters, and then through a copper tube in the upper part of the plastic tube, hanging above the ground like the discarded skin of a snake, waiting for the candle.

So begins a Wednesday in late June in Winnemucca, a lonely mining town in Northern Nevada, which narrowly escaped oblivion, located on highway I-80. Among the few attractions in Winnemucca is a testing ground for ProjectLoon, ambitious project, launched in 2011, which should provide the Internet the places on earth where commonly hard to get from cell towers and wires. Instead of having to build and maintain terrestrial structure with a range of only several kilometers, Loonзапустит couple of balloons with antennas to a height of 18 000 meters, each of which will shed a good kilobytes on an area of about 3000 sq. km.

the

What to do in a secret laboratory Google?

Not so much in Northern Nevada what it was. But that's what it takes if you want to start a huge, glowing with Internet balloons up to 18 kilometres into the sky.

Loonпроходит testing in Winnemucca for the most part because of the clear sky and because there is a nearby airport that can accommodate private jets of Palo Alto. At the moment the team is experiencing a new iteration of their system of communication.

Half an hour Later the balloon is ready to go. It is held in place by a red horizontal bar and protected from winds by walls on three sides. At the command of the engineer, this whole structure is known as "Big bird", rotated 90 degrees to the left. Like Rafiki holds a newborn Simba in the opening scene of "King-lion", a few manipulators, and cranes pushed the balloon up. As he carries precious cargo — a range of solar panels, antennas and various electronics — for a second he freezes. And then leading upward, lifting at a speed of 300 meters per minute.

Despite the routine, it looks beautiful. "Never boring," Nick says Koli. "No-when".

When Kohli joined the nascent Project Loon in 2012, his job was to look at the world and collect the balloons fell from the Mojave desert, rural Brazil and to the coast of New Zealand. Loon was part of Google X, the offshoot of the search engine, whose task was the development of unusual projects that had to solve age-old problems in new ways. One such project was a self-governing automobiles. (In 2015, after the restructuring of Google, in which the parent company Alphabet, Google X is renamed X).

Kohli is not the normal "gurgler" — miraculously survived the Apocalypse. He could not go to medical school, so I studied on technique in the emergency Department — actually, this skill in combination with a pilot's license and eight years experience in search and rescue operations in the Sierra Nevada and helped him to get to Loon. This set of practical and operational skills made him one of those people who can not only sit at the computer and pushing buttons, like most in the Alphabet.

With the support and resources Alphabet, If (who now manages flight operations) observed, the development of Loon, first just looking at the balloons, flying hundreds of kilometers, and then launching them as if nothing had happened. This is another step on the path to implementing complex systems Loonв the future.

Today, the X takes a big step forward in this mission, declaring that the Loon "is available" — becomes an independent company under the umbrella of the Alphabet. Along with the Wing, another initiative X for the delivery of goods by the forces of the Autonomous drones, Loonначнет to build your own team, HRи PR. Leaders will receive badges as a CEO, and employees will participate in their share of the company if it succeeds. In the end, generate revenue and make a profit is as important as changing the world.

Loonи Wingдалеко not the first, projects that have received diplomas from the hands of X (employees actually receive these diplomas). In 2015, the creation of contact lenses to monitor glucose levels in the blood allowed a huge leap forward. And in December 2016 the breakthrough carried out self-driving cars named Waymo. The draft cybersecurity Chronicleполучил autonomy in January.

Dual issue Loonи Wing– big, ambitious projects – is critical for X, and perhaps marks the moment when a secret unit of the research and the design starts to perform well their mission. For a technology giant that has made billions of dollars on advertising, X is not just a trash unusual projects that do not fit into the corporate structure. This is an attempt to find a formula of release of the revolutionary products that not just appear on the screen, but interact with the physical world. Starting Loonи Wingв the world, X will have to see if she can sit through the new Google– and to put the Alphabet at the head of the industry, which doesn't even exist.

But the attempt Alphabet to create a new generation of companies-freaks raises two questions. Could this monster to grow exponentially? And who needs that?

Laboratory Loonв headquarters X in mountain view is literally Packed full of the results of generations of UPS and downs. Loonоснован on a simple idea – to replaceground-based cell towers-flying balloons – but this simple idea opens up a number of technical problems. In 2013, after a year of work, the balloons still had a bad habit to burst or fall to the ground in a few days. (They carried parachutes, which mitigated the impact of electronics on earth, as the command warned mission control center about the fall of the ball). Before the installation of the launch was similar to the one that was called "Big bird", the startup process was similar to how a bunch of kids trying to fly a kite, but a random gust of wind puts an end to the idea.

Now specially designed "mother of all boxes" keeps the system safe during transport. The key components are coming in a silver metallic box from polystyrene, which reflects sunlight and conserves heat. 30-meter-long flat scanner explores plastic bolotnoe on the subject of microscopic defects, which can reduce the survival of the balloon at the altitude of 18 kilometers from several months to a couple of days. Map monitors the flow of air over continents and oceans, and machine learning determines the flow of wind, you need to use to achieve a given goal. All of these tools allowed the team to launch a balloon every half hour and keep them in the air six months or more.

This kind of development is engaged in the laboratory of X. for six years, engineers and designers Loon and operators on restoration of the balloons did not worry about the funding streams, or looking for people who will squeeze out all of the HRили PR. They have access to the expert knowledge of machine learning Googleи to the "kitchen designers" — a huge space for prototyping any type of mechanical device, which can only dream of. They didn't have a detailed business plan, revenue or profit. They were allowed to try again and again, experiencing setbacks, each time gaining a little knowledge.

The Chief X chief Astro teller makes X place that will improve the world, but it does not hide the advantages for Alphabet, including the new revenue stream, strategic advantage and value to HR. And although he did not disclose the budgets or the costs of employees "moonlight" at the lab, he also makes clear that, regardless of how much, in your opinion, spends laboratory X, this is minor compared to the value it creates.

All commands in lab X, who are fighting for the creation of amazing projects, mistaken in its own way, but have the untouchable status. Welcome any ideas, offers new ways to solve old problems. Ideas come from everywhere. Some are born in the bright minds of the developers. Others throw up teller, Larry page or Sergey Brin. Staff X prowl in academic work and going to the conference in search of fertile ground in which to drop the seed of the idea. One of the unnamed projects was born during an interview with NPR scientist: someone in the X heard and offered to join in the work.

Wherever they were, most of the ideas first pass through the team evaluation. This small group meets once or twice a week, but not to protect ideas, but in order to destroy. "The first thing we ask: is it possible to implement this idea with the technology that will be available in the near future, and decide whether you need it part of the real problem?", says Phil Watson, the head of the team. Violation of the laws of physics – a definite refusal. "You have no idea how many perpetual motion was offered to us", he said.

At these meetings the smooth reflections in smoke filled hostels combined with the hard rigor of the thesis defense. The team considered the production of energy from avalanches (not feasible), the placement of copper rings around the North pole to generate electricity from Earth's magnetic field (too expensive) and construction of coastal ports to facilitate shipping logistics (nightmare controls). Discussed even work on a device to create invisibility. Technology seems to be allowed. "We said Yes, let's do it, it's cool, but no, we can't do that because he'll just problems and problem solving it will not provide," says Watson. "Tonkoprosejannogo".

Ideas that burst through the Bureau of primary grades, fall in the "foundry" where the head of the fledgling project is considering the issues and functions of the business, which creates; usually engineers don't work. At this stage, the Obi issues Felten that came in X in 2012, after working for years on products Googleв Europe.

In his first encounter with the teller Felten learned all that he was preparing a secret team X, including balloons and drones delivery. And she began to ask questions that expect from the man who runs the products. How legally justified the aerial balloon on the territory of another country? What are the issues in terms of safety? Will you cooperate with cell phone companies or compete with them?

"Astro looked at me and said, wow, actually no one thinks about these issues. There is only engineers and scientists who only think about how to make the balls fly."

"Casting" uses this intense interrogation to root out anything that can kill the project further downstream, before X will be pouring in piles of money and a lot of time. Take, for example, Foghorn, attempted to create a carbonless Xfuel from sea water. The technology was stunning, the problem to be solved is huge, but it's been two years and the team realized that there is no way to compete with gasoline on cost and technology were closer to the stage of research than development. X killed Foghorn was given to all the team bonuses and told them to work on something else. Ideally, the "Foundry" must guarantee that failing projects will be strangled as soon as possible.

The Premise is simple: the sooner you kill one idea, the faster your time and money on your next. Trying to change the world and create a huge new company means abandoning the traditional signs of progress. The only thing that will help to succeed is to discover what you are doing is most likely doomed. Because once you get rid of it, will only move forward.

Any project that hopes to qualify in the lab X needs to get to the center of the Venn diagram of three circles. It needs to solve a huge problem. It needs to provide a radical solution. It needs to include breakthrough technology.

This is a definition, which X uses to separate delivery drones from of invisibility cloaks did not exist in 2010, when the lab was just taking shape. It all began with an experiment: Larry page asked a Stanford computer science Professor Sebastian Trun to make him self-driving car. At the time of Troon knew about this technology more than any other: he helped Stanford win the Darpa Grand Challenge 2005, a race among Autonomous modes of transport on more than 200 miles across the Mojave desert near the Nevada. When Darpaобъявила about a new race in the 2007 Urban Challenge, the Agency has complicated the task, forcing the transport to wander layout of the city, where he had to follow the traffic rules, markings and Park. Stanford took second place, and Troon, who has worked with Google, joined the company full time and helped develop the Street View.

The Competition Darpaдоказали that cars can drive on their own, but more races were not held. American automakers are focusing on how to survive economic collapse, where we talk about the development of technologies that can destroy their business. Googleбыла company-producers, but she had mountains of cash, and it was clear that this idea has the potential to save lives, generating a stable flow of revenues and expand holdings Googleза the limits of mobile technology.

So Troon has quietly assembled a team and refused prominent academics in favor of a young team, veterans of Darpa Challenge, with less rigid notions of the impossible. Paige has set the team's own condition, noting 2.5 thousand kilometers of roads of California, which he wanted to see cars with self-control.

This transition to the physical world gave Google a fresh ground to work, since the main projects of the company outside of its Central business Gmail, Google Maps, and Google Books — cool stuff, but still software. But the arrogant Toyota Prius cutting yourself on the streets of mountain view, only added fuel to the fire of ones and zeros. But self-driving cars still appeared in the ranks of Google. The search for such complexity, serious problems require a certain infrastructure. Paige made the first Trun of the company "Director for the rest" that he was involved in everything that didn't interest investors. Since Troon focused on self-driving cars with its Director Astro teller took over the reins of the ship with no specific destination.

In one of the first conversations with Paige, the teller tried to clarify the situation. "I said we an incubator?", the teller says, leaning back in his chair. Not really. But not a research center. They created a new business, but it was not the most precise definition of their work.

Finally, the teller picked up the unexpected word. "We poke a finger at the sky?", he asked Paige. "Yes, indeed," answered Paige.

The Creation of research units to create innovative products is the core of companies whose value is inseparable from their ability to innovate. This tradition dates back to Bel lLabs, founded in 1925 AT&T and Western Electric. Created by many of the smartest scientists in the country, bell's lab is known for the creation of the transistor, the basis of modern electronics. She also helped in the development of the first lasers gave rise to the field of information theory, which has created the mathematical framework for understanding the transmission and processing of information. Eight Nobel prizes, three awards Turing, the Unix operating system and programming language C++ is also the merits of Bell Labs.

A Broad approach was key to the success of Bell Labs. There was no way to know how it will look next breakthrough, so it was not necessary to require a detailed plan of action. Its leaders are quite satisfied with "uncertainty in relation to the objectives," writes Jon Gertner. "The employees of Bell Labsбыли ready to explore anything remotely connected to human communications, whether they are tied to wires on the radio or on the recorded sounds, visual images".

And yet, Bell Labs worked according to certain rules. Her most valuable tool was fundamental research. Scientists Bell Labsпровели years, learning the basics of chemistry, physics, metallurgy, magnetism and other disciplines in search of what could be monetized. And although "human communication" is a broad definition, their work has not gone far beyondthat could improve business AT&T, which was the phone.

Silicon valley got its first major innovation lab at the Xerox research center in Palo Alto, where scientists stood out, not only their scientific achievements, but also the ability to take existing technologies and adapt them to new purposes that were never developed. PARCсоздала the first laser printer and Ethernetв 1970s – early 80s and laid the groundwork for modern computing, leading the shift from punch cards to distributed interactive machines – personal computers.

But in Silicon valley it is best to recall why Xerox chose this job. Laboratory literally invented graphical user interfaces, the icons on the screen that are controlled by a computer mouse, but Steve jobs brought them to the masses. Bosses Xerox has not pushed technology, they just didn't understand how that matters, says Henry Chesbro studying corporate innovation at the business school of Haas at the University of California at Berkeley. "Xerox was looking for things that are appropriate to the business model of copier and printer".

Providing its residents with almost unlimited rights and may not that unlimited funding, X thinks he can create products and services that other laboratories would never have discovered – or was discarded. It does not conduct fundamental research, relying instead on other agencies (mostly governmental and academic) to create instruments that she can imagine. She's not trying to gather the smartest people in the world in the four walls and see what they get. And, importantly, it intends to expand the range of business Alphabet and not to improve what is already there. With all these Nobel laureates, BellLabsимела value to their owners because they did phone calls better and cheaper. Shareholders Highpriestly PARC, because it brought them billions of dollars at the expense of the laser printer.

X will not make such mistakes, because its purpose is not to make search better. She must ensure that the mother ship – Alphabet– never stop expanding.

So, project X was not to be the first in the world to create self-driving cars or launching balloons Internet or deploy Autonomous drones. His real goal was to create units that can embody this business. The fetishization of failure and love of ideas that makes everyone look up, even if they fall down, they all serve this one purpose: if you don't make mistakes all the time and sometimes even stupid, you are not trying hard enough.

This is great for Alphabetи people who like the idea of self-driving cars (especially those who can't drive), or health tracking using non-invasive sensors, or cheeseburgers and toothpaste, which do not contribute to air emissions.

But Google, through Google, is already having a huge impact on our lives: how we communicate with each other, the way we get news as makes its way through the traffic jams. For most people life will be much harder without free email detailed maps and free access to almost unlimited information. X looking for an opportunity to multiply the impact of the withdrawal of it outside of the virtual world. Critics have always called Googleмонополистом. Now imagine that his dominance will go on cars, food, goods, physical things – in the same extent that we are dependent on Googleв the Internet. Are we willing to accept Googleв every aspect of your life?

Andre Prager enters the room, pushing a trolley with something similar to garbage. Basically it is cut from cardboard pieces, plus several plastic bags of sawdust. Prager has worked on Porsche engines. In his spare time he did the chain saw with the jet engine and the Vespa turbo. Now he's a mechanical engineer Wing, project delivery forces drones in X. In the truck – his failure.

Shortly after Wingзапустили in 2012, the team realized that the landing drones on the ground is wasting energy on greedy energy vertical takeoff. No, the aircraft instead have to hover and drop your cargo to earth – something like that.

The First attempt was the system based on the coil when the load is attached to the cord, which unfasten from the UAV. "The idea sounded great, because it just" says Prager. But in reality all were the elephant: how to wrap a package very difficult. Each package needed its own coil, and since the cord was in addition to the package, it was not the best user experience.

They experimented with less complicated mechanical systems, but nothing worked: either the packet did not unhook or hook broke, or something broke. "And then we said, what if you remove the moving part, in principle," said Trevor Shannon, another engineer from Australia, where experiencing Wing.

This thought brought them to the current design, which in size and shaped like a potato with a projection, which is superimposed on the packaging. It is easy to post manually, and when the precious cargo hits the ground, the weight of the hook naturally tightens it. Praeger like this modest device. "We measure our own success by how indifferent people in the form of this device".

The Purpose of Wing is to facilitate the delivery of for people to notlost time and reduce carbon emissions that are associated with moving things around in cars and vans. 2014 Wingзапускает pilot program across Australia, first in Queensland and then in Canberra, the capital. The company began to offer drone delivery on a ranch in remote areas and is now preparing to begin flights to the suburbs closer to the city.

A Real obstacle to scaling this initiative is not a delivery system and not the technology: batteries and control systems air navigation has made sufficient progress in recent years to control an Armada of drone delivery companies. The problem is how to do it safely, especially in crowded, controlled airspace over the United States and Europe.

Therefore, in 2015, the team has begun to establish a management system of an unmanned air traffic, which would connect all its planes, and would provide each drone is its own corridor from the place of delivery of cargo to its destination. The hardest part of this is not develop a system that will track the movement of aircraft, and to make all things in the sky worked on the same system. Wing works with the FAA and made part of their system open to all, so anyone can join.

The Most fun in this problem that X was not created to deal with those. This problem does not require the work of engineers or prototyping with the brainstorming. It requires careful relationship-building and communication with regulators and competitors. And if Wingне will make it all work, will be questioned long-term survival of the idea in General.

This means the change that needs to take the new company as Loon: the issue of Oznachaet a different attitude to failure. They become companies, which should succeed in the traditional sense, offering real services and attracting real customers with real money.

Loonналетала more than 18 million miles (30 million kilometers). It provided Internet Puerto Rico after hurricane Maria and Peru after devastating floods. Now the company will have to do something more complicated. "Time to fly from the nest," says alaster Westgarth, a veteran of the Telecom industry, which came in Poltora years ago and became CEOLoon. It was his task will be to establish agreements with providers around the world and to make the balloons brought together people around the world. It is important to stay bold, continue to innovate. At the same time, we can't risk what is already done.

X to observe the fledgling Loon and Wing as they enter the market, but soon will turn their attention to the new "shots from a cannon to the moon." It will be years before Internet balloons and delivery drones will take to the skies or crashing to the ground.

But in the Winnemucca launched the balloon gradually ascends. He goes to the desert and spend the night in the area before heading to Denver and then to Nebraska. If nick says there are already three balloons, which were launched a few months ago from Puerto Rico. They are approximately 0.005% of the way to the moon, which, in the opinion of all, not so far.