Google X Lab is a division of Google, overseen by Sergey Brin, one of the co-founders of Google. For an outside observer, this is a real factory of wonders where more than a hundred promising projects are currently being developed, from an unmanned vehicle to a self-learning neural network. But few people think that this laboratory is already an amazing and unparalleled experiment in itself. In this article we will talk about one of the projects of this secret laboratory – Project Loon.
Google X Lab
Нужно четко понимать, что Google X не является обычным R&D центром. Rather, this laboratory can be compared with the fantastic Academy of the Limits of Knowledge from the novel 'Andromeda Nebula', which worked on problems only barely outlined by science. Even if Google X is a little more practical. In Google itself, the work of the laboratory is often compared to shots from a cannon at the moon – the chance of success is scanty, but if successful, a sensation is guaranteed.
There are no forbidden topics for Google X employees. It may well be that behind the closed doors of the laboratory, they are seriously working on the teleportation of people or an anti-gravitational vehicle. What any other commercial organization would consider outright insanity is a common reality at Google X.
For example, it is known for sure that the Google laboratory at one time was developing a space elevator project that would deliver people into orbit. True, the project was soon frozen for an indefinite period – for the implementation of this idea, materials would be required that simply do not exist yet. Yes, and on skateboards flying above the ground, work was also going on at one time – unfortunately, the project was also suspended so far, since the cost of producing such devices, even according to the first estimates, turned out to be astronomical.
There are only three main conditions that a project must meet. First, it must be some kind of global task, one way or another affecting millions of people. Secondly, the proposed solution should be radically new, head and shoulders above all existing analogues. And thirdly, the project should be based on technologies that really exist today, it should not be some abstract idea, the implementation of which the author himself does not know how to approach.
Someone might say that Google's management is just hovering in the clouds and investing money in obviously useless projects. It is one thing to invest in a specific development, when the approximate cost of work, the time of its self-sufficiency, and so on are known. But what's the point in funding crazy, but very expensive ideas, most of which are simply not feasible in practice?
However, there is a reason for this. The technique has long ago reached the level when it is simply impossible to improve one or another solution 'on the knee'. Often the struggle is for fractions of a percentage of efficiency, and each next step is not only worked out in detail by many narrow specialists, but also requires the introduction of new technologies and the modernization of production lines. To create in a small laboratory a car surpassing any one in its performance characteristics Mercedes is not just a fantasy, it is a utopia. Google simply does not have the necessary specialists in this profile, nor experience in this area.
Photo from core77.com
But a futuristic self-driving car is a completely different matter. No one will watch how much fuel it consumes, how ergonomic its body design is and what top speed it can reach on the race track. Moreover, it is almost guaranteed to be inferior in these parameters to existing machines – but who will pay attention to this, when in any case it simply has no analogues?
In addition, the fact that the laboratory's management never rejects this or that idea on the fly does not mean that its development will automatically be given the green light. To assess the prospects of new projects in the laboratory, there is a separate Rapid Evaluator team, which the employees themselves prefer to call Rapid Eval. Its leader is a man with the speaking name Rich DeVol, who previously worked in Apple. The task of the group is not only to assess the possible risks, but also to try to immediately 'drown' the idea, throwing tricky questions at the author. Or even completely demanding to assemble at least a conditionally working prototype that demonstrates the main theses of the speaker. Is it cruel? Maybe. But if the project is not initially viable, then it is better to find out about it immediately, and not after many years of fruitless (but very expensive) work.
It would seem that with his profession, Rich DeVol is simply obliged to become a cynic, mortally tired of the fantasies of half-insane inventors with burning eyes. In fact, everything is strictly the opposite – he loves crazy ideas like no other. But, unlike many dreamers, he prefers to embody crazy dreams in life, and not just reason on abstract pseudo-scientific topics. The famous line from the Soviet air march 'we were born to make a fairy tale come true' could well become not only the motto of Google X, but also the Rapid Evaluator group.
Photo from fastcompany.com
However, when it comes to shots at the moon, the overwhelming majority of them are doomed to failure from the very beginning. И в этом еще одно большое отличие Google X от R&D центров остальных компаний. In science, a negative result is also a result. And often a scientist is simply obliged to undertake a bunch of different studies and projects, even realizing in advance that 99% of them will not bring any practical benefit. After all, there is no way that would make it possible to make the opening of the century strictly on schedule and with minimal expenditure of time, money and effort.
The trouble is that commercial companies have never been happy with this approach. The goal of any company is to make a profit, and in this regard, financing crazy projects is not much better than playing at a casino. For one lucky person who has hit the jackpot, there will be several thousand poor people who have gone bankrupt to the last penny. If we are to invest in research, then only having a clear idea of what benefit (read – profit) may be in them for the company in the future. And trying to control this process as much as possible.
Google has much less stringent requirements for its experimental laboratory. From the very beginning, everyone understands perfectly well that one cannot do without a series of failures and that there may not be any concrete return on these investments for many years to come. But if by the will of fate you have at your disposal a large cannon, then it is already better to shoot from it at the moon than at the sparrows. And Google, at the moment, clearly has funds for additional research.
Photo from core77.com
Failures are the working days of the laboratory, and we can only guess how many falls for each step forward. They also know how to benefit from them, analyzing the path traveled, drawing conclusions and adding the knowledge gained to the common piggy bank. Moreover, any at least more or less successful project will be immediately removed from under the wing of Google X, and work on it will be entrusted to a completely different department. And the laboratory staff will again try to find the way for the next technical revolution.
And it is for this reason that Google X projects are usually in no hurry to show the press. Failures may be necessary, but there are few people who want to demonstrate them to others. Yes, and for the project itself, this can go sideways, by the time its commercial version is released, an impatient buyer may well 'burn out'. An example is the same Google Glass that was shown to the general public too early. Many have already recorded these 'smart' glasses as failures of the company, although work on them has actually just begun. And whether the company will now be able to re-interest buyers with a toy that has already bored them is an open question.
It is also worth noting that projects that initially cause a 'wow-effect' to the public do not always have any perspective. Conversely, what ordinary people might not understand can often be of much greater benefit than any space elevators or flying skateboards. One of such necessary, but not attracting special attention projects, will be discussed below.
Project Loon
We used to think that the Internet is available everywhere and everywhere. Cable networks, Wi-Fi, mobile and satellite Internet – there are many options. And if we are complaining about something, it is only about the high cost of data transfer in roaming or the low speed of communication. Even if, due to some global accident, the lights are turned off in the entire area, the Internet is likely to be – the base stations of mobile operators have backup power supplies, and instead of a computer, it is quite possible to use a tablet or smartphone.
In fact, in most of our world, Internet access does not even smell. And for a huge number of people, this is as fantastic as a time machine or a flight to Mars. It is simply impossible to find an answer to an important question, get expert advice, or write an e-mail to relatives. But in some cases it is a matter of life and death.
[iframe url = '// www.youtube.com/embed/m96tYpEk1Ao' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ']
They say that the Internet has everything. It may very well be – the only problem is that this 'everything' is by no means for everyone. For many people, Internet access could be the only and unique chance to get a normal education – but alas, in places where there are no normal schools, there is usually no Internet. The same applies to communication – a situation when for years it is impossible to contact loved ones and find out if they are alive at all, for many people the only reality they know.
Project Loon's goal is to provide Internet access to all people, even in the most remote corners of the globe. And not just come up with some abstract concept, but bring it to life with your own hands. Even by Google X standards, this project is considered completely insane and it is not surprising that Rich DeVol himself is one of its leaders.
In the video below, Cliff L. Biffle, in charge of all flying systems at Project Loon, shares his memories of starting the project.
[iframe url = '// www.youtube.com/embed/5mD6m3V02n8 ″ width =' 640 ″ height = '360 ″ scrolling =' no 'frameborder =' 0 ″ marginheight = '0']
The Rapid Evaluator team then crowded around a white plastic board on which a droid flying in a balloon was drawn with a marker. “Complete madness,” smiled Rich DeVol. “Let's try to prove that it can actually work.”
Most of the necessary equipment was bought at the nearest store, some were ordered online. A radio module, a pair of antennas, a simple mini-computer, a cardboard box as a case, a cheap weather balloon and some adhesive tape – that's all, the first prototype is ready.
The very first launch showed that the problems would start not at all where they had originally expected. The radio itself worked without problems, but a bunch of other things went awry. And even just following the balloon and then finding the box with the electronic 'stuffing' that fell to the ground was not easy. But the main thing is that the project was started. Even if the following prototypes looked even more bizarre, resembling giant jellyfish more than balloons.
One of the first problems that the development team had to solve was the navigation problem. A balloon is not a full-fledged flying drone and, at first glance, it is not at all easy to control it. But in fact, the answer to this question has long been known. Strong winds walk at high altitudes and by lowering and raising the ball up and down, you can catch the air flow blowing in the right direction. Thus, the operator of the ball was simply required to control the altitude at which he was flying. For these purposes, the device was equipped with a special mechanism that could pump in or, on the contrary, release excess air from the ball.
[iframe url = '// www.youtube.com/embed/E4D0c6yQftc' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ']
The second global problem faced by the project participants is the need to provide constant access to the Internet, and not a short-term connection with a ball flying by accidentally.
Dan Piponi became interested in this question, who, like Rich DeVol, is one of the regular members of the Rapid Evaluator group. From his point of view, this problem was interesting in itself, and although no one specifically asked for help, he decided to think about how to solve it.
[iframe url = '// www.youtube.com/embed/mjyLynnQuC4 ″ width =' 640 ″ height = '360 ″ scrolling =' no 'frameborder =' 0 ″ marginheight = '0']
To do this, he had to create a special emulator, in which it would be possible to quickly simulate different positions of the balls, taking into account their speed, height, headwind and other parameters. And then – to develop an algorithm for the operation of balls, in which each ball independently determined its place in the scheme based on the position of its neighbors.
It turned out to be no less difficult to find out where and at what height the wind is blowing, which will carry the ball in the right direction. This problem was partially solved with the help of NOAA (National Oceanic and Atmospheric Administration) forecasts. However, their data on the situation in the stratosphere were insufficiently detailed. The thing is that information about the speed and direction of wind movement is collected using special 'weather' balloons, which simply burst at such a height.
[iframe url = '// www.youtube.com/embed/gaY_1_2UBas' width =' 640 ″ height = '360 ″ scrolling =' no 'frameborder =' 0 ″ marginheight = '0']
I had to think about how to collect the missing data on my own. There was nothing fundamentally difficult in this either – its own 'air fleet' for Project Loon will in any case be equipped with GPS modules that allow tracking the speed and direction of movement of each ball. This means that the air currents that carry them. It remains only to learn how to organize this data and use it for their own purposes.
For Project Loon, this was a turning point – a beautiful, but vaguely practical idea, for the first time began to take on quite visible outlines.
However, the team still had a tremendous amount of work to do and faced a host of very different challenges. The fact is that although the components of the ball for Project Loon are quite ordinary by themselves, they have to work in the most extreme conditions.
Take at least a battery. According to the idea, in the daytime the ball should not only work from solar energy, but also charge the built-in battery with it, with which it will survive the night. One problem is that the temperature in the stratosphere is very low; the battery is not designed for such cold at all.
[iframe url = '// www.youtube.com/embed/EPDzpxFkTvM' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ']
The idea of a multi-layer protective casing that allows you to keep warm without wasting battery power was borrowed from NASA. Despite the fact that he showed himself well already on the first prototypes, work on it continues to this day. After all, there is no limit to perfection, and ensuring the safety of the battery charge is one of the highest priority tasks of Project Loon engineers.
The durability of the ball itself is no less important. As planned by the developers, the lifetime of each ball should be about 100 days, which will allow it to make three circles around the Earth. The main danger here is micro-damage to the shell, which will be very difficult to avoid. The point is in strong temperature drops and in periodic inflation and blowing of air. Experiments in this direction are still underway (or, to be more precise, at the time of the video recording). To do this, prototypes of balloons from different types of materials are launched into the air, each of which is equipped with a special sensor that helps to determine the location of air leaks.
[iframe url = '// www.youtube.com/embed/zULbu1xfdwE' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ']
Putting such a ball on the ground is also a whole thing. He should not fall on someone's head, he should not get caught on the plane and should not be at the bottom of the nearest lake. The entire flight of prototypes must be strictly controlled from the first to the last minute of the flight.
[iframe url = '// www.youtube.com/embed/zpzwQajQxZ4 ″ width =' 640 ″ height = '360 ″ scrolling =' no 'frameborder =' 0 ″ marginheight = '0']
The balloon is landed in uninhabited places, and during landing, you have to not only monitor changes in air flows, but sometimes also interact with the local ATC service (Air traffic control). After a successful landing, a special team flies out for the device, which determines the exact position of the ball and collects all available equipment from the landing site.
An equally important part of the project is the antenna, which is located on the ground. As mentioned above, the appearance of the Project Loon balls is quite impressive. But none of the users would be able to evaluate it at such a height anyway. But the design of the radio antenna that is installed on their home is a completely different matter. She will always be in front of the user's eyes and should look as stylish as possible. After a short discussion, it was decided to try it out in the style of a balloon.
[iframe url = '// www.youtube.com/embed/2D4tMf1w2xQ' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ']
According to the designers, no amount of 3D modeling will ever give a complete picture of the object. After all, it is important not only how the antenna looks by itself, but also how beautiful it will look on the roof of the house. Google X solved this problem in a radical way – they built a small wall with a roof visor and checked how the prototype antenna looks on it. And at the same time, we tested the installation process itself, trying to understand how it can be simplified as much as possible.
The antenna device is demonstrated in the video below. And it is not at all as simple as it seems at first glance. After all, a balloon, unlike a satellite, rotates slightly around its axis, and an ordinary transmitter is not suitable for receiving such a signal – communication will periodically disappear.
[iframe url = '// www.youtube.com/embed/gpu0VZFKm5U' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ' ]
[iframe url = '// www.youtube.com/embed/iFUjp495gUA' width = '640 ″ height =' 360 ″ scrolling = 'no' frameborder = '0 ″ marginheight =' 0 ']
Outcome
For more information on the Google X lab, check out this article from Fastcompany. One of the most interesting moments described in the article is the end of the workshop, which was attended by the leaders of the laboratory, CFO Patrick Picchette and one of the co-founders of Google Sergey Brin. The meeting was not easy – at that moment one of the key groups of Google X was going through difficult times, the development of the project (wasn't it Google Glass?) Was almost at a dead end and additional funding was required to continue the work. Both Brin and Picchette reacted to this news calmly enough, promising to settle all financial issues. And then we went out into the street and approached the rope, which was pre-stretched between two trees at about waist level. And for more than twenty minutes they tried to walk on it in turn, constantly stumbling and falling, but not giving up their attempts.
This moment perfectly characterizes the whole essence of the Google Lab. When you take on a complex and unfamiliar business, you cannot do without misses and falls. Google executives are not afraid to fall in front of all employees, gradually learning to walk the tightrope. And all those who work at Google X always know that falling is not ashamed. It’s a shame to give up and after another failure and stay lying.
