“Already now you can build a robotic world of the future. But the economy is not ready "

This year, the first issue of the scientific journal Soft Robotics proclaimed a new stage in the development of robotics: robots become soft.

One of the main articles of the room, for example, talks about a robot fish developed by MIT specialists. This is an autonomous organism that can make quick movements with a body, in a split second, avoiding a meeting with an obstacle and changing its direction through convulsive compression. In the left and right halves of the tail are channels along which gas - carbon dioxide from the “stomach” enters there. Giving different portions of gas into the tail, the robot can wag the tail at an angle of up to 100 °. This is controlled by a computer system with which a robot can make 30 different maneuvers.

But the robot fish is only the most famous example. There are already many projects related to robots made of soft structures, and soon it will become even more.

Barry Trimmer , one of the main ideologists of the new direction of robotics, the founder of Soft Robotics and a professor at the University of Tafts, told Slon that the world of soft robots today and whether this technology is able to become a generally accepted standard.

- Why did they talk about soft robots right now? Several projects related to this technology, which has not yet become the mainstream, but seems to be close to this.

- In the past two years, it became clear that outside the robotic conveyor systems, military robots and all kinds of gadget robots, this area was at a dead end. There are very developed areas with great achievements - mechanics, dynamic systems, machine learning, artificial intelligence. In this part, robotics made a huge jerk in just a few years.

But the question is in the forms, namely the forms will determine the content. We, users, are used to it, and the developer community is used to the fact that robots are complex metal structures. This approach was justified, it was explained by the desire for reliability, and most importantly - it assumed traditional types of production, the approaches of work with the material, with its processing, its integration with electronic systems. There are already ready -made methods in the automotive industry and electronics, which were easily transferred to the new sphere. This meant that the invented can be created and collected here and now.

The main problem remains what experts call the “level of close contact”. No matter how various sensors are developed, the robot becomes another participant in the movement, any - human, automobile, natural, etc. This strong reliable design itself protects it, but is dangerous for the surrounding world.

If we talk about personal robots, in houses and offices it would be necessary to create its own system of movement, similar to the one that has been built on the roads, with its rules, the rights of participants, analogues of traffic lights, etc. For decades.

In addition, the traditional form is ineffective: it is simply impossible to make its energy by the power of energy extremely clever and adaptive; Whatever innovative materials are used in addition to it, the machine will still be limited in motion. The search for new, non -traditional forms of robotics, which would not be just an additional approach, but a completely new technology, and led to what several research groups are doing now in different universities. The first results are already destroying stereotypes. But the ultimate task is not just to create an interesting alternative, but to make this approach dominant in research, development and production.

- Judging by the first projects, this is a movement towards biomimimicry, which many and have long been said: robotic structures will mimic under a living environment.

- I myself am a biologist, I am engaged in neuromechanics and biometria, so for me this direction of thought is completely natural. But if we talk about biomimics in robotics, then it will look a little different than everyone imagines. This will not be blindly copying animal images and their dynamic qualities. Although there are experiments with modeling images - for example, Italians are now trying to make soft robotic tentacles.

It is rather about internal filling and animal logic: about the structure of muscles, the musculoskeletal apparatus, the principles of conservation of energy and the principles of action. The advanced robots today can choose the ball, but for this they need to first find it, calculate the exact distance to this ball, and then calculate where to put the hand and which effort to put. This is a period of long fluctuations, calculations, where to put your leg, how to move your hand. In a sense, this is the principle of human action, which is why some robots are humanoid, and not just because they have a head, eyes and limbs.

Animals come differently, their body allows them to make more relaxed, quick and clear movements. At the level of robots, this means a rejection of the idea that it is necessary to control all variable values.

-Can we talk about some practical use of soft robots now? Is this so far exclusively experimental technologies or are there already ready-made areas of application, perhaps even business models?

- Soft robots have more space for maneuver - in the broadest sense. They are more effective in an unpredictable and uncontrolled environment, whether it is a house, a natural disaster area, an underwater cave or a children's room. The absence of springs and hinges allows them to avoid the same thoroughly calculated accuracy of movements as the “hard” robots. Hence the spheres of possible use. True, now it is important not so much to think over practical models as to test the technology, to understand the limit of its possibility. The finished technology will easily attract developers.

If we recall the most large -scale and successful experiments, I would mention research in Harvard, where they are developing soft robotic limbs for exocostyums, jumping legs with acrobatic functions, etc.

There is also a laboratory of soft robotics in the Woster Polytechnic Institute, where scientists create a murmur robot. There is Otherlab: it builds inflatable robots, including hospitals, they are completely safe for patients, have psychotherapeutic functions and can transport people on themselves.

You can also fantasize: for example, the problem has not yet been solved by the complete automation of egg packaging - robots of hard structures cannot cope with this delicate task. A mass of soft robots in the shape of a grid can be filled with water and discarded in the area of ​​shipwreck, where they will swim to the survivors and save them.

There are space studies where the weight factor is very important, and soft robots can be very useful. There is a market for prostheses where the softening of structures while maintaining functionality is a number one task. There is a problem of personal homemade robots, the same vacuum cleaners that are dangerous for pets, young children and old people in the current rigid structure.

- It turns out that mostly these are niche decisions. Soft technologies will not become universal, but will only be a traditional alternative?

- In the case of robots, the versatility is generally greatly exaggerated. Now is the transition period when everyone competes in wit. This is right - after all, without constant search and sorting out interesting ideas, a golden middle will not be found - but a little naive. I see this: niches will be determined by money, and not our ideas about the beautiful.

Popular conversations about cross-hairdressers, nun robots and policemen robots are rather meaningless. Forget about small business and public sector (with the exception of the armed forces).

For various reasons, robots will not become a mass phenomenon there in the coming decades. Small enterprises live at the expense of individual work, there are robots, no matter how perfect they are, contradict the very principle of a small margin. The public sector, if governments and parliaments do not lose their minds, will deliberately minimize the introduction of robots in order to maintain the correct balance of employment and social calm.

Robots will appear where there is free money, a large private market and complicated tasks that require new technological solutions. And soft robots due to a new approach to plastic may be in demand.

- You mentioned military developments, which often ahead of the speed and innovation private initiatives. Are the military soft robots are also interested in?

- I think the army will play a significant role, but not decisive. If we talk about military robots, then in their case there are two priorities - the speed of movement and the ability to overcome complex obstacles. The whole industry is currently working on this, and judging by the last prototypes, it has achieved significant results. Another priority - superpowered software is a task that concerns not only robots, American developments are engaged in it as part of other programs.

As a result, thanks to the army, we can get more flexible, dexterous and safe structures. But their improvement and search for specialization will be engaged in companies operating for the private sector.

- Few people argue that this concept itself is very attractive. Most doubts are due to how to ensure a combination of softness and sufficient energy intensity. For example, a carbon dioxide robot is enough only for several dozen sudden movements, which is only interesting as an attraction, no more.

- What we already see is pneumatics, hydraulics and various experiments with mechanisms for the supply of compressed air and fluid under pressure - this is the first level. The second level is elastic materials that could expand and contract, reacting to stress. The third is special metal alloys that would change the shape at a certain temperature or spring, responding to light.

A certain ultimate task is to create universal artificial muscle tissue and artificial fat that would have it. They are already working on this in different laboratories.

- Another frequent reproach is primitivization of functions. Soft robots will not be smart due to the structure, artificial intelligence will simply not fit into them.

- This is a mistake in priorities.

Artificial intelligence is a super -task with superpowers. But if we do not build a body that can determine the function of intelligence, all these developments will be meaningless.

Moreover, I will say, perhaps, an unexpected thing: the trouble is that the intellect of robots is now better developed than the body. A team of programmers is able to create a complex decision -making system and teach it to change these decisions depending on different introductory ones. But the mechanical body of robots in its current form can often not be able to comply with these decisions.

-But there is already a rather large robotics market, there have been developments that already have demand for, there are business models that are interesting to investors. And this is the market precisely traditional, “hard”, robots.

- When analyzing the robot market, it is important to divide it into two parts, otherwise all subsequent constructions will be too general. There is a segment of robotic gadgets, often directly related to the global mobile ecosystem. Partly it is at the junction of the Internet of things. If you introduce important criteria, namely the long -term autonomy of complex actions, the lack of manual control and the multi -level system of movement and functions, this border will become more obvious. All the many complex domestic gadgets can be ranked here, controlled through smartphones with a single-line scenario of actions: from robots-perclescouses, superpowers relative to the market as a whole, to television and various toys robots. This part of the market already has its own sales, they successfully attract venture and crowdfunding money and gradually leave the niches of marginal demand.

The second segment is complex multi -level robots. They will also definitely intersect with the mobile market, but this is a different level of self -sufficiency, a different level of complexity, a different cost and potentially a different level of influence on production, the national economy and household economics.

At the technological level, these are also incomparable things: robotic gadgets today require marketing innovations today, this is the approaches and ideas market. Here, due to understandable demand, there is already a reasonable investor. In multi -level robots, a market idea at this stage is secondary. These are complex technologies that still require adaptation to existing realities.

- The robot market is often compared with the markets of smartphones and computers, conducting analogies of the speed of their introduction into everyday life. Do they seem correct to you? Smartphones and PCs were nevertheless created on the basis of traditional materials, there are also experiments, but they cannot be called revolutionary.

- I am not ready to analyze purely market circumstances, although it seems to me that common analogies in this case are not fully correct. Both desktop computers and smartphones were evolutionary devices - they seriously affected consumer practices, but were part of a long chain of transformation of life and business, stretched for a long period of time. Robots and their derivatives are technologies of very compact time. What we mean by them now appeared as a potential market product in just a few years, and this product has only a few years to become a market. If it doesn’t work out, something else will appear, but later, and this will not be evolution.

If we talk about the speed of implementation of technology, about their readiness, then they are already ready. And they are ready not only in DARPA laboratories, but also in private developers laboratories. Already now you can build a robotic world of the future. But “smart money” and the economy as a whole are not ready for such a development of events, which is another difference from the same smartphones. It requires very large one -time investments.

- At the same time, there are relatively few venture money in the robot market. What are they waiting for?

- And the venture is not able to form such markets, he can only support them afloat. Obviously, we are at the stage of revising the existing principles of the economy. The banking system is changing, approaches to industry lending and long money are changing. It is clear that fundamentally new introductory in such conditions are not considered. Simply put, everyone is not up to it.

In addition, the dynamics of technology development paradoxically negatively affects their financing.

There is a common erroneous feeling that this is not the limit, that it will be even cooler further: today there are soft technologies, and tomorrow robots will grow like flowers in pots.

The expectation of some absolutely recognized level of achievements that will not leave any doubts is not the best atmosphere for large investments.

Being now inside this world, I realize that this is a mistake, but I also perfectly understand those who are not ready for bold steps right now. The effectiveness of a robotic economy is still a subject for discussions. And while leading economies do not come to some consensus on this subject, the market will be in the experimental stage.