Living Machines
2014-07-11 | Text: Aydar Fahrutdinov | Photo ©: archive of "Living machines" project; Malcolm Burrows /;;; | 4689

Nature is an ingenious designer and unsurpassed inventor. It is hardly surprising that much of what we use in our everyday lives arose thanks to the study of living things. The Eiffel Tower was constructed with the human shinbone as its prototype; in dentistry crown buildup is based on the growth principle of sea sponges… The list is probably endless. Nor is it surprising that today’s robots are also made according to the laws of nature, their external appearances and internal structures copying those of animals and man. 

Nature, however, is limited in its possibilities, since it can only use protein structures as building materials for living things. This is why solutions worked out by evolution are optimized for protein polymers. Can man, using a wider array of materials, come up with completely different principles for the organization of life – principles that are possibly more perfect than those offered by nature? Both individual enthusiasts and entire groups of specialists the world over are wrestling with this question. One of these groups is based in Russia. We would like to introduce you to Dmitriy Davidovich, head of the project Zhivye Mashiny [Russian for “Living Machines”].

Nature really is a brilliant engineer. Many mechanisms generally considered to only be within man’s power to create have actually existed for millions of years. Even the gear train, used everywhere in mechanics, is one of nature’s creations. Illustration: a magnified image of a mesh gear, found on each rear pair of legs of insect larvae in the family Issidae. The size of the mechanism is only 400 micrometers, but ideally synchronized movement of the legs, achieved by means of small teeth, allows the insect to jump at a speed of 4 meters a second.

– Dmitriy, please, start by telling us about yourself. Who are you – a scientist, a businessman, or an engineer?

I’m a scientist who wants to see the fruits of his scientific research turned into commercial products. I don’t want to be just a businessman – the laurels of Warren Buffet do not tempt me. On the one hand, I sincerely believe that with the help of science one can make outstanding, progressive, and commercially profitable creations. On the other hand, I play the role of a merchant who wants to sell the things he creates on the market under the heading “planet Earth”.

– What’s special about the area of robotics that you work in?

The more of these objects, which I call robots, I construct, the more I understand that they’re not robots. Silicon Valley, which has practically monopolized the ideology on this concept, defines a robot as an absolutely soulless piece of metal programmed to fit human needs. It shouldn’t possess any questionable features, it has to impeccably serve mankind and carry out the functions assigned to it. But when you have the task of making the objects independent, able to move within their own logic and occasionally absolutely ignore the interests of man – when we give this construct some sort of right to its own existence, in other words – then it is hard to call that a robot in the conventional, stereotypical sense. I was forced to give this phenomenon a different name – an “alive machine.”

Naturally, living machines are my work. In a way it is alternative robot construction. You thought that a robot has to be an android? No! Here are robots that in no way resemble humans. You thought that a robot has to be made of iron? No! Here are robots made exclusively from soft materials. You thought that a robot has to be stuffed with electronics? No! Here are robots in which mechanics takes center stage, and where everything begins to spin and whirl due only to the movement of wind or water. The main quality of living machines is their ability to being moving even when only two components have been put together. And these movements actually do give the impression that we have living mechanisms.

– Is anyone else working on this?

My role model is the great Dutch engineer and mathematician Theo Jansen, who makes amazing figures that are moved by the wind. I know him – I’ve been a guest of his and have had the honor of speaking with him. If you believe that the epitome of engineering is the creation of objects that do not require electricity and programming, while at the same time behaving extremely technologically and with cybernetic expedience, even with cybernetic intelligence, then that’s exactly what Theo does. It’s a very thorough solution to the problem of making a machine live – a tangle of levers, rods, and tubes comes together in coordinated movement.

However, I understand that when viewed through the lens of current expectations this is already too little – it’s like perfecting piston-engined aircraft in the era of jet aviation. That’s why to me Theo Jansen is a great craftsmen and designer, but at the same time an artist who is outmoded in his tastes. He is trying to solve problems that he set before himself in his youth, thirty years ago, when first professor Prigogine, then Doctor Haken revealed to the world the science of synergetics, showing that from a series of uncoordinated systems a miracle can arise – something completely new.

Dutch artist and kinetic sculptor Theo Jansen creates self-propelled structures that externally resemble the skeletons of animals. For material Theo uses plastic piping, bottles, and other materials that come to hand. Wind sets the sculptures in motion. The sculptures can also “breathe.” When the wind is strong enough, the excess air pressure is stored away in special reservoirs, so that even when it’s becalmed the structure does not stop, instead falling back on its reservoirs. And vice versa – in dangerously high winds the construction drops anchor.  

– In this case what is the next step in development?

The building-up of synergy – this is exactly the type of research I carry out. For me what’s important is a coordination of elements that makes some sort of governing nucleus that determines the activity of the entire system. I can show this a little more simply by showing you one of my models – a pneumatically self-regulating pump. At its heart is a balloon connected to a compressor. If you turn on the compressor, the balloon will expand until it just about bursts. But this state of nearly bursting, by means of a mechanism built into the balloon, makes the compressor turn off. The balloon deflates until it is practically completely deflated. The condition of almost complete deflated, though, turns the compressor back on. The balloon acts as if it “doesn’t want to” burst and “doesn’t want to” deflate, and to this end it brings in different gadgets that are capable of reacting to these states. So, again and again, moving from minimum to maximum and from maximum to minimum, this balloon becomes the heart of the mechanism, at the same time setting in motion a world of elements that simultaneously perform an incredible number of tasks – they turn on, shift, whirl, turn, switch places. One of the tasks I put before myself in this model is to learn how to construct things so that each detail had up to a hundred functions – so that it simultaneously twirled, and scuttled, and turned, moved itself, and stopped… In the future this entire thing will be able to unhook itself from the compressor, move around, find another one and hook itself up to it. And this will also be behavior “with elements of reasoning”, although initially it was born of a “desire for a certain emptiness”: on the one hand, the desire to not expand into infinity (burst), and on the other, to not become nothing (deflate).

Dmitriy Davidovich

And in this structure there are no electronics at all. It’s what I call 3D mechanics, since each element of the structure plays its own important spatial role within the entire technological ensemble. And all of this is controlled by a certain unseen hand. Sooner or later I will hide, or “dissolve”, this balloon within the mechanism, and no one will even be able to figure out why the structure moves, why it acts like a rational organism. And here’s the surprising thing – in a living cell something similar happens. It’s also 3D mechanics, only on a molecular level, where the driving force is provided by “molecular gears”, twirling in an enormous number of interwoven biochemical reactions.

We’ll bring the machine to life one step at a time, algorithm after algorithm – there are something on the order of forty experiments of various types.  Many questions go into these algorithms: questions of utilization (death), self-preservation, self-awareness, formation of a special mechanical sense perception (sensors), even of emotions and future transformations (development). Of course, an entire set of construction experiments is necessary, and it is these experiments that excite our minds with the interesting results we might get.  You can’t exclude the element of chance – at a certain point I will have to relate not to the reality I create, but to the reality that the machine creates for itself.  Exactly like a child who for a time is raised by its parents, but then begins to amaze them with the decisions it makes for itself. This, by the way, is also one of the forty algorithms.

– Where do you get the ideas for your designs?

I painstakingly study the history of technology, devoting a great deal of time to it. I have an enormous library of books on the history of technology, and not an evening passes that I don’t read something from it. I could tell you a lot, beginning with the automatons of Antiquity up to contemporary mechatronics. I know a thing or two about the evolution of many types of technology (weapons, aircraft, androids, adding machines, clock mechanisms, mechanical men and animals) and specific devices (levers, cranes, presses, printing equipment, steam, diesel, and electric engines), the work of brilliant masters of the past (Hero of Alexandria, Ctesibius, Archimedes, Leonardo da Vinci), and all sorts of clever things they made, like trap labyrinths with secret doors. I am constantly systematizing this wide spectrum of things, getting down to the very roots – what, when, for what purpose and how it appeared, with research into dates and names, into the interrelations between these masters, the instruments and techniques they used. It’s in these areas that ideas shine out.

History repeats itself. I can point out that the type of mechanics that I work with has been used more than once – beginning with Archimedes, Hero of Alexandria, and a series of Arabic and Chinese inventors. There are a large number of ancient mechanisms so complicated that we can’t reproduce them today, and if we can, then only for very large sums of money. For example, a piano-playing automaton, consisting of more than 2,500 ingeniously assembled mechanical components and created in the seventeenth century by Henri and Pierre Jaquet-Droz, can today only be duplicated by two companies. They would nearly three years and 1.5 to 2 million dollars on the production of this delight. 

The prototypes for the automaton used in Martin Scorcese’s film Hugo were the automatons of the watchmakers Henri and Pierre Jaquet-Droz. The most complicated – the Calligrapher – consists of 6,000 parts. Using a selection of forty letters, it can write out a text encoded on a codex, choosing letters one after another. The boy uses a goose feather which he periodically dips in an inkwell, then shakes so as to avoid inkblots. The automaton’s eyes follow the text, and its head turns toward the inkwell when it dips the feather in it. Although the Calligrapher was created in 1772, it is still in working condition. 

Before the nineteenth century the creation of mechanical “toys” was very advanced, and but was crowded out by the efforts of Faraday, Davy, Maxwell, Tesla, Edison. Mechanics went over into the fields of the arms industry, the internal combustion engine, electrical energy and engineering tools.

An attempt to revive the situation was made by the first generation of cyberneticists – they created a grand revolution in the rethinking of models of life. But yet more children of progress – this time rocket science, microelectronics, and digital technology – once again drew all the attention on themselves. The work begun by the mechanists of the past was never finished, even though the potential is enormous. For this reason this work is both a subject of study and a source of new ideas.

– And why are you doing all of this?

I understand that I could just run my own business – the commercial assembly and sale of 3D printers; I could just make money while amazing things are happening in science and technology. And later, sitting in front of the television, I would be excited about how this bright new epoch has arrived, that I’m living in a golden era. It’s an interesting prospect, a pleasant one, but it would mean that I have absolutely nothing to do with it – that I don’t make any person effort to become a partaker in those incredible things that are going on around me. And how could I do that when I have ideas, the opportunity to challenge preconceptions and attempt to propose alternative options that are already before us?

What I am working on, and what our entire laboratory is ultimately striving towards, is the creation of a parallel mechanical civilization that can evolve, that possesses a fate all its own, one not at all related to the fate that arose from the primordial soup of organic substances, as it’s described in Oparin’s theory. My team doesn’t have the luxury of those billion years that nature had to work with, but we have an “accelerator” called “reason” that has been given to her. We try to take full advantage of reason so that we can make a desperate attempt to “pump up” our pieces of metal and initiate evolutionary processes of self-development in mechanical systems; turn on the growth areas, so that right before our very eyes the machines began to make themselves, and even remake each other into conceivable and inconceivable models that transform, develop, compete with one another for certain resources and all the while fear death, although admittedly not death in the moral-psychological sense that we humans are used to. We have yet to find an interpretation of the concept of “death” on a mechanical level.

Anticipating what is probably your next question, I’ll say right away: I am categorically opposed to taking upon oneself the role of a Supreme Creator. I am a man of deep faith. I believe that the Maker united the many hearts of my ancestors so that eventually I too would appear on this earth. And created a mass of variations so that I would become the way I am, and so I would think the way I do.  I don’t even consider myself the inventor of these devices – I look for, derive, and give structure to what already exists. Kurchatov, after all, didn’t create the atom, but he was able to unleash the power hidden inside it.

– How do you put together a team to solve problems like these?

I formed the team over a period of two years in terms of “the collective intelligence of interrelation and mutual support.” With this team everything gets done quickly, everything is coordinated and done in order. There is a set of rules for the creative mechanist. There are no stupid workers, and everyone’s in his own place and knows what he needs to do. They keep their minds on the job. Each person has his own specialization: research, modeling, mock-up, assembly, debugging, testing, checking, introduction of parallel processes... Center stage are two people – the technical moderator, who manages the process as a whole, has a good feel for the production process, keeps track of the time spent making the next model and answers for the final result of the process, and the coordinator – the person who quickly gets in contact, sets up additional cooperative ties, who when the need arises contacts, writes, calls, negotiates, corrects the sequence of cooperation – in general, holds things together.

The team of the project "Living machines"

When new employees come to me, I throw them into the assembly of some complicated unit right away, but in a way that they don’t mess things up, and after an hour or two I already know what will come out of them. It becomes completely clear who has a technical vein, who has the marks of an inventor, and who doesn’t have either. By the way, there is an angle of our work tied to the humanities – technical journalism, which takes the form of keeping a blog.

The surprising thing is that big-shot specialists come to us and can’t handle simple tasks, even though they are able to do a lot of things. Only Bauman Institute graduates have never disappointed me – they’re competent to an unbelievably high level. If you give them a task, sometimes they’ll say: “All right, that’s impossible, of course, but we’ll come up with a way of doing it.”  And they do!  And at times like that I really get excited – I feel that these are the people I can move mountains with. My only caveat is that I’m talking about people of my generation (the ‘80s). Young people these days aren’t quite as capable.

– How much more time is needed to completely realize your concept, and where are those key points through which your developments can be capitalized?

I would like the “symphony of parallel mechanical life” to be written in the next 10 years, since I long to see the fruits of my labor in my lifetime. In general I like to compare my work with film production, and, as a person who has “shot the movie”, I want to take a seat in the cinema myself and look at my finished product. At the same time I’m a realist and very well understand that you can’t sell this grandiose mechanical civilization offhand. The public is merciless; it has its own well-formed stereotypes and everyday preferences which have broken more than a few innovators. For that reason we have to make this happen by means of a constantly-developing series, like a TV series, dividing it into seasons and showing each episode at a strictly-determined time that is convenient to the market. An episode is a production line of five or six units serving as a source of income, which will then be reinvested into the most interesting thing – what the viewer will see next. And well, to be honest, I don’t have the money right now to go start on the high-rise Institute of Mechanical Civilization.

I will begin my first season this year. It will be a series of toy construction sets with clever and diverting mechanics. Using these sets, it will be possible to assemble relatively simple devices that you can put on a table to liven it up. For example, different types of nightlights that swivel, blink, and follow movement;  mechanical flowers that react to human activity; different kinds of breathing, moving, snorting creatures that are interesting to observe… It’s a small segment of that fathomless and rapidly growing sphere that’s called “smart house” or “smart office.” I think that in the future attaching all kinds of entertaining devices to the smart house platform will become the norm.

– What niche market are you focusing on? Toy construction sets for children?

While making the first episode of my film, I’m trying to stay within the realm of fun and useful devices that the customer can assemble himself. My consumer is a person who needs to enliven – in the literal sense of that word – the environment in which he lives and works.

As for children, today’s child has more preferences in high technology than even the most advanced adult. But the spectrum of preferences of children today is very selective. When I show some incredible gadget to preschool-age children, it doesn’t really get them interested – but then you have a lot of kids from ten to fifteen who get indescribably excited and want it all for themselves. That’s why for me teenagers, without a doubt, are the guides in the market.

On the whole progress always needed and always needs good promotion:  if you take part in exhibitions, organize showrooms, shoot interesting videos and put them on YouTube, then the group of interested parties, without a doubt, will form.

– How are you preparing to organize production, and on what scale do you hope to plan production?

Organizing production is probably what I’m able to do better than anything else. In my life there were many projects that required me to solve systems organization issues. Machinery, people, internal ergonomics, internal logistics, the manufacturing cycle with its timetable, rhythms, schedule, flowcharts with time assignments, bookkeeping – it’s a world of its own. And it’s my inheritance – my father managed production, my grandfather managed production. Even my great-grandfather managed production. From the age of 5 I was with them at the factory because there was nowhere else to put me.

I learned a lot from the Chinese, and I’m still learning. Since 2010 I’ve gotten rid of all my false stereotypes and false knowledge from textbooks about how a company should be organized. I forgot everything and went to work at a factory – the Chinese and I rented a few machine shops, and from the inside I saw how they work and organize things.

Now, thanks to my knowledge, contacts, and agreements in China, I can put any test model into serial production fairly quickly. Approximately half a year is all I need to launch serial production of an item. We will turn out from 2,000 to 10,000 units – if we have the corresponding financing, of course. Five or six items is enough for me – that’s my current organizational resource, or the limit to my personal maneuverability, you might say. Without a group of assistants I can’t do more. For things to take off, it’s necessary to put out a collection of 6-10 types of toy construction sets; otherwise no market will take me.

– Aren’t you afraid that the Chinese will copy your developments and fill the market with their own product – especially seeing as how you plan to organize production on their territory?

I’m not afraid. Of course, they could put together their own toy construction set, but they wouldn’t understand what step to take next. A construction set is only one little square in a comprehensive picture that is the result of lengthy, painstaking research work. Over the course of each season we will unveil a few of those algorithms I talked about earlier, and that will enable us to always be ahead of our imitators.

– How do you plan to stay afloat until production is set up?

With a business that I’ve been working on for a year now – the sale of electronic devices, and of specialized component parts on the Do-It-Yourself (DIY) market in the sphere of machine tools with CNC (Computer Numerical Control). Right now I have something on the order of 70 clients who need component parts for 3D printers, milling machines, lathes, plate bending-and-rolling machines, laser metalcutters with CNC, all sorts of controllers, electronics. Many people know that I can get any hard-to-find part for machinery. Rack gear, cogged pulley, cogged belt – by all means, flange bearings, ball-and-screw rails, slide rails – no problem, a stepper motor of any size and driver’s for it – that’s easy. I make money on this, which I then put towards my research on the creation of a civilization of living machines that is parallel to human civilization.

Default AJAX