Alexey I. Borovkov, Professor, Vice rector for advanced projects, Head of Engineering Center (CompMechLab®) and Center of Excellence in New Manufacturing Technologies of the National Technology Initiative (NTI) in Peter the Great St.Petersburg Polytechnic University (SPbPU).
Under his leadership, the Center has evolved into one of the most successful innovative organizations in Russia with branches nationwide, bringing together intellectual initiative of young scientists and engineers. In 2017, SPbPU spin-out hi-tech engineering company CompMechLab® was awarded the Industria National Prize of the Russian Federation.
In his exclusive interview to “New Defense Order. Strategy”, Mr. Borovkov spoke about the trend in the field of digitization and industry digital transformation that should be developed today to enable the necessary innovative breakthrough, and what has been already done to this end by developers of the Engineering Center.
– The Industria National Prize of the Russian Federation was first awarded in 2014. It is annually contested by hundreds of Russian innovative developments and the prize goes to the single winner firm. The competition is strong. What was the cause of CompMechLab® Engineering Center winning the prize in 2017?
– We have been awarded, speaking in modern vernacular, for creating CML-Bench™ digital platform intended for building Digital Twins. In fact, it is digitization in its most knowledge-intensive form expressed in digital transformation of business processes and business models, ensuring creation of a globally competitive and relevant product in a highly constrained environment in terms of finance, time, production capacity, technology, nature etc.
– Mentioning competitive products, do you mean software, i.e. the written platform itself, or the product that has been or will be developed using this platform?
– Certainly, my initial implication was the knowledge-intensive and hi-tech product. It can be cars, engines, aircraft, helicopters etc. We have built a cutting-edge toolbox – a computing technology, more specifically – a digital platform that brings together or integrates dozens of best-in-class global level technologies.
– Are you intending to export the technology?
– No, not in the near future. We are not even taking it out to the broad market so far. Why? This unique Russian development is currently unprecedented worldwide. CML-Bench™ digital platform gives us a critical edge in the process of new product development. Having this competitive advantage, who would transfer or sell it? Long story short, we are talking about a business model which would help maintain and, above all, increase the competitive edge.
– Do you mean that the platform gives a benefit of innovation and it would be irrational to export it?
– Yes, that’s exactly what I’m saying. It is the basic critical technology. We distinguish three stages (a triad) in the technological development: technological breakthrough, technological breakaway, and technological superiority. Today we focus on solidifying our technological breakthrough through building the technological breakaway and technological superiority, and understanding the momentum of our growth that must keep pace with or somewhat exceed the global momentum. Finally, we need to make an estimation of the resources we are going to need for systematic work.
– What stage of this marketing strategy are you presently at? And what should be done in order to create this technological edge and obtain a substantial economical result on a national scale?
– Speaking of the marketing strategy, another triad comes into play: Hi-Tech & PR & Marketing. It is a common belief that PR is a part of marketing but I make a distinction between the two. The example of Elon Musk’s car launched into space perfectly illustrates how this triad works. When we discussed this case in Roscosmos, their specialists said: “Come on, he only launched it once, and we did it multiple times”! Yes, we did launch, but blocks of concrete, and without bothering to make it public. Musk launched an e-vehicle into space and the news reached everyone, even those who didn’t have any previous knowledge about Musk or interest in space. The outcome was investors queueing to offer money and scientists/development engineers begging to cooperate. Hi-Tech & PR & Marketing in the right time and place are sometimes indispensable.
– Neglectful attitude to PR is quite common in our people.
– In advanced technology marketplace, and this is our case, all new products follow the Gartner Hype Cycle (Fig. 1). Our pilot project was in 2015, when we published our book Bionic Design and thereby introduced both a new term and a comprehensive technology to create products with conceptually new features. In our understanding, Bionic Design is additive technology plus optimization technology stack, e.g. multi-criterion, multi-parameter, multi-disciplinary optimization and, of course, topological and topographic optimization. A set of solutions generally termed “Bionic Design” provides an opportunity to effectively apply new or existing technology. The term “Bionic Design” instantly migrated into documents, e.g. tender documentation for engineering centers.
– The term’s authorship belongs to you together with its “stuffing”? I mean the “Digital Twins” and “Digital Shadows” that followed…
– Yes, if it is about hi-tech industry, mathematical simulation-based development, optimization technology and additive manufacturing. But for today, “Bionic Design” is past the entire Gartner Cycle. Now this complex technology develops simultaneously with the related Generative Design technology and by design, as a component, fits into the more powerful Digital Twin technology.
This solution is essential for development of any knowledge-intensive or high-tech product for any hi-tech manufacturing area, and is therefore especially relevant for the Defense Industry.
Importantly, the Digital Twin should not be confused with the “Digital Shadow” of the production process or product in operation. Digital Shadow has long been adopted in a range of industries worldwide, usually to monitor technical condition of an object in operation, but it can do little to prevent emergency as its only function is “memory”, i.e. the ability to “remember” what has already happened. Due to multiple sensors installed on (in) the physical object, it can warn about “normal wear” of units and assemblies, but will not predict an emergency condition which often results from a fateful combination of many causes.
Use of “Digital Twin” technology encircles the entire product life cycle, from design (including avant-project, research and development) through manufacturing to after-sales service / maintenance and repairs. Even more importantly, the system is trainable and represents an ever-expanding storage of data, solutions, and knowledge. It is critical and irreplaceable for the Defense Industry as it can significantly decrease the amount of expensive testbed and full-scale testing, due to the enormous number of virtual testing performed at the stage of development using mathematical simulations with a high level of approximation to real data, products, physical and mechanical processes. Tests are often performed on custom-made virtual testbeds / test ranges.
– Do you have any ready use cases? What point of the Hype Cycle is this technology in?
– Yes, we do. But people often hurry in an effort to turn the technology into a state standard, which is premature at this stage, because the critical mass of knowledgeable and trained engineers with sufficient background in physics and math and prepared to use it to develop, e.g. armaments, military and special purpose equipment, has not yet emerged.
– How do we bring forward this critical mass?
– Only through solving real-time challenges, e.g. pilot projects that end with obtaining measurable results which were previously thought of as unachievable.
Imagine the Polytech’s Engineering Center in the eyes of old-school engineers from manufacturing and Design Bureaus? “Here are the bunch of youngsters trying to teach us!” This limits our possible entry point to pilot projects where we can solve “unsolvable problems”. These are the problems that were for some reason previously deemed “insurmountable”. Take, for example, the renown Cortege project (with FSUE NAMI as the general contractor) implemented with deep involvement of CompMechLab® Engineering Center specialists. In early 2014, the automotive industry unanimously declared that the pressing deadlines (before inauguration of the new president) and with the allocated resources (around 12 billion rubles) made it impossible to achieve the result comparable to the world’s best-in-class counterparts. I recall famous industrial designer Vladimir Pirozhkov saying in early 2015 that such projects were “completely adventurous” and their successful implementation was possible only at a high level of domestic automotive industry, with the budget of 120+ billion rubles instead of the available 12 billion ($2 billion, he said).
Nevertheless, the project was successfully implemented within the budget ten times less than the amount deemed acceptable by the global automotive industry. In this project we strongly relied on the Digital Twin development technology based on the matrix of claims / target indicators and resource constraints amounting to 125 thousand. This is what advanced digital design and engineering, or, more precisely, the Smart Design & Engineering, looks like…
But it is followed by another critical question: is our technology in demand with the state defense order contractors? Their usual way is to say, “Give us one billion (or multiple billions), bring forward the deadline, and we will do what you need.” And here we are, saying, “No, our advanced ‘Digital Twin’ technology can do it cheaper and faster”. We are only being rational. But in today’s economical reality, we act to shrink the contractors’ budgets…
This kind of a situation (“cheaper and faster development”) is typical for civil hi-tech industry which has no other way but to navigate global markets through making competitive products in less time and for less money…
– It is an enticing perspective to reduce time and cost from concept to implementation. Government and private business can benefit from it significantly, but state defense order contractors may lose income.
– Another very important thing. Along with many countries, Russia is currently in the technological “Death Valley”. The hi-tech products become increasingly more sophisticated year by year due to the global competition and emerging new players like Elon Musk in automotive industry and rocket engineering. Boundaries between traditional industries are being erased, new markets pop up. There is less and less time to solve these more complex problems that we are facing, or we simply don’t have it. We also have less and less money to solve these increasingly complex problems. Clearly, we will never have the financial opportunities of our predecessors such as Kurchatov and Khariton, Korolyov and Glushko, Tupolev and Ilyushin, Mil and Kamov, and, last but not least, Kalashnikov.
To sum up, the problems are becoming more complex, and there is less time and money to solve them. This is our common reality, or, metaphorically speaking, the “Death Valley”.
One may conclude that obviously, traditional approaches will not work for crossing the Death Valley, being more time- and money-consuming when applied to solving more complex problems. Therefore, the solution will only come from using new, advanced, best-in-class global level technology. The technological breakthrough that the president has been mentioning, without properly being heard should be built systematically and purposefully. In our development we have reached a certain stage which is challenged by a problem requiring a way more advanced level; we are facing a leap in complexity. Overcoming it with less time and money would require the matching kind of leap – a technological breakthrough.
– Can Digital Twins provide this opportunity?
– Yes. Take Figure 2. We are looking at a hypothetical but highly realistic situation. Our level of development is described by value RL(t) (Russian Level), t – time, and our development pace is, for the sake of simplicity, shown as angle А. Worldwide development level is WL(t) (World Level), and its development pace is represented by angle B. From comparison of curves in Fig. 2 we can see that WL(t) > RL(t) at any t, as B > A, with breakaway WL(t) – RL(t) > 0, moreover, for any t2 > t1 we have WL(t2) – RL(t2) > WL(t1) – RL(t1).
What does this mean? It means that “whatever we do, our disadvantage increases day after day”, despite our development with “A” pace that we keep proudly reporting to our seniors (without mentioning, of course, the “B” pace of our opponents).
So again, we are facing our hereditary “catch up and outdo” objective. How, in solving this kind of a complex task, we can draw on the Digital Twins technology? Let us assume that global level WL(t0 = 0) = WL0 is a Benchmark, and take it as a landmark that we need to outreach. We are at a level RL(t0 = 0) = RL0, meaning that we only have a short time (t* – t0) (preferably months, not years!) to “jump up” to the World Level DT(t*) = DT* ~ WL(t*) = WL* – and it is only possible using the DT (Digital Twin) technology. Next we will have to set the pace for DT* development according to angle D – this kind of improvement will be ensured by rational balancing of increasing numbers of characteristics in the matrix of claims / target indicators and resource constraints, as well as use of optimization and advanced manufacturing technology.
Since D > B > A, by the time t** we will “catch up with” the World Level DT(t**) = DT** = WL(t**) = WL**, however at times t > t** we will have: DT(t) > WL(t), t > t**, i.e. a family of Digital Twins surpassing the global level: DT(t**) < DT (t3) < DT (t4) < DT (t5) < … , t** < t3 < t4 < t5 < … . The product we take to the market will be at the level of Digital Twin DT(t3), which is, notably, above the World Level. DT(t4), DT(t5), … are Digital Twins “in an ambush” that can be released to the market at any time, whenever the global marketplace requires it.
Experts raised by Klaus Martin Schwab, pioneer of the Fourth Industrial Revolution, termed our approach Double Leap Frogging. In fact, this figure shows the technological breakthrough (t < t**), technological breakaway (t > t**), and technological superiority (t >> t**). Evidence that proves it might have already happened. The Cortege project, for instance.
– Could you explain how this technology works and describe the advantages of its implementation, including cost impact?
– It is important to understand that a “Digital Twin” is not a digital copy of a real object; that would be too simplified an idea. Come to think of it, it is the real object that is a “copy” of a full-on Digital Twin. General Electric has taken its digital copies (replicas) of real objects earliest and farthest along the way. In 2017 they declared, “By year end we will have a million of Digital Twins.” That said, they termed “Digital Twins” the part we refer to as “Digital Shadow”. They only studied the construct’s behavior in operation through “stuffing” it with sensors and capturing huge data arrays. Clearly, this approach only works well if the process follows the standard pattern or has minor deviations that unfold relatively slowly.
A “Digital Shadow” cannot predict an emergency situation because “nobody taught it” to do so, and it can be trained only on real cases. Conversely, a “Digital Twin” can simulate probable critical situations suggested by experts, feed them into his “digital brain” and perform multiple virtual tests.
Basically, it is important to realize that the “Digital Shadow” only has its memory function and is not designed to predict emergencies, whereas the “Digital Twin” possesses a unique predictive potential.
– You mean that the Digital Twin can “anticipate”, “imagine” and simulate a situation? How?
– Yes, it is capable of predicting, and that’s the main thing. Consider an example. Assume that at a development stage of a complex object there is a chief designer who has, say, five senior designers, each of them responsible for designing a certain plant unit. Imagine that each of the senior designers can “hold in sight” up to one hundred target indicators: five multiplied by one hundred is five hundred, meaning that the design encompasses five hundred target indicators.
In fact, a real object in all modes of operation can be described by about five thousand target indicators and characteristics. The next step in development and elaboration is fifty thousand, since a Digital Twin keeps adding target indicators, parameters and characteristics, in the process of both development and operation. Once the addition reaches an order of magnitude, its effect becomes obvious as the improvement in design quality is also substantial. This is the proverbial technological breakthrough…
For information: In project Cortege for development of a single modular platform the tests concurrently took into consideration one hundred and twenty five thousand target indicators and constraints – this is what was implemented on the basis of CML-Bench™ digital platform and got us the Industria National Prize of the Russian Federation in 2017. It was a world record back then. Notably, one year before, in June 2016, Aurus (a sedan from project Cortege at that time) got the top grade in the first attempt during tests at an independent test range in Berlin for passive safety in a frontal crash test. Never before the indigenous automotive industry got a top grade; in its previous days it “got nothing”. I gave you an example of successful implementation of the Double Leap Frogging strategy in a particular project.
From the cost perspective, the leap effect is reached through reducing the amount of testbed and full-scale tests, in some instances – by an order of magnitude or more. However, don’t forget that the amount of virtual tests, including those performed on custom-made virtual testbeds/test ranges, grows by two or more orders of magnitude. Looking at the marginality curve of any product of global leading manufacturers, you can see that the most marginal stages are design and aftersales service/maintenance and repair. Manufacturing is not marginal any more; global leaders have long made it lean, in many instances digital, and are now making it intelligent, “smart”.
Thus the Digital Twins of a product, manufacturing steps and modes of operation, in combination with the Digital Shadows of the products in operation, help considerably decrease costs at all life cycle stages of an innovative product’s development and maintenance, first of all, at the stage of design.
– Does the successful experience with Aurus from Cortege project mean that the breakthrough technology already exists and it can and should be upscaled?
– Yes. And this is how it looks. The design process generates the so-called Type 1 Digital Twin of a product, while the manufacturing stage brings forth Type 2 Digital Twin of the production process operations. Once the product and manufacturing process Digital Twins merge together, the resultant Digital Twin becomes a Smart Digital Twin which possesses with a “genetic memory” about how and in what sequence it was manufactured. Next, the Smart Digital Twin informs us of critical zones and characteristics at the stage of operation, answering critical questions of Where and What to measure. This, in turn, helps build the Smart Digital Shadow and substantially reduce the amount of “garbage data” generated at the stage of operation. On the one hand, the Smart Digital Shadow makes our Twin smarter with each operation stage; on the other hand, it enables three important kinds of feedback – about the stages of operation, manufacturing and, most importantly, design and creation of new, more competitive products in the shortest timespan.
– Ok, but what’s next? How do you factor in external influences and foresee changes in the operational environment?
– Exposure of a material to aggressive environments can result in degradation of its properties. Nuclear power industry leads the research in this area. Speaking about other hi-tech industries, their learnings, one should mention cross-industry technology transfer performed under the umbrella of the trans-disciplinary approach. We simultaneously cooperate with ten hi-tech industries, and each of these industries faces its own challenges and addresses them, “outrunning” the other industries in this respect. As a result, superiority and learnings in development of Digital Twins in one of the industries can be extended to another.
– Do you mean that soon the Digital Twin development technology is going to become a blueprint for all design work in a variety of industries?
– Yes, but implementation can be challenging. People that never had such opportunities are “suddenly” given “unlimited opportunities” and of course they struggle to believe it. They are used to not having a tool to ensure quick results; in their experience, no matter what design path they opt for, there is an inevitable risk of missing the deadline. They also tend to choose just one path, one design trajectory, which they deem the most correct.
On the other hand, in our experience, the design process is inherently non-stationary and non-linear, with many bifurcation points; which means that we launch 10, 20 or 30 design trajectories without knowing which will lead us to success. Imagine that designers are sitting together brainstorming and generating ideas, one of them saying “I would do it this way” and another, “And I suggest this thing.” Traditionally, the chief designer would respond, “Stop that. Cease talking. This is what we are going to do”. And this cuts off all other options.
Certainly, some of these neglected options were better, but it couldn’t be verified at that point; conversely, the advanced design methods suggest “design beyond chief designer’s experience and intuition”, and also design tools fundamentally different from what the industry employs today.
– Does it mean that the Digital Twin development process expands the opportunity space?
– Yes, this is exactly what happens, and it is a critical change. We can launch several design trajectories. Afterwards, in a while, we can bring the experts together again to review the intermediate results and see which idea, hypothesis, or concept had worked, controlling enormous amounts of target indicators and limitations along the way. In fact, we reinvent the process of design to include “continuous quality gate”.
In the opportunity space, there is no limit…
Interview by Alexandra Grigorenko
Translated by EGO Translating
©New defence order. Strategy №5 (64) 2020