AI calculates the best option
Improving the efficiency or the search for the guilty?
Alexey, what, in your opinion, prevents Russian companies from introducing digital products and services and transforming their business processes to a greater extent?
I have even made a report on this topic at the Agency for Strategic Initiatives. There are at least five constraints. Firstly, not everyone wants to be transparent. After all, any digitalization is accompanied by data collection, building any analytical models, publishing errors and various kinds of ineffective actions of the companies that took it. Look at the current cross-section of the ownership of manufacturing enterprises in Russia: up to 80% are state corporations. And every year their share is growing. Let’s take top management and mid-level managers: they are undoubtedly high-class pros, but they are not able to resist the rules of monopoly. If they give the green light to transparency of business processes, then everyone will see the inefficiency of not only the company, but also specific managers. In this case, state structures often act clumsily – they search for the perpetrators, dismiss them, but do not solve the essence of the problem. After analyzing the data for the past year, our software showed that the company could fulfill the annual plan by 20-30% faster, and in fact at the end of the year, everyone worked around the clock with high costs to fulfill the plan. However, private companies do not always want to be transparent.
Secondly, Russian companies do not like or, rather, are afraid to take risks. They are afraid to work with startups and diversify their products and services. After all, most of those who work in the field of digitalization are somehow new to the business. Thirdly, certain barriers arise because of an unwillingness to change. Digitalization for the sake of digitalization is not necessary for anyone, it involves changing the business model and processes, which requires a lot of work, and very often meets serious resistance on the ground. Fourthly, not every company manager can think strategically and look a step ahead. Like, why do we need upgrade the plant, if everything works, everything is in order with profitability? But in fact, it may turn out that there are reserves inside the company to earn many times more. But finding these reserves is not easy without special predictive analysis and analytics tools. The fifth factor lies in the low level of education of both top managers and mid-level specialists.
Rostec claims that only 10% of Russian manufacturing enterprises have a digital technology strategy, and only 15% of factories are ready for digital transformation. Do you agree with this assessment of the situation?
Recently, BFG Group has been trying to work with private companies where there is a possibility of direct access to the owner. As for Rostec`s data, in my opinion, 10% is a very optimistic figure. While working with various companies, I have not seen a ready-made digital development strategy, at best there were some elements of it. The development strategy is not just a road map with a list of instructions for action, the development strategy should be aimed at cardiac improvements. That is, the company must understand which transforming is worth for the business. For example, speed up production processes several times or make them 20% cheaper. Industry 4.0 technologies are only meaning to achieve a global goal. If we give people a microscope, and they try to turn self-tapping screws with it, and then they are surprised that they fail, then the question is first to the performers, but not to the tools.
Tell us about successful examples of your product implementation. How does your digital twin concept differ from similar solutions from major vendors like Dassault Systemes, Siemens, and others?
We have implemented a number of successful cases. For example, for "BTK Textile", the foundry of PJSC "KAMAZ", the company was digitally reengineered, BFG Simulation decision support systems and BFG iMES rapid response production management systems were implemented. The digital twin reduced product production cycles by eliminating chaos in operational planning and decision-making. We simply compressed time, reduced work in progress, increased the speed of working capital movement, and as a result increased competitiveness. Now in Russia and in the world, a modern production enterprise is, in fact, a black box. Yes, there is a commercial service that receives orders, there are warehouses that cope with the storage of components and their delivery to the shops with varying degrees of success, and then the wilds begin. At the output of such an enterprise, the finished product, as they say, "floats" in timeframe and quality, and no one can figure out the problem, despite the fact that the company is equipped with the latest CNC machines, uses a monitoring system for industrial equipment and the industrial Internet of Things. Not everyone understands that production should be treated as a business, because that is where the heart of the enterprise is located, and it is production that makes money, not the commercial or financial department. The production's digital counterpart sees all the problems and solves them in real time. The key difference between our system and competitors ' products is that we create not just a simulation model of the enterprise, but an optimal model of AI usage, that is, an algorithm that can choose the most successful option from all possible ones. International giants, as a rule, offer a large platform that has to be refined, described all the processes, and programmed in it. In the database of such products, we must describe all production processes, business processes and rules and transfer them to a figure. But we already understand that the company is not in order, and here the existing chaos turns into a figure, which does not solve the problem, but only exacerbates it. And then in the chaos, but already digital, we try to analyze ourselves and look for solutions for optimization. In addition, description is a very time-consuming process.
Is it a kind of statement of fact?
Of course, some processes are optimized, but the output effects are very small – from 5 to 10%. The BFG Group system, on the contrary, breaks existing business processes and rules and builds new, optimal ones. The most important criterion that guides our system when building a digital double is the minimum time cycles at maximum performance.
Another difference is the possibility of operational management using our platform, that is, we created a twin, built processes, and now they must be executed and followed. To do this, we issue shift-daily tasks, monitor and control their execution, and if there are any changes and deviations, then there is an opportunity to check different scenarios, choose the best option and issue an already changed plan for the next shift. Thus, the company will always be effective at any time, no matter what happens, you can make the best decisions and execute them. The company begins to respond quickly to any changes that occur outside or inside, which means it becomes more competitive and cost-effective.
Finalize the PLM system
While creating your system, did you use framework or develop an intelligent program from scratch?
At the beginning, we tried to create a digital twin for our customers based on AnyLogic, Siemens Plant Simulation, and other simulation environments. We tried different software and searched for the most suitable system for us for a long time. But all systems in the world are based on the process description method, so creating of digital twins in them is a time-consuming process. For example, the description of the plant can take it up to one year and require the participation of many people. And when we describe the system for a year, we may already have new technologies, products, resources, and in the end - the twin will not correspond to reality. In addition, as already mentioned above, we will also enter a bunch of errors and get a non-working tool. As a result, we decided to create a system from scratch that is designed exclusively for production companies. There are about five or seven products in the world that can be used to make a digital twin, if not of the entire enterprise, then at least of the workshop. All these products are a programming environment. We do not need to pre-program the system for individual needs.The output is a ready-made digital twin of the entire enterprise, rather than a single shop. Of course, a few existing systems have more extensive twins that are perfectly suitable for tasks that do not lie in the plane of industrial enterprises. For example, AnyLogic can simulate traffic lights, passenger traffic at airports, and logistics. Our product is not a universalsolution for everything, but we focused on a specific niche, a segment of manufacturing enterprises. In its element, our system is able to produce results almost 10 times faster than solutions from Siemens and other manufacturers, and even achieve the effect of 10 times more!
How attractive is your product in terms of price?
The license for our software is a little more expensive than the same Siemens Plant Simulation, but the ownership itself will be much cheaper, as you don't need to hire a few programmers or wait a whole year for them to enter data and "finish" the system according to the features of your plant. And in a year, whatevermay happen, for example, the product range will change or new machines will be purchased, etc.Our system can be used by an ordinary specialist (engineer, technologist) and get the result in two or three months.
In addition, there is a SaaS solution where you only pay a monthly fee for a subscription, which is comparable to the salary of one qualified middle-level manager.
Well, about the second level of operational management, here we need to be compared with MES systems. We are definitely cheaper than foreign twins, given that we create MES for the entire plant as a whole, where all flows are synchronized from the warehouse to the finished product.
To sum up, we are essentially improving the efficiency of the company with our tools and trying to offer the customer a price that suits them, and the effect of our implementation would allow us to cover the cost of our solution in less than a year.
Nowadays, companies often get confused between the related concepts of "digital twin" and "digital shadow". We should understand it like that: the shadow is a kind of statement of fact in the form of taken and possibly processed data from the past and present, while the digital twin in its essence and purpose must have predictive mechanisms and predict how this or that object will behave under different circumstances. Am I right?
When we talk about the digital shadow, we really mean data collected from various objects some time ago. That's right, the shadow is a bit about the past. Within this concept, we can talk about monitoring systems for industrial equipment, and partly about the industrial Internet of things: first we take data, and then we analyze it. But it is only possible to predict, simulate the future, and test different scenarios with the help of a digital twin.
Does your platform have predictive analytics?
On the digital simulator, you can run different scenarios of development and see the consequences of making certain decisions. For example, what will happen if a new product is introduced: how will this affect the implementation of the current production plan, how will it expand the order portfolio without affecting the main production, etc.? We call our software "CAD of production systems". Everything has gone from the drawing board with the drawing paper and the notion of the digital double products almost never cause problems. Companies acquire a lot of various industrial software and have been designing and modeling specific products on computers for a long time, making sure that machines and aircrafts have optimal characteristics. We also strive to track the entire life cycle, with the only difference that our system is not focused on the final product, but on the tool for creating and managing this product. Instead of an airplane or a car, we suggest creating a business (production system) with specified properties (characteristics) and predictable production processes and managing it in a volatile environment.
The ocean of opportunities
The digital twin is not a boxed solution, but rather a reflection of the entire infrastructure of the enterprise. PDM, PLM systems, ERP, MES, BI, IIoT – in fact, there should be a bunch of these tools, only then the digital twins will be able to fully function on an enterprise scale.
Of course, this is one of the levels of development of the enterprise, which is worth striving for. Gradually move from point A to point B and on. In any case, you need to start with the design and modeling of the enterprise itself, to build a framework and understand what the plant should ideally be, and then, when the twin has indicated the path and the result, gradually increase the degree of digitalization. In the same way, we start production of a product with its design in PDM, PLM and only then launch it into production. So why not start designing a business as well? Therefore, we created our solution. Usually the company is literally stuffed with advanced digital assistants, but they do not give an effect for the business as a whole, the machines work in a warehouse, there is no clear interaction between departments, there is no coordination, production cycles are long, and there is a large turnover. Following the e-Manufacturing concept, Russian companies are gradually learning to use the digital twin for the product. And if product`sPLMs answer the question "what", then the digital twins of enterprises helps to understand "how" to produce.
Is the BFG group digital twin suitable for industrial products?
No, our systems only work with businesses. The PLM market is quite diverse, it is not our niche. Our company is focused on the vast blue ocean, where there is almost no one. At the moment, there are no solutions where you can both see the future and manage the present on a single platform.
Why do you think there is such a rush around artificial intelligence? What is the phenomenality of this technology?
AI is not a phenomenon at all, but rather a necessity dictated by life. This is just a mathematical algorithm based on data analysis, it can prompt and help in making decisions. There is so much data that you can't do without AI. First came the abacus, then the calculator, the computer. The AI-related trend will continue to develop. We will see a lot of useful things in the same video analytics, and in the further development of digital twins. We are already using it and are planning further development in our product strategy.
Synergy of engineering and IT
One of the frequently asked questions now is why the AI considers some criteria optimal and others not. You can also train AI using false data, and the algorithm will produce errors in the future.
Imagine a cake. Artificial intelligence is the icing on the cake, but the effect of this icing is just crazy. But if the cake is shapeless and tasteless, then the icing will not save. Remember the recent boom around big data – and big data, in fact, is not necessary for everyone, and not always in huge quantities. After all, data can be just chaotic. Correct verification of information is of great importance. Building causal relationships.
A chronic problem for Russian companies is the lack of qualified personnel. How do you think it is necessary to train personnel for the economy of the present and future?
The problem is that universities prepare graduates who do not meet the requirements of business. We have the same problem – a shortage of qualified specialists. Of course, we cannot do without fundamental knowledge, but from the point of view, of cultivating innovation, educational institutions are not brilliant today. In my opinion, we need to promote modern topics and technologies that are relevant for business to universities.
Venture funds are increasingly investing in its companies in Industry 4.0. This year, the SKOLKOVO Industrial Fund, whose investors include UAC, Russian Railways and Russian Helicopters, announced investments in your company in the amount of up to 200 million rubles. What do you see as your competitive advantages?
The BFG command Group cannot be considered as a pure IT company, first and foremost, we – the manufacturers. Vladimir Kutergin is co-founder and Chairman of the Board of Directors of BFG Group, in addition, he is a doctor of technical Sciences, mathematicianand also had worked as Deputy General Director of a large machine-building holding. Our team also includes plant managers, technologists, and design engineers. We are out of production and therefore know all its pain points, and programmers have transferred our knowledge to the software. Our uniqueness lies in the symbiosis of fundamental knowledge, industrial expertise and its infrastructure, that is probably why we include venture funds. I recently went on a business trip to the Czech Republic. So, there is a lot of interest in our product. BFG Group plans to enter the foreign market with its system and establish cooperation with international venture funds in the future.