2022/08/05 07:13:34

Smart manufacturing

Technological futurologists foreshadow that soon all workers in industrial workshops will become "blue collars," who in a separate comfortable room press the buttons, remotely controlling the work of melting furnaces, baths where molten steel cools, etc. However, the path that enterprises have to go to achieve this ideal state is very long and difficult. At what point in this path are the Russian industrial the enterprises today? The article is included in the TAdviser review "Artificial Intelligence Technologies"

Content

Features of smart digitalization of the Russian industry
IoT-Based Production Intelligence
Platform approach
Predictive analytics of technical systems
Digital Plant Twin
Digital Twins and Virtual Tests
Digital Twins for Organizational Models
Manufacturing Logistics Solutions
Deserted manufacturing technology
Smart agriculture
Smart technology
Other Review Materials
Other materials on the topic of AI

According to the research company, MarketsandMarkets the global smart manufacturing market (Smart Manufacturing Market) will grow from 214.7 billion. in dollars 2020 to $384.8 billion in 2025, showing an average annual increase of 12.4%. The development of this segment of digitalization of industry is pushed by the general growth in demand for digitalization of industry in - Russia it, according to the CNews Analytics IT in Industry study published in December last year, will increase 14 times by 2030. For example, the demand of the manufacturing industry for digital technologies, which, according to the to data study, was estimated at 41.5 billion rubles in 2020, by 2030 may grow to 587.5 billion rubles. The list of the most popular digital technologies includes,,, industrial robots artificial intelligence machine learning digital prototyping, sensorics.

Source: CNews, December 2021

Features of smart digitalization of the Russian industry

According to expert estimates, automation of production in Russia is taking the first steps. Mass complete automation of production in the format of so-called deserted production is unattainable in the foreseeable future. A number of reasons inhibit these processes.

All technological processes are quite difficult, each unit in production consists of complex mechanisms, and most of them were made under the Soviet Union. This is one of the important factors hindering digitalization, says Yaroslav Shmulev, head of the Jet Infosystems machine learning group. - In the coming years, the industry is unlikely to be able to replace the entire technical base with a more modern one.

Our enterprises are now only at the beginning of the path of building the IT landscape and internal processes, the expert notes.

Source: CNews based on SAP data, Deloitte, December 2021

Take the steel smelting process as an example. This is a complex process related to chemistry, physics, thermo- and hydrodynamics, - says Yaroslav Shmulev. - To create a mathematical model, you need metallurgists, physicists, mathematicians who will study the process and make an honest model.

This means that the system should know and see everything that a person sees, and have historical data on the process for a year, or better for two, the expert emphasizes.

But in production, it often happens that technologists know their part of the process well, but do not have an end-to-end vision of the process as a whole, and this also becomes a problem for implementation.

If this is possible, but the resulting model does not work well, while a person in production copes with the same task, then most likely it is about data and/or its quality. For example, the model may not receive some of the critical information that the technologist has, - explains Yaroslav Shmulev. - If the forecast model works, then we move on to creating a recommendation service that works in real time 24/7, giving advice to the technologist and helping to optimize the process.

For example, at the Novolipetsk Metallurgical Plant (NMLK), artificial intelligence helps a steelworker who smelts steel to introduce additional chemical elements, such as ferroalloys, into steel during smelting. Ferroalloys are quite expensive materials, they significantly affect the cost of metal produced, but even a very experienced steelworker does not always manage to accurately fall into the range of permissible chemical parameters the first time. Actually, this is the optimization problem solved by the recommendation system, which was developed and implemented at NLMK by the specialists of Jet Infosystems: to determine the minimum required amount of ferroalloy in order to fall into a given interval in terms of chemical composition, to use the minimum amount of materials themselves, and if possible, then use the cheapest material.

The key element of the created system is a mathematical model based on algorithmic machine learning, which predicts what the chemical composition will be if certain materials are added at a particular moment in time. The plant as a whole is a large socio-economic system. To describe the complex dynamics of processes in such systems, the mathematical apparatus of simulation modeling is best suited. This is the most adequate way to study systems and processes in the absence of complete, accurate and reliable information about their properties, notes Alexey Gintsyak, head of the laboratory "Digital Modeling of Industrial Systems" at the NTI SPBPU Center.

In this context, the introduction of projects using AI is part of the process of digitalization of production, which in general has the goal of a gradual transition from manual control of processes and individual units to automated control of the entire production chain, says Yaroslav Shmulev, but while in Russia industry automation takes only the first steps, most processes are manually controlled by the operator.

Today there are many interesting technologies on the market - autonomous equipment, artificial intelligence, etc. It would be tempting to investigate them. However, the current maturity of these technologies is not enough to get an economic effect from them right now, "says Artem Natrusov.

Yaroslav Shmulev lists those AI technologies that have already reached a certain maturity for use in industrial enterprises. Firstly, this is machine vision - a combination of video cameras with computing power that processes data. With the help of cameras and machine learning today, you can count the volume of harvested forest, sort minerals, track violations of safety rules and solve a number of other tasks of the industrial industry.

When using computer vision, the model learns faster and works more accurately, says Yaroslav Shmulev. - Nevertheless, even in this area, each project is still carried out as an experiment that may be successful or not.

Therefore, in his opinion, a pilot project is required, which shows, perhaps in a particular case, the use of AI methods or this is impractical.

One model predicts what will happen, acting as a simulation of the process or its digital twin. The second model solves the optimization problem: it submits several options for control parameters to the forecast one, selects the best set and issues it as a recommendation to the technologist.

This is how the system implemented at NMLC works.

Over the next 5-10 years, they are likely to appear in every production.

At the same time, it is important to understand that these are precisely the systems-advisers, the final decision and responsibility for it still remains with the person.

As a rule, an industrial unit has a limited efficiency potential, and it can be improved only partially, and not several times. But even small percentages give a good profit in total. More complex projects that optimize the operation of not one machine, but the entire chain, pose a greater risk for us, which can be realized in the form of failure to achieve the expected effect or exceeding costs, says Artem Natrusov.

IoT-Based Production Intelligence

Industrial solutions Internet of Things are a market segment that is actively developing around the world. According to MarketsandMarkets forecasts, the market volume Industry 4.0 in 2021 will amount to $64.9 billion, and by 2026 will grow to $165.5 billion. The average annual growth rate during this period will be 20.6%. Key factors contributing to its development include the rapid adoption of artificial intelligence and the Internet of Things. The company that IDC investigated the Russian AI market at the end of 2021 noted: {{quote 'The production sector of both discrete and continuous, often uses elements of artificial intelligence for automated preventive maintenance and quality management in projects related to the use Internet of technology of things. }}

For example, the operator MTS has completed a pilot project "Digital Vodokanal" for the municipal institution "Kopeisky water supply and sanitation systems" in Kopeisk, Chelyabinsk region. An IoT solution has been implemented to collect and analyze information from pressure sensors and water supply metering devices. The result of the system: detection of illegal tie-ins and unaccounted consumption on water supply lines, the ability to record pressure drops in pipes and prevent accidents at the earliest stages, reducing the costs of resource-supplying organizations by reducing the balance of water consumption.

According to resource supplying organizations, the total losses in water supply networks amount to more than 100 billion rubles. per year. Water channels are faced with data distortion and unaccounted consumption, and it is difficult to control calculation services, since most operations are performed manually.

The introduction of an integrated fault-tolerant infrastructure provides an increase in the level of availability of data for making operational management decisions, - emphasizes Alexander Eder, director of business development at CROC in the agro-industrial complex.

High-precision positioning of the bucket allows you to extract the desired 3D cube in order to draw precisely the ore that contains gold, and not empty rock. At the same time, the explosion model is taken into account, the company says.

The pit dump truck connected to the IoT platform chooses where to go, based on data on the ore content in different faces, in order to ensure stable quality on the crusher. At the same time, the recommended dump truck system prompts the operator to use the best gas and brake pedal to save fuel and pads as much as possible.

You understand exactly where you are extracting from, where you are taking, where the material is extracted, what its characteristics are, you can calculate the resources for its processing and determine the most effective way to enrich the rock.

Platform approach

Zyfra uses a platform approach to creating IoT solutions. The Zyfra Industrial IoT Platform (ZIIoT) is a single digital platform that includes all the necessary components to create and innovate your enterprise. It enables the enterprise to get a unified data management environment in order to optimize production based on AI mechanisms, big data and the Internet of Things. Within the framework of a single platform, it becomes possible to build the management process through a single system, ensure high response speed, integrate existing automation systems into a single complex, as well as prepare data for ML solutions and create digital twins. This is essentially a constructor that contains pre-configured industrial automation scenarios for different industries.

With the help of machine learning, recommendations are formed on how to accurately go in an oil-bearing formation and avoid drilling complications, because each stuck can lead to the loss of a whole shaft and cost up to 100 million rubles, and the technology itself allows you to subsequently increase production and get an additional 10 tons of oil from one well per day, - say the specialists of the company "Digital."

In fact, we have created a kernel on the basis of which any company can write a solution for itself, - emphasize in the company "Zyfra."

As a result, combining the entire production process and the economy on one digital platform, companies get the effect of end-to-end inter-channel optimization.

The experience of Zyfra in the automation of work in mining and processing, industries, implemented in metallurgical machine-building the form of the ZIIoT platform (Zyfra Industrial Internet of Things Platform), interested the company. At MTS the end, the companies entered into a partnership agreement, which involves the deployment and replication of the ZIIoT platform cloudy in the infrastructure# according to the CloudMTS model PaaS and the integration of the platform into MTS digital industrial solutions based on dedicated Private/technology networks. LTE 5G

The complexity of numerous aspects of the digitalization of industrial enterprises determines the demand for ready-made solutions tuned to the specifics of a particular area of activity. Analysts Gartner predict that by 2024, 50% of manufacturers of industrial platforms in the world will offer them immediately with ready-made solutions, and 40% who have abandoned this path will leave the business. Experts talk about the upcoming consolidation of the market due to the fact that customers vote and will continue to vote in rubles for platform solutions with fast business effects. For example, Datana offers a set of ready-made digital solutions based on a single Datana Smart digital factory stack, including:

Slag cutoff on MLS (continuous casting machine): gives a recommendation to the operator on the need to close the gate valve "for advance," adjusted for the operator's response time and the operation time of the gate mechanism.
Slag detection when metal is discharged from an arc steelmaking furnace or acid converter: reduced consumption of ferroalloys and deoxidizers
Out-of-furnace process control: reduction of argon and electric power consumption.
Service of temperature ordering on MSL: reducing the consumption of electrodes and electricity on the ladle furnace.
Optimization of the decarburization process on the vacuum cleaner.
Optimization of ferroalloys consumption.

Predictive analytics of technical systems

Another mature area of intelligent industry automation is predictive analytics for the tasks of monitoring the state of technical systems. In 2018, as part of the NTI, the Center for Big Data Storage and Analysis Technologies was organized at Moscow State University. It was created in the form of a consortium, which included a number of leading Russian technical universities, as well as integrator companies LANIT and Rakurs.

The created analytical platform is a software package and a set of predictive analytics models for predicting equipment failures and improving production efficiency.

Structure of predictive analytics analytical platform

Source: NTI Big Data Storage and Analysis Technology Center

On the basis of the platform, for example, a predictive analysis system for wastewater based on the spectroscopy method is implemented (the UV-VIS spectrometer is used). For example, 150 solutions and spectroscopic measurements were prepared to determine the concentration of phosphates and chlorides. Then the optical spectra were processed using machine learning and a model was built to determine the concentrations of salts in the distillate solution.

Predictive Wastewater Analysis System Based on Spectroscopy

Source: NTI Big Data Storage and Analysis Technology Center

Severstal A predictive model for monitoring the failure of gas converter superchargers has been created for the company, which includes predictive models for various types of equipment in steel production with a prediction horizon of 1 to 3 days and a metric system.

According to Ilya Mukha, head of software development at the NTI Competence Center in the direction of "Big Data Storage and Analysis Technologies" at the Lomonosov Moscow State University, according to the results of the project, the fact of the possibility of reducing the downtime of equipment by more than 2 times has been practically proven,, the possibility of reducing emergency failures at the 30% was proved, it was possible to reduce the number of WEPs carried out by 55 hours per year, and the economic benefit from the introduction of models can be up to 60-70 million rubles. per year for one type of equipment.

More than 150 different parameters were analyzed and a model was developed for early notification of the operator

Source: NTI Big Data Storage and Analysis Technology Center

In addition, within the framework of the project, the concept of introducing a digital twin was developed to simulate the operation of the hot rolling mill stand and determine critical loads during rolling and simulate various malfunctions.

Digital Plant Twin

We simulate a real process and get an accurate result. Any parameters can be changed and the resulting results can be compared, then select the optimal solution and configure the execution of the production plan, the company says.

Datana Smart is an industrial-class system built on open source code, easily embedded in the enterprise infrastructure, providing a single environment for all digital production solutions based on the digital twin management platform. The company explains: a digital twin, receiving feedback from a real process, simulates its work and automatically selects parameters for the effective operation of a particular unit. At the same time, several such projects can be launched as part of the digitalization of the enterprise. This solves the problem that is often encountered today: pilot projects of various smart solutions for improving various production sites are launched simultaneously and independently of each other, which leads to the emergence of another "zoo" of incompatible AI solutions at the enterprise.

In order for digital twins to give a tangible result, they must work in a single information infrastructure, the company says, and for this you need to model the entire digital supply chain. To support this capability, the Datana Smart platform implements a collaboration mechanism between different digital advisors and their models, which has an integral effect on the business, including traceability of the entire supply chain and dependencies, comparison of advisers from different suppliers, uniform rules for the work of advisers for all suppliers (degradation reports, business monitoring, user interfaces, technology stack and software interfaces). From January 1, 2022, the national standard GOST R 57700.37-2021 "Computer models and modeling began to operate in our country. Digital product twins. General Provisions, "which was approved on September 16, 2021. As noted by Alexey Borovkov, Vice-Rector for Digital Transformation of St. Petersburg Polytechnic University Peter the Great, the initial scope of the standard is mechanical engineering products, however, if necessary, standards may be developed on its basis, establishing requirements for digital doubles of products of various industries, taking into account their specifics.

Digital Twins and Virtual Tests

Modern digital twin technologies can significantly reduce the cost of product development. A striking example is development in the field of engine building, says Alexey Borovkov. - So, if the implementation of several dozen prototypes is necessary, then a number of tests must be carried out, which increases the cost of the product.

Reducing the volume of field tests, optimal organization of maintenance of repairs at the operation stage, reducing the cost of development, reducing the time for launching new products on the market are the economic effects of advanced digital and production technologies, the expert emphasizes.

In early January, it became known about the launch of the first software and hardware complex in Russia (PAC) for virtual tests of electronics.

PAC "ELBRUS-ASONIKA" - a joint development of INEUM named after I.S. Brook and the company "NII" ASONIKA "), which has no world analogues. The PAC includes:

a computer based on the Russian processor "Elbrus-8SV," developed by PJSC "INEUM named after I.S. Brook" together with JSC "MCST";
Russian CAD electronic engineers in terms of virtual tests;
automated system for ensuring reliability and quality of ACONICS equipment;
national standards of the Russian Federation in the field of modeling and virtual tests of electronics, developed by the ASONIKA Research Institute.

In addition, the ELBRUS-ASONIKA PAC includes such unique developments that comply with Russian GOSTs as

database of the entire domestic electronic component base (ECB) and materials according to geometric, physical and mechanical, fatigue, thermophysical, electrical and reliability parameters;
design cores and specialized graphical interfaces for analyzing and ensuring the resistance of electronic equipment (EA) and ECB to complex thermal, mechanical, electromagnetic effects, fatigue strength to thermal and mechanical effects;
subsystem for creation of ECB operating mode maps based on integrated simulation of physical processes in EA;
subsystem of EA reliability indicators analysis based on ECB operating mode maps;
subsystem for creating digital twins of EA and ECB.

The creation of this complex is in line with the strategic direction in the field of digital transformation of the manufacturing industries, approved by order of the Government of the Russian Federation No. 3142-r dated November 6, 2021. In accordance with this order, within the framework of the digital transformation project of the manufacturing industries industries "Digital Engineering," it is planned to create by 2030 a national standardization and certification system based on virtual testing technologies.

On the basis of the ASONIKA Research Institute in 2020, a technical committee for standardization TK 165 "Computer-Aided Electronics Design Systems" was created, and during 2021 Rosstandart put into effect 7 national standards in the field of modeling and virtual tests of electronics developed by the ASONIKA Research Institute. Currently, within the framework of TK 165, together with FSUE MNIIRIP and JSC TsKB Dayton, national standards in the field of CAD electronics included in the plans of the national standardization of Rosstandart are being developed.

Digital Twins for Organizational Models

It is in Russia, where the industry represents a set of vertically integrated holdings, that the effects of creating a single digital layer for each industrial group can be maximum, according to the Zyfra company. This aspect was highlighted by Elena Tishchenko, Advisor to the Dean of the Faculty of Economics of Moscow State University Lomonosov on Digital Economics, speaking at the Round Table "Economic Effects of Digital Transformation of Industry" at the III International Forum "Advanced Digital and Production Technologies" in December last year.

She noted the fact that digital twins allow modeling industry cooperatives: "Digital modeling or digital polygons allow scientific and technical laboratories to move away from vertically integrated communication to the formation of long cooperation chains, both intra-industry and inter-industry." As an example, Elena Tishchenko cited the effectiveness of the model of modular assembly of solutions for industry problems created at Rosatom State Corporation. "We see that reengineering of industries is taking place," emphasizes Elena Tishchenko. In particular, digital twins in BIM allow the use of models for monetizing the results of intellectual activity.

In addition, digital design normalizes ontological maps. For the first time we can look at the convergence/reengineering of industries as an engineering process, - said Elena Tishchenko.

We are talking about a digital tool for creating the ontology of an enterprise. Note that in this context, the concept of ontology is used, rather in the philosophical sense than in the form of mathematical formalism.

For example, Dan Rose is professionally engaged in ontological modeling (design and implementation of applied ontological models of activities of specific enterprises and organizations) and decision support systems (SPS) based on the accumulation and structuring of experience and system-situation analysis of activities.

Ontology of the enterprise: expirientological approach. Technology for building an ontological model of an enterprise based on the analysis and structuring of living experience ":" Not reaction to events, but the development of optimal solutions based on continuous scanning and advanced forecasting of the behavior of the Environment - living space.

In the figurative expression of the author, the development of information systems for the inheritance of experience based on QuaSy's own methodology and technology is "an attempt to create a virtual" brain "that is able to accumulate experience and ensure the work of the collective mind of any organization at the level of a well-integrated and extremely synergistic (coordinated) non-stop and online system."

Example of BEOM Ontological Model

Source: Shvedin B. Ya. Ontology of the enterprise: an expirientological approach. Technology for building an ontological model of an enterprise based on the analysis and structuring of living experience. - M.: LENAND, 2010.

Manufacturing Logistics Solutions

According to IDC research, the use of algorithmic machine learning to optimize the supply chain has not yet attracted large investments, but in the period until 2024, these expenses of enterprises will grow on average three times faster than the market as a whole.

The main solutions of production logistics include:

Manage shipments, purchases and inventory
Production planning
Operational planning
Fleet Management
Manage shipments, purchases and inventory

Zyfra PSP (Production And Shipping Planning) is a supply chain management system of Zyfra, designed for adaptive and automated planning and management of logistics and service processes. It provides accurate forecasting, real-time recommendations, and automation of the planning process to ensure supply chain stability in a dynamically changing world.

To achieve these goals, Zyfra PSP uses advanced planning technologies:

Dynamic planning. The production and shipment plan is automatically recalculated, adapting to the current situation. Moreover, this process does not require human intervention.
Planning in conditions of uncertainty. The accumulated historical data on past operations allows Zyfra PSP to identify risks in real time and make realistic plans that take into account typical problems, time and resources to solve them.
Cost and efficiency planning. The digital twin for planning the production process automates the calculation of optimal prices, as well as communication with the client.

The company "Zyfra" talks about the experience of switching to a flexible system of logistics planning and risk forecasting of a large oil refining company: the possibility of rescheduling in 50 seconds and making daily plans for 30 days in advance is realized, while planning for 90 days in advance is carried out every 10 days. This flexible approach to planning led to the fact that the number of fines decreased by 4 times, which corresponds to 1% of the cost of the product (before the start of the project it was 5%).

Deserted manufacturing technology

Robotic dump trucks, drilling rigs and excavators, deserted technologies for extracting natural resources, managing an enterprise using a computer mouse are all characteristics of the ideal digital future of industry, which, however, is already being implemented in individual advanced projects.

• Smart quarry. Zyfra and Huawei agreed at the end of December last year to develop technologies for autonomous robotic complexes using 5G networks, which will become the technological base for the implementation of the Intelligent Quarry concept. This concept involves the creation of an infrastructure for deserted mining, which implements the technological cycle of loading and transporting TPI in autonomous mode.

The partners will carry out comprehensive (End-to-End) digital projects, which include the introduction, ON AI application of 5G technology, 5G/CPE modules, 5G network infrastructure, as well as the use of services Huawei Cloud for autonomous driving of mining trucks in open-pit mining enterprises. In particular, the use of 5G networks will allow transferring part of on-board computing to servers enterprises, which will lead to cheaper autonomous robotic complexes.

The first project of this kind was successfully implemented at SUEK, where mining trucks and drilling rigs were successfully robotic.

• Drones to scan underground tunnels. At the end of April 2022, Norilsk Nickel announced the start of using drones to scan underground tunnels. You can connect a lidar scanner, a video camera, a thermal imager, and an echo sounder to the devices.

With the help of such drones, it is possible to carry out geological exploration in places where there is no convenient access for humans or a direct view at a depth of up to 100 meters: for inspection of treatment chambers, when shooting hard-to-reach workings, ore runs, rock outcrops and non-workings. The video recording photo mode allows you to quickly and in detail view the state of the attachment and the volume of spent cleaning space.

Source: Norilsk Nickel Company

As explained in Norilsk Nickel, earlier filming took a long time and was carried out by two teams of surveyors. Thanks to the use of drones, in good weather conditions in the photo, video recording of a rock dump with a volume of 2 million cubic meters. m takes only 10-15 min.

"Underground" projects are also conducted by Evraz. For example, a solution based on sensors, a video surveillance system, mobile devices and dashboards with convenient monitoring allows you to more efficiently extract natural resources. The company is also developing a joint project with the MTS operator: an underground communication infrastructure is being created based on the LTE network, followed by an upgrade to 5G, which will be used, firstly, for monitoring equipment in mines, and further - for remote control of machines with autonomy functions.

Smart agriculture

Programs of "smart farming," "precision farming" operate in dozens of countries around the world. The introduction of artificial intelligence technologies in the agricultural complex, according to Markets and Markets, is growing by 22.5% per year, and in 2025 the global volume of this market will amount to $2.6 billion. Analysts at J'son & Partners Consulting counted about 2 thousand companies supplying high-tech solutions for agricultural automation. They were merged into a separate segment of Agrotech (AgTech), which, according to J'son & Partners Consulting, has been overtaking FinTech in terms of investment growth for several years. The most popular innovations among American farms are:

collection and analysis of soil samples (used by 90% of respondents);
yield maps, yield monitors, navigation GPS-systems (about 80%);
technologies of differentiated application of fertilizers and prescribing maps (60%);
satellite images and plant vegetative index analysis (30%).

In our country, the rise of agrotech is expected in the near future. Including, due to the attention of the state. Thus, the Ministry of Agriculture of the Russian Federation is implementing the Digital Agriculture project, which sets ambitious goals - digital technologies should help double the productivity of agricultural enterprises by 2024.

In early June, the chairman of the Cabinet RUSSIAN FEDERATION Mikhail Mishustin announced state support measures for the introduction of digital technologies agricultural in the field. It is planned to allocate more than 900 million rubles. for the use of digital technologies in. agriculture This year, in 8 pilot regions (,,,, and Voronezh Kurskaya,, Bryansk) Tula Nizhny Novgorod Moscow region Perm Territory Tatarstan AI technologies will be introduced, initially in relation to seven agricultural crops: spring and winter wheat, sunflower, corn, sugar beets, buckwheat and potatoes.

It is also planned to launch the register of federal property of the agro-industrial complex. It is expected to help gather together disparate information, analyze and make thoughtful decisions based on them. Moreover, records of business books will be digitized, many of which have been kept since the 30s of the last century and often exist only in paper form. {{quote 'This will allow collecting data on manufactured products into a single information platform. As a result, the state system for monitoring and forecasting food security will become more accurate, - said Mikhail Mishustin. }}

industries The state corporation "" takes part in the transition of the agricultural country to "digital Rostec." The enterprises of the radio-electronic cluster, which are part of the holdings Roselectronics"," "" and Schwabe, " Concern "Automatics" work in the field of precision agriculture and effective crop production. Developed software complexes for intelligent farm management, robotic systems, drones for monitoring agricultural facilities, solutions for precision agriculture on the basis. Internet of Things

In particular, unmanned complexes manufactured by Zala Aeroconcentre "Kalashnikov" are actively used for aerial photography of agricultural land. Their use makes it possible to assess the state of soil and plants, increase the yield of land, optimize the costs of fertilizers and plant protection products, and determine territories that need additional irrigation.

Technologies of differentiated application of fertilizers assume that in order to determine the required volume of fertilizers, samples are taken at each site, the obtained results are analyzed, field maps are drawn up, tasks for equipment in the field are determined using satellite navigation technologies and specialized software for remote control of equipment.

AI-based solutions have direct economic effects.

As an example, he cites the Digital Agro platform, created specifically for the company's customers, where it is planned to place a large number of services that are useful to customers, for example, field surveys or precision farming (Precision Farming).

To achieve maximum effect, it is important to implement not only separate smart equipment, but also comprehensive solutions for automating enterprise processes. For example, an intelligent information and analytical system for digital crop production was developed by specialists from NIIIT JSC (Roselectronics). The system allows agricultural enterprises and farmers to switch to rational use of fertilizers, based on the needs of a specific field site. Based on the data of chemical analysis of the soil, the so-called "digital map" of farmland is compiled. Farmers are given recommendations on the optimal planting of crops, the amount and type of fertilizers and plant protection products prepared taking into account the condition of the soil. Then the task cards are received by smart agricultural machinery - seeders, sprayers, spreaders. Farmers can manage this IT toolkit through their personal account. It displays all field logs and recommendations created by the system, and notifications arriving on a mobile phone are reminiscent of carrying out certain works.

Smart technology

According to a study by Statista on the state of the artificial intelligence (AI) market in 2021, it is noted that the market for collaborative robots will more than double by 2026.

Collaborative Robots Market Growth Forecast 2020-2026

Source: Statista, RBC

For example, in Ust-Luga, Leningrad Region, robotic dogs, self-propelled platforms, tiltrotors and drones take part in the construction of a production complex for the processing of ethane-containing gas. Their task is to scan, photograph, shoot video, and then process information and transfer data to a special digital platform. All this is necessary in order to monitor the dates, quantities and quality of work performed in real time.

Source: it weekly, January 2022

Smart software is able to compare the materials used with those specified in the estimate, is able to count the number of workers on the site, and also monitors the observance of safety rules by each of them (the presence of a helmet on the head, etc.). - if an employee does not use, for example, a helmet, the digital controller will notice this. All the information collected is consolidated: work areas with equipment, dozens of technological processes, hundreds of contractor companies and thousands of employees - are combined in one digital space. All collected information is processed in the analytical system, which automatically generates summary analytics and various reports: on mounted structures, used materials, planned dates and other parameters.

The Moscow State Real Estate Inspectorate is testing a robot house. Its task is to identify property and land violations.

Source: it world, December 2021

The robotic system, made in the form of a small friendly dog, is equipped with a GPS module, a lidar (light radar that measures distances) and a built-in video camera. It can reach speeds of up to 11 km/h and use lidar to form an accurate three-dimensional map of the surrounding space. Thus, the state inspection gets the opportunity to remotely and without the participation of people to measure real estate objects. The built-in camera makes it possible to conduct video broadcasting online.

The smart device has training abilities: it is trained in the skills of moving along given routes and conducting photo and video recording. During the current year, it will be tested in real urban conditions. At the KAMAZ factory "" successfully tested Russian the robot manipulator A12, manufactured by the Kazan company. Eidos Robotics The robot worked at a press and frame plant at the welding site, where, according to Vestnik KAMAZ, it performs a monthly rate of work in one to two weeks.

Russian robot manipulator "Eidos" is busy welding the bracket of the cabin hydraulic lock

Source: it weekly, March 2022

In the workshop, a full-fledged robotic cell and a turntable are mounted on the chassis, where electric arc welding is carried out with the supply of wire in semi-automatic mode. The task of the worker in this area is only to fix the next workpiece on the turntable.

Photo protection is provided at the robot workplace: if an employee crosses the perimeter of the work area delineated by a yellow line, the operation of the entire system is blocked.

The Perm company Promobot, a Russian manufacturer of autonomous service robots, has created a cyber cafe where the anthropomorphic robot cashier Dunyasha offers visitors ice cream, coffee and other drinks, while maintaining the conversation.

The humanoid electronic seller is based on the Robo-C-2 droid model, which Promobot introduced in January 2022. This model uses a new type of servo drive, which allows the robot to express emotions more vividly using micromimics control software. A new technology for making silicone skin is also used, which allows it to have greater elasticity. In addition, Robo-C-2 have new hyperrealistic eyes. The achieved level of "human similarity" made it possible to implement the robot in the image of a real-life person: the first "Mrs. Perm 2014" Diana Gabdullina became the prototype of "Dunyashi."

"Dunyashi" has its own character and manner of communication. While waiting for an order, she can support the conversation, tell a fairy tale or offer to take a selfie together. The robot software includes a whole range of entertaining and cognitive interactive skills.