Denis Kononchuk, First Deputy General Director of RT-Techpriemka, on current trends and practical experience in implementing robotic solutions
In the context of the implementation of key national tasks, domestic enterprises are moving from disparate solutions to managed production systems. Today, robots are being actively integrated into existing workshops, AI is being scaled in mechanical engineering and processes are being automated. One of the drivers on this agenda is the competence center for the quality management system, robotics and automation of Rostec State Corporation — RT-Techpriemka Company. Denis Kononchuk, First Deputy General Director of RT-Techpriemka, spoke about current trends and practical experience in implementing robotics solutions in an interview for the website of the conference "Efficient Manufacturing 4.0".
— Metalworking is a historically conservative industry. What types of operations — loading/unloading machines, milling, welding, quality control — are most "ripe" for robotics today in terms of economic benefits, and where is the robot still an expensive experiment? What payback period do you currently plan for such projects and how is it achieved?
— From the point of view of economics, operations with high repeatability and a clear cycle show the greatest maturity today. The effect of their automation directly affects the rhythm and stability of production. These are the shifting, loading and unloading of CNC machines, welding of standard products, palletization, as well as some basic operations for controlling geometric parameters and product quality. In such areas, robotization provides predictable results by reducing downtime and reducing dependence on scarce staff.
Processes with high variability are much more difficult to robotize, such as complex machining or milling, and operations with frequent changes in nomenclature. In such cases, either deeper technological training is required (standardization, unification, and sometimes a revision of the product design), or the use of additional technologies, such as computer vision, which affects the economics of the project. Then robotization turns into an expensive experiment with no guarantee of stable site utilization.
To calculate the payback correctly, you need to look at how robots help reduce downtime, marriage and cope with the shortage of employees. In practice, a good result is obtained only where robotics solves a specific production problem and is accompanied by process restructuring, rather than being implemented as a separate element. If these conditions are not met, the probability of achieving the payback targets decreases, and the purchased solution is highly likely to be economically inefficient.
— There is an opinion that one of the main "pains" in the introduction of robots in enterprises is a diverse fleet of machine tools (different CNC, age, condition). How is the task of integrating a robot with existing, often not new, equipment being solved today? Which turns out to be more difficult: technical integration or adaptation of production processes and changing the roles of operators?
— Integration with a diverse fleet of equipment begins with diagnostics: the condition of the machines, available interfaces, the level of repeatability of operations, the actual logic of the site and safety requirements. Retrofitting is often required on older equipment. However, the key difficulty is not so much technical as organizational. The robot requires tougher production discipline: stable tact, clear logistics, well-defined roles. If this is not built, even a well-integrated system will not give the expected result.
For RT-Engineering Center (part of RT-Techpriemka of Rostec State Corporation) As an integrator of robotics solutions, it is important to ensure a systematic approach: the implemented solutions must be organically integrated into the existing production and IT infrastructure. Therefore, in practice, integration is always a parallel process: technical integration and adaptation of the production model. Without this, robotics will remain a local solution and will not affect the efficiency of the site as a whole.
— Flexibility is critical for enterprises with small-scale and single production. How to build robotics in such conditions: strive for a "robotic cell for a typical part" or implement quickly reconfigurable complexes? What is the actual changeover rate of the robotic complex now, and who should perform it — the integrator's service department, the CNC programmer, or the machine operator?
— For small-scale production, it is not a universal robot that becomes optimal, but a flexible robotic cell for typical repetitive operations. This approach allows you to maintain flexibility and at the same time not turn the automation of the production site into a labor-intensive and long-term project. For such enterprises, modularity, unification of equipment and the ability to quickly change scenarios without completely stopping the site are important.
The changeover time is of key importance. It is determined not so much by the equipment as by the level of standardization: tooling, products themselves, programs, etc. In the prepared areas, the changeover takes hours, not days, but only on condition that such changes are technologically planned from the beginning.
A separate question is who performs the changeover. At the start, this is the integrator's area, but as the project develops, these functions should be transferred to an internal team of trained specialists: the operator, the installer, or the responsible one. This is critical for the independence and sustainability of production.: Without this, the company remains dependent on external support and loses flexibility in management.
— Today, robotics is increasingly going hand in hand with computer vision and AI. At what stage do these technologies really begin to bring benefits: is it the recognition of parts in the cassette, adaptive correction of the trajectory for the deformation of the workpiece, or on-the-fly quality control? Where is the gap between the promises of vendors and the real possibilities of industrial exploitation?
— Computer vision and AI are already used in industry, but their effectiveness directly depends on the maturity of the processes. The most reliable scenarios are the recognition of parts, monitoring their presence and orientation in space, primary control of the geometry of products, and monitoring the condition of equipment. At this level, AI is already helping to reduce the human factor, especially if the products are fairly typical.
Adaptive management and quality control in real time is a more difficult task. This is a really promising area, but it requires a much more stable production environment. Not only algorithms are important here, but also data quality, hardware stability, and compliance with technological discipline.
The main gap today is related to the fact that the demonstration capabilities of technologies do not always match the conditions of real production. Therefore, AI works as an amplifier of an already established process, standardized and ordered, but not as a tool capable of compensating for its instability.
— After the introduction of the robotic section, what changes in the work of the replacement foreman, adjuster, operator? Is there a new role for the "supervisor of robotic cells"? What challenges do you face when handing over such a site to the factory team after the project is completed by the integrator?
— After the introduction of the robotic site, the very logic of the work being built changes. The proportion of manual operations is decreasing, but the role of managerial competencies is increasing. The foreman is increasingly responsible for the flow and its stability, the adjuster is getting closer to the technologist and the diagnostician, the operator — to the controller of processes and quality. On mature pre—prepared sites, a new role appears - a conditional supervisor, a specialist responsible for the operation of the robotic cell as a whole.
The most difficult stage is the commissioning of the system. Even with successful commissioning, the site quickly loses efficiency if the staff is not trained to deal with typical failures, basic diagnostics and does not understand the logic of the system. Therefore, the key factor is the preparation of the team and the formation of a new culture of work with modern production logic. In this sense, robotization is changing the management culture in general.
— Many enterprises are undergoing a pilot project on robotics. What, in your opinion, distinguishes successful cases when the pilot is scaled to several sites or runs from those where the project remains isolated and does not develop? What are the most common organizational or technical barriers to scaling?
— Pilot projects are scaled when they initially solve a typical production task with a clear economic effect. If, after the pilot, the company sees reduced downtime, stable quality and a clear maintenance model, the chances of replication increase.
If the pilot is of a demonstration nature or is implemented as an "isolated", local initiative that is not integrated into the overall architecture, it usually does not receive development. The main barriers to scaling lie in the organizational plane: heterogeneity of equipment, industrial conservatism and the lack of a managerial decision on system robotics.
Of course, the key focus on the current agenda is shifting from individual implementations to replicated solutions. This is a sign of the transition from pilots to a mature model of robotics.
— Robotics is often perceived as a substitute for humans in production. How correct is this for quality control tasks and what is actually changing in the role of a specialist?
— I would not say that robotics in quality management processes simply replaces humans. Rather, it removes the most repetitive, time-consuming, and human-sensitive operations from him. In quality management, stability, repeatability and objectivity of the result are especially important. Robotic and automated solutions allow measurements and inspections to be carried out using a single methodology, without fatigue, subjectivity and differences between shifts.
But the role of the specialist does not disappear, but changes. A person moves from manually performing the same type of operations to setting up a process, analyzing deviations, making decisions, and dealing with the causes of defects. Therefore, the main value of robotics, in this particular case, is not in reducing human labor, but in increasing the controllability of the process and adaptability.
