In the exhibition zone of the stand of the VEB.RF group "Digital Fortress" Rector Skoltech, Academician of the Russian Academy of Sciences Yulia Gorbunova presented technological solutions to the Institute Deputy Prime Minister of the Russian Federation Dmitry Grigorenko. The exhibition includes five projects:
- Transmission module for inter-satellite laser communication (Project Center of Applied Photonics).
Laser communication will increase the data rate between satellites up to 100 Gbps. The presented coherent transmitter based on photon integrated circuits uses complex signal modulation formats, multiplying the content density to the channel. The sample passed the flight qualification.
- Coherent transceiver for backbone lines (Project center of application photonics).
Fully functional prototype of the transceiver on a photonic integrated circuit. It will become the basis for the first domestic coherent transceivers in the main networks and data centers.
- Temperature control system of the spacecraft (Project center of applied photonics).
It is based on fiber-optic sensors and a compact questionnaire based on photonic integrated circuits with machine learning algorithms. The flight sample for the satellites CubeSat 3U is scheduled to be presented before the end of 2026.
- Photon integral circuits for signal transmission and processing (Center for Engineering Physics).
Samples are manufactured according to the standards of mass microelectronic production and passed tests. They are ready to integrate into real electro-optical systems and confirm the prospects of the domestic photon integrated platform for the digital industry.
- Model of a quantum sensor for navigation and mapping (Project Center of Applied Photonics).
Based on the Ion trap of Paul and elements of photonic integrated circuits. Such sensors allow you to create high-precision gravitational maps of the Earth, and also open up opportunities for autonomous navigation in the absence of a GPS signal.
As part of the business program of the conference, Skoltech became the organizer of two sessions at the VEB.RF site. At the discussion “From Models to Machines”, the leading developers and customers of technology discussed what hinders the introduction of laboratory models on the way to real business. The session was moderated by Skoltech Vice President for Technological Partnerships Shamkhal Dzhabrailov. Main points of the participants:
- Alexey Zaitsev, an associate professor at the Skoltech Center, believes that 10% in the production is not a failure:
“Companies actually had a year to bring to the production of heavy technology, which is changing business processes. 10% have already coped, the rest are still in the process. I look at the situation very optimistically.”
The speaker stressed that the effect of the introduction of AI is not always expressed in one direct figure. Often, together with the model, new competencies, computing infrastructure, data ontology, the logic of their collection and storage appear in the company. This changes the culture of working with data and helps the following AI projects to run faster.
- Oleg Tretyak, Director of Digital Transformation of Gazprom Neft, highlighted the key barriers to the accelerated development and implementation of AI:
“Technological leadership in AI is achieved not by the number of models, but by the ability to turn them into products, to count efficiency and provide infrastructure. Businesses need to solve business problems. Until we move from scientific research to the stage of productivity, nothing will work.”
No less important component for the AI industry is trust in models, constant monitoring and stability of AI models.
With large LLM-models, it is impossible to drive away the entire set of tests, it all depends on the industrial scenarios. Updated open-source models appear every two weeks, so there is no architectural binding to one model in the perimeter of the company, the internal platform includes a set of repositories of models that quickly switch to solutions.
- In the continuation of the theme of productization, Nikolai Truzcal, Director for the Development of AI Technologies, FabricaONE.AI (Somftline Group) noted that when talking about productivization, it is necessary to go down to a level lower - the level of platforming:
“With the spread of AI, the input barrier to the development of software is reduced. We need a platform where each company will assemble its own architecture.”
At the same time, it is critical to change the culture: the risks from working with AI should not try to blindly prevent, but learn how to correctly evaluate and calculate them.
- Denis Surzhko, Deputy Head of Data Modeling and Analysis Department, VTB Bank:
“According to studies, only 5-7% of organizations from those that use generative AI in their processes felt the economic effect of its implementation. The key difficulties are the presence of deep internal expertise, the development of interaction with AI users and the high cost of computing resources.
He also shared an interesting case: VTB, together with T1 and Chinese partners, was able to successfully adapt alternative Nvidia solutions, creating the largest cluster independent of the American monopolie in Nvidia Free, which allows you to make fully import-placed solutions.
- Igor Drozdov, Deputy Chairman of VEB.RF, began his speech with the fact that Skoltech is part of the VEB.RF group, and through it the development institution supports the AI industry. But for any bank, the key criterion is to prove the economic efficiency of the project and ensure the return on investment.
“We are not on the sidelines of foreign progress, we are developing and implementing different solutions, but so far this is the development of hypotheses. AI is a support tool, and the decision is made by a person in the end.”
Opening the second session of Skoltech at the VEB.RF platform “Developing 5GA/6G. Promising technologies through the cooperation of science, state and business”, moderator Natalia Kosmodemyanskaya, head of the Skoltech Project Support Department, noted that although the commercial introduction of next-generation technologies is not expected until 2030, the whole world has already actively joined this race.
“We have starting conditions more difficult: 5G networks have not yet been fully implemented, and there are many other priority tasks in the country. However, in recent years, its own scientific and technological agenda 6G has begun to form.
- Deputy Minister of Digital Development Alexander Shoitov noted that the introduction of the data economy depends on the speed of the Internet connection. According to him, new effective systems are needed. In 2025, the Ministry of Digital Development launched a roadmap for the development of key technologies 5G Advanced and 6G - the winner of the competition was Skoltech, the deputy head of the department recalled. He stressed that by 2030, all critical technologies should be fully developed in the direction of 5G Advanced. By the same date, it is necessary to create a maximum reserve for 6G, he concluded.
- Alexey Frolov, professor, director of the Project Center for Wireless Communication and the Internet of Things Skoltech, spoke about the eight-year reserve of the Institute for 5G base stations.
“In 2023, the first experimental sample of 5G base stations was created. These works continue within the framework of the roadmap, and we thank the Ministry of Finance for their support. As part of the roadmap, we chose critical areas: software with AI elements, digital twins and antenna systems based on photon integrated circuits. And it is also important to cooperate with the state, vendors and operators to clarify the tasks, to learn new ones, to receive feedback.
- Dmitry Lakontsev, CEO of Irtea, noted that it is difficult to build direct communication with the industry.
“Let’s be honest: some fragmentary things have been preserved in our universities. I hope that the roadmap will build a systemic cooperation between us. Another problem that is present in our universities is a rather low level of technology readiness.”
- Yulia Klebanova (GC "Picle") stressed the pragmatic approach.
“To go further and plan the implementation of 5G Advanced and 6G, we need to adhere to the basic principle – this is the principle of economic feasibility and recognition of the decision. The theme of AI is not just some fashionable hype, but an obvious need at the moment of transition, since it is impossible to manage such a number of tasks with the help of classical superstructured systems. ”
- Vladimir Franchman (Protei) added that the industry is changing quickly, so industry players need to constantly monitor this development.
“We are trying to make sure that the product can be offered, among other things, to the global market. We want to understand what’s going on in the industry and we want to follow it. We have some part of the team always working on promising technologies.”
- Grigory Seregun (MIPT) shared his experience in creating the first pilot zone and identified a number of problems.
“In the process of development, we are faced with the lack of initial data and with the inconsistency of what is written in the specification. It is necessary to get rid of false information in the marketing materials of manufacturers. To do this, it is necessary to create independent pilot zones in Russia.
- Evgeny Novikov (DIT of Moscow) presented the work of DIT in the field of promising technologies:
“We tested about 80 startups on the basis of our Democenter. In Moscow, it is planned to open two or three more 5G landfills, which will become pilots.
According to the speaker, in terms of promising research, DIT closely cooperates with universities. The study on the world trends in cooperation 6G can serve as a starting point for the formation of domestic models of cooperation.
militarnyi.com
www.defensenews.com
