The Skoltech Artificial Intelligence Center and the Skolkovo Foundation held the Second Annual Scientific and Technological Seminar on Generative Design in Industrial Engineering. The platform brought together more than 120 leading scientists and practitioners from oil and gas, metallurgy, chemical, machinery, and automotive industries, universities, and federal agencies. In particular, Skolkovo residents presented innovative solutions for high-tech industries.

“Generative design is an innovative technology that uses generative AI to create optimal design solutions. This creates a new process model where instead of performing routine tasks, the engineer sets goals and constraints, while the intelligent system generates the optimal solution. Due to its cross-industry nature, generative design can significantly influence economic development,” noted Evgeny Burnaev, Doctor of Physical and Mathematical Sciences, Professor, and Director of the Skoltech Artificial Intelligence Center.

Specialists from the Skoltech AI Center plan that the future generative design platform will be based on Skoltech's Multi-Agent Engineering Artificial Intelligence System, which includes a library of agents for solving complex engineering problems. The system is being developed by Skoltech under the program “Engineering Artificial Intelligence: Generative and Multi-Agent AI Systems for Natural Science and Engineering Applications” of the third wave of AI research centers within the federal project “Artificial Intelligence” of the national project “Data Economy.”

To develop the generative design field, Skoltech representatives announced the creation of an expert community on generative design. Within this framework, participants will be able to exchange experience in developing and implementing generative design products, create new collaboration models, and gain access to world-class expertise in generative AI, CAD/CAE, and BIM modeling.

“Together with our partners — Skoltech and leading vendors — the Skolkovo Foundation is implementing a number of initiatives united by the ProGenAI platform. This work includes not only market research on generative AI development but also stimulates pilot projects aimed at transforming domestic industry, achieving technological leadership, and implementing science-intensive technologies in production processes. This form of collaboration supports corporate R&D projects involving leading research centers, bringing them to implementation and industrial operation. We see the ideal scenario of this work in the emergence of new science-intensive companies — participants of the Skolkovo ecosystem, including through joint work with Skoltech’s expert community,” emphasized Sergey Dutov, Director of Corporate Innovations at the Skolkovo Foundation (VEB.RF Group), while opening the expert platform.

Dmitry Parkhomenko, Deputy Head of the Federal Center for Regulation and Standardization, delivered a report on the requirement registry as a digitalization tool and software development driver. Dmitry spoke about prospects for automating expert review and design processes of information models using the requirement registry, as well as the use of AI as an assistant in forming and applying requirement registries.

Mikhail Gusev, the director for development at Cyberphysics (a Skolkovo resident), spotlighted simulation modeling for optimal production planning. The company develops industrial software for predictive analytics and optimization of technological processes. The simulation modeling direction is also being developed at the company, allowing creation of digital twins of entire production facilities for optimal planning, justification of investments in production equipment, and productivity increases. A simulation model is a set of agents containing equipment operation logic, site operation rules, workshop specifics, all logistics flows, and operations between equipment. Successful testing of digital simulation models at production facilities has shown impressive results. For example, simulation models can calculate production options for a day, week, or month with high accuracy, optimizing production planning, and ensuring 2-4% production increases. The simulation model’s accuracy in production volume (in tons) reaches 99.4%, and it also allows evaluation of technical requirements for new equipment and analysis of technologies and technological constraints. In the future, simulation models of complex technological objects could become data sources for generative AI.