Published on: 2026-05-06
Source: Saint Petersburg Polytechnic University of Peter the Great –
An important disclaimer is at the bottom of this article.
Specialists from the Saint Petersburg Polytechnic University of Peter the Great have developed a set of algorithms for optimizing technological processes in industry and construction based on a multi-agent approach. The development allows increasing the efficiency of resource distribution in decentralized production systems and, consequently, reducing the consumption of energy and material resources.
Work is carried out in partnership with the companies “Gazprom Neft”, “IdealStroy”, “Gazpromneft NTC” with the support of the federal program “Priority-2030”.
Multi-agent system — is a distributed system where multiple independent participants-agents (people, machines, robots, software modules, logistical nodes or even individual production sites) interact with each other and the environment so that the entire system works effectively. The project is aimed at solving the problem of the limitations of existing tools for optimization of technological processes by means of a multi-agent approach, which leads to increased efficiency and stability of the obtained solutions.
SPbPU specialists have developed a system in which optimization is implemented through iterative (step-by-step) communication of intelligent agents that clarify production management scenarios, taking into account individual plans and other incoming information. The multi-agent system itself is part of the digital platform for multimodal data analysis POLANIS, being developed by SPbPU specialists within the framework of one of the university’s three key scientific and technological development areas — “Artificial Intelligence for solving cross-industry tasks.”
The developed algorithms have already demonstrated an increase in optimization efficiency compared to existing methods. Thus, they allow finding solutions of the same degree of optimality 10 times faster compared to classical optimization methods, and with an increase in modeling time — exceed the technological parameters of the optimized system.
For example, the development by SPbPU specialists in the oil and gas sector will help to more accurately allocate resources performing geological and technical activities on the base fund and drilling of new wells, which will allow increasing oil production volumes. Currently, politicians are improving a prototype program for optimizing technological processes based on a multi-agent approach and are testing it on actual industry data.
Each managing agent in the modeled system has its own level of knowledge: some know the current state of wells and infrastructure better, while others, regional ones, see the whole picture as a whole, but less detailed. Each agent, based on its own data, proposes its work plan and options for resource exchange. Then they repeatedly coordinate decisions among themselves, gradually improving the result — this way both local and total production in the region increase. This is a feature of our development — instead of one centralized solution, we have agents working who themselves propose scenarios and negotiate among themselves.This results in more stable and realistic plans that solve the tasks of agents’ management at all levels,” explains the project leader, head of the “Digital Modeling of Industrial Systems” laboratory at the SPbPU IPSh, Alexey Gintsak.
Cooperation with industrial companies will allow politicians to develop a set of software solutions for optimizing technological processes based on a multi-agent approach, including a set of industry libraries that provide fast and efficient integration of the developed product into existing decision support systems to improve production management quality.
Please note; This information is unprocessed content obtained directly from the source of information. It represents an accurate report of what the source claims and does not necessarily reflect the position of MIL-OSI or its clients.