bg
Industry and import substitution
10:00, 21 November 2026
views
4

YASON Power Modules With Artificial Intelligence Enter Mass Production

St. Petersburg-based Russian Electrotechnical Society LLC (REO) has launched mass production of a new generation of YASON power modules with built-in artificial intelligence. The technology enables predictive diagnostics and automated control of industrial equipment.

REO’s latest development is built around power cells from the YASON family, designed for power conversion equipment. Each unit contains three semiconductor assemblies featuring insulated-gate bipolar transistors (IGBTs) and freewheeling diodes, which together form a power half-bridge. These components handle high currents and voltages, making them well suited to high-power inverters and industrial electric drives.

The platform has already demonstrated its reliability under real-world operating conditions. In late 2024, REO equipped a research vessel under construction as part of Project 17050 with converters based on the platform. In 2025, the vessel’s 1.2 MW electric propulsion system successfully completed testing. Then, in summer 2026, the modules were deployed in an energy router for a microgrid at a research center operated by PJSC RusHydro on Russky Island. The proven hardware platform now incorporates AI algorithms.

New Algorithms

The updated module analyzes deviations in operating parameters in real time, enabling adaptive control and protection of power systems. Its embedded AI tracks changes in equipment performance, helping prevent failures before they cause component damage and predicting potential faults. The hardware uses EtherCAT as its primary high-speed control channel, with CAN as a backup. The modules are available in 400 V and 690 V versions, with rated phase currents of 600 A and 300 A, respectively. REO plans to use this technology as the basis for developing variable-frequency drives, inverters, energy storage systems, charging stations and uninterruptible power supplies.

Greater Autonomy and Faster Response

Analyzing equipment condition through electrical signals is an established global practice. In 2023, Siemens AG launched predictive maintenance services. In Russia, researchers at the Higher School of Economics (HSE University) unveiled an equipment failure prediction system in October 2026. The system completed pilot industrial testing at the Volodarsky Waterworks facility in the Leningrad Region, operated by JSC Leningrad Regional Utility Systems (LOKS).

REO’s approach differs in that its algorithms run directly inside the power cell, using edge computing. The module analyzes data and makes decisions independently, without sending large volumes of information to an external system or the cloud. This enables an immediate response and allows the equipment to operate independently of network connectivity quality.

A New Product Category

For Russia’s IT industry, the REO project points to the emergence of a new product category: sophisticated hardware-software systems for industrial applications, with algorithms integrated directly into power electronics. The development expands the capabilities of intelligent power conversion equipment and enables the creation of industrial Internet of Things (IIoT) solutions that do not depend on imported components.

Russia’s IT industry is gaining a practical tool for modernizing the real economy, where system response times and autonomy matter more than simply collecting data.

Reducing Downtime in Critical Industries

REO’s solutions are aimed at industries where a drive failure can lead to prolonged downtime, including power generation, shipbuilding, oil and gas, manufacturing, data centers, utilities and transportation infrastructure. Deploying these systems is expected to improve equipment reliability and reduce the duration of unplanned technical downtime.

Industrial automation remains the most promising application area, as companies move toward condition-based maintenance. Research by Izhevsk State Technical University into the use of machine learning to predict the remaining useful life of electric drives further underscores the relevance of this approach.

Flexible Configuration as a Competitive Advantage

Success in these sectors could pave the way for REO to expand its exports. The most realistic target markets are the Eurasian Economic Union (EAEU) member states and countries investing in the modernization of their power and transportation infrastructure. Entering these markets will require international certification and the development of a service network. However, the ability to tailor software configurations to individual customer requirements gives the company a competitive advantage.

Over the next two years, REO will focus on its first commercial deliveries of modules with AI-powered diagnostics and on collecting operational data. By 2028, predictive analytics based on the YASON platform could already be deployed in marine power systems, energy storage facilities and data centers.

Industrial companies are increasingly adopting predictive analytics and AI-powered machine vision to reduce the risk of accidents at their facilities and improve production efficiency. These technologies are cost-effective while also making operations more reliable.
quote

like
heart
fun
wow
sad
angry
Latest news
Important
Recommended
previous
next