Against Russian Standards
Students in the Information Technology program at the Kalashnikov Academy have developed software that automatically checks engineering documentation for compliance with Russia's GOST standards. According to the developers, no comparable end-to-end solution currently exists either in Russia or internationally.
Experts expect the application to reduce the time required to review engineering and manufacturing documentation, lower the number of human errors and ease the workload for compliance specialists responsible for standards verification. The project was supervised by engineers from the Izhevsk Mechanical Plant (JSC IMZ), part of the Kalashnikov Concern. The software is also expected to become the foundation for broader AI adoption at the plant.
What Can the New Software Do?
The application verifies whether engineering drawings comply with GOST standards and the Unified System for Design Documentation (ESKD), while automatically identifying formatting and design errors. It then generates reports highlighting deviations from established standards for drawing layout and documentation, along with recommendations for correcting them.
Those reports allow design and manufacturing engineers to identify issues quickly and update both drawings and related manufacturing documentation. Another major advantage is that the software operates entirely within a closed local environment without requiring an internet connection, significantly reducing the risk of sensitive engineering data leaks.
The application is built with modern Python-based algorithms and uses computer vision and data analysis libraries. It supports PDF engineering drawings, accurately recognizes text, dimensions, lines and hatching patterns, and automatically checks documents for compliance with GOST and ESKD requirements.
Built Around Standards
The new tool is expected to accelerate reviews of engineering drawings and manufacturing documentation against GOST and ESKD requirements. In turn, that should improve documentation quality while minimizing human error. Also, the program is expected to shorten review times for engineering and manufacturing documentation, reduce the workload for compliance specialists, lower the number of human errors and speed the preparation of new products for manufacturing. Over time, the technology could help manufacturers bring industrial products to market faster, improve product quality and reduce production costs.
But for now, the software is still under development. Its creators are working on a secure document repository and additional information security capabilities. Looking ahead, the application is expected to be valuable for Russian manufacturers because standards compliance reviews remain a labor-intensive process that directly affects production readiness.
Designed for Secure Industrial Environments
The developers plan to deploy the software at the Izhevsk Mechanical Plant and later across other Kalashnikov Concern facilities. If the application demonstrates consistent accuracy and performance when processing large collections of real engineering documentation, it could be adapted for companies in mechanical engineering, instrument manufacturing, aerospace, shipbuilding and the defense industry.
One of the software's most important advantages is its ability to operate entirely inside a closed domain. Many industrial companies cannot use external cloud-based AI services because engineering documentation is subject to strict security requirements. Local deployment could therefore become a significant competitive advantage for the product in the Russian market.
In the future, the software could also be exported beyond Russia, primarily to countries within the Eurasian Economic Union (EAEU) and the Commonwealth of Independent States (CIS). Entering additional international markets would require adapting the application to each country's national engineering standards.
