Abstract. The article examines the potential application of neural network technologies and computer vision to automate the standard control process for design documentation. Standard control, a critical stage in engineering design, traditionally demands substantial time, labor resources, and a high level of specialist expertise. Modern deep learning methods offer the potential to significantly reduce these demands, eliminate human factor influence, and minimize the risk of errors. The study highlights algorithms that ensure the verification of design drawings against regulatory standards such as GOST and ESKD. A concept for an automated standard control system (ASC) is introduced, which integrates with CAD systems and incorporates stages such as data preprocessing, image analysis, comparison with reference standards, and generating reports on detected discrepancies.

Keywords: neural networks, computer vision, design documentation

For citation: Penshin I. S. On the possibility of applying neural network technologies to optimize standard control of design documentation in defense industrial organizations. Aerospace Engineering and Technology. 2024. Vol. 2, no. 4, pp. 138–148. DOI 10.52467/2949-401X-2024-2-4-138-148. EDN APKGVE (In Russian)