Information model of the algorithm for the application of machine-readable normative documents in the design of building structures
https://doi.org/10.37538/0005-9889-2025-2(627)-16-26
EDN: ZHVRBU
Abstract
Introduction. In modern conditions, the design of reinforced concrete structures requires prompt processing of a large amount of regulatory data presented in text format. The task of automatization of the extraction and analysis of information from regulatory documents is becoming urgent, as it makes it possible to increase the accuracy and efficiency of design work.
Aim. Development of the information model integrating algorithms for processing regulatory documents (using the example of SP 63.13330.2018) with design systems.
Materials and methods. The Mistral language model, deployed locally through Ollama, is used to process text data. Key entities (material parameters, loads, requirements) were extracted automatically, and their relationships were visualized in a graph database. The integration of the calculation and design systems is carried out using the IFC and XML formats. The formulae are implemented (for example, the calculation of the length of the reinforcement anchoring) in Python, and the automation of reinforcement selection is carried out in the Revit environment.
Results. The application of the proposed algorithm has made it possible to automate the extraction of material parameters, design characteristics and design requirements, which significantly reduces the complexity of routine operations. The integration of the calculation model with the design environment (Revit) has provided automatic selection of reinforcement and welded meshes, contributing to increased design accuracy and reducing the likelihood of errors.
Discussion. The developed system demonstrates high efficiency in automating the processing of machine-readable regulatory documents, which leads to a reduction in design time and an improvement in the quality of construction work. Further research will be aimed at expanding the algorithm to work with other types of documents and adapting the model to specific engineering analysis tasks.
Conclusions. Machine-readable documentation formats and LLM models increase the accuracy and speed of processing of the regulatory data. Integration of the calculation and design systems via IFC/XML reduces labor costs and errors. Automation of reinforcement selection demonstrates the potential for scaling to other design stages.
About the Authors
S. V. SnimshchikovRussian Federation
Sergey V. Snimshchikov, Cand. Sci. (Engineering), Vice-rector
for E and APE, Moscow State Technical University of Civil Aviation, Moscow
e-mail: s.snimshikov@mstuca.ru
tel.: +7 (499) 459-04-90
I. P. Savrasov
Russian Federation
Ivan P. Savrasov*, Cand. Sci. (Engineering), Assistant to the Vice-rector, Moscow State Technical University of Civil Aviation, Moscow
e-mail: i.savrasov@mstuca.ru
tel.: +7 (499) 452-47-60
E. V. Sumarokov
Russian Federation
Evgeny V. Sumarokov, Head of Information Technology Department, Research Institute of Concrete and Reinforced Concrete named after A.A. Gvozdev, JSC Research Center of Construction, Moscow
e-mail: evgeni.sumarokov@yandex.ru
tel.: +7 (495) 602-00-70
References
1. Documentation on the Ollama tool. URL: https://ollama.ai (accessed: 24.12.2024). (In Russian).
2. Information about the Mistral language model. URL: https://mistral.ai (accessed: 24.12.2024). (In Russian).
3. SP 63.13330.2018. Concrete and reinforced concrete structures. General provisions. Updated version of SNiP 52-01-2003. Moscow: Stroyizdat Publ., 2018. (In Russian).
4. Mahamud I., Asif A.A., Rahman Md.M., Sarker S., Shome A.R. Python eats python: A Burmese Python (Python bivittatus) preying on a Reticulated Python (Malayopython reticulatus) in Bangladesh. <i>Reptiles & Amphibians</i>. 2024, vol. 31, no. 1, pp. e21418. DOI: https://doi.org/10.17161/randa.v31i1.21418
5. Lyannik V.A. Comparative analysis of methods for calculating the value of reinforcement anchorage for bending elements. <i>Innovatsionnaya nauka</i>. 2019, no. 4, pp. 215–217. (In Russian).
6. Bazhenov Yu.M. Reinforced concrete structures: Textbook for universities. Moscow: ASV Publ., 2017, 512 p. (In Russian).
7. Setkov V.I. Design of reinforced concrete structures. Moscow: Infra-Engineering, 2020, 368 p. (In Russian).
8. State Standard R 57361-2016/ISO 29481-1:2010. Building information modelling. Part 1: Methodology and format. Moscow: Standartinform Publ., 2017. (In Russian).
9. Timoshenko S.P. Method of finite elements in building mechanics. Moscow: Librocom, 2015, 280 p. (In Russian).
10. Eastman C. BIM Handbook: A Guide to Building Information Modeling. Wiley, 2018, 688 p.
11. Khemlani L. Interoperability in BIM: Challenges and Opportunities. AECbytes, 2021. URL: https://www.aecbytes.com (accessed: 10.10.2023).
12. ISO 16739-1:2018. Industry Foundation Classes (IFC) for data sharing in the construction and facility management industries. Geneva: ISO, 2018, 54 p.
13. Autodesk Revit 2022: The official guide. Moscow: DMK Press, 2021, 480 p. (In Russian).
14. Smith J. XML-Based Data Exchange in Civil Engineering. Springer, 2020, 198 p.
Review
For citations:
Snimshchikov S.V., Savrasov I.P., Sumarokov E.V. Information model of the algorithm for the application of machine-readable normative documents in the design of building structures. Concrete and Reinforced Concrete. 2025;627(2):16-26. (In Russ.) https://doi.org/10.37538/0005-9889-2025-2(627)-16-26. EDN: ZHVRBU