On 2–3 November, the Ural State University of Architecture and Art held the VI International Scientific and Practical Conference “New Information Technologies in Architecture and Construction” (NITAC-2023) .

PhD student at the SPbGASU Department of Information Systems and Technologies Yury Zgoda made a report at the conference “Features of the development of neural network web applications for calculating thin-walled shell structures,” for which he was awarded a first-degree diploma by the conference organizers.

The use of neural networks is one of the current areas of research in modern architecture and construction. Yury Zgoda examined the capabilities of neural networks in the context of modeling thin-walled shell structures. Based on the results of the study, high-performance software was developed that allows almost real-time calculations of cylindrical shells via a web browser. The application does not require expensive hardware, making it easy to integrate into teaching or engineering activities. The ability to interact via a web browser makes the application cross-platform (allows it to be used via mobile devices and computers running different operating systems), and also simplifies the setup process (the web application only needs to be deployed once on the server, after which all users on the network have access to it).

Elena Maznyak, a second-year master's student at the SPbGASU Faculty of Environmental Engineering and Municipal Services, gave a presentation on “Application of information modeling technologies in landscape design.” Elena works as a landscape architect and is mastering new competencies at the university.

According to the author, landscaping and planting are an integral part of almost any architectural and construction project. Currently, most landscape projects are carried out separately from the information model of the construction site (IM OS), using 2D design and 3D visualization software. This approach does not correspond to the concept of TIM as a systematic approach, since it does not provide a set of interrelated information about the object.

Elena Maznyak noted that an important characteristic of the TIM approach and software for TIM is the unified information and coordinate space, the general structure of model classifiers and the settings of general parameter files. This allows to compare, analyze and evaluate the feasibility of design decisions at early stages, including the construction and landscape sketch stages.

Software for TIM also provides the ability to analyze, automate and manage data, including programming tools. This opens up a wide range of opportunities for optimizing work on a landscape project in terms of the functional organization of space. The functional organization of space at an urban improvement facility is a search for the optimal location and type of improvement elements based on a set of specified restrictions, norms and rules, the display of which in the OS IM will allow many collisions to be avoided. In addition, the customer will have the opportunity to check the viability of the project, doing this at the sketch stage, and not during or after its implementation.

In her report, Elena Maznyak proposed methods and means of mathematical modeling, solving optimization problems and simulation modeling in relation to specific and relevant aspects that arise when designing landscape projects.