Artificial intelligence in construction began to be implemented relatively recently, starting in the early 2010s. Initially, technologies focused on solving narrow tasks such as structural calculations, construction data analysis, and automating project documentation. Over time, algorithms began to integrate into all stages of construction.
The first experiments with AI in design were related to optimizing architectural solutions. For instance, the use of genetic algorithms allowed for finding the best construction options in terms of strength and cost minimization. Visual recognition and data analysis began to be utilized for quality control on construction sites.
Currently, the capabilities of artificial intelligence have significantly expanded. It is not only capable of designing buildings but also predicting maintenance needs, identifying potential errors, and even suggesting optimal ways to implement projects.
According to Petr Manin, modern AI systems successfully tackle numerous tasks. Among them are modeling and optimizing projects: AI helps assess the viability of architectural concepts at early stages, taking into account various parameters, including soil type, climatic conditions, material costs, and future building loads.
Another task is quality control in construction: visual recognition systems analyze photo and video data from cameras on construction sites, comparing actual construction processes with digital models, quickly identifying deviations from project norms.
For the operation of facilities: AI is used for monitoring buildings, with the data collected being used to predict the need for repairs or maintenance. Finally, in managing construction processes: AI optimizes logistics for material delivery, reduces the risk of errors in planning, and allows for more accurate calculations of work completion timelines.
A significant aspect of the current stage of development in the construction industry is robotic design. According to an expert, this direction allows for the automation of a large portion of project processes while maintaining flexibility and variability.
“We are moving towards a scenario where humans will set parameters for the machine. Based on these parameters, the machine will generate project options. Moreover, these options will be shared very quickly and in a highly variable manner, ultimately providing us with a fully ready project at the model level, at the level of drawings,” said Petr Manin.
An important achievement is that some projects completed by robots are already undergoing certification alongside those created by humans. According to the business development director of the NTs "Platform," experts could not distinguish between the drawings made by machines and those developed by professional architects.
“Let’s consider a specific example. At the building layout stage, the robot offers a large number of layouts, including those with shifts, with different axes, and varying efficiency coefficients for usable area. We choose and move forward,” explained the speaker.
He noted that AI-powered platforms can generate detailed building models in just a few minutes. For example, one of the presented technologies allows for the complete creation of project documentation for simple objects (such as residential towers) in less than 10 minutes. Importantly, these systems design several solutions for a single object simultaneously, taking into account parameters of economic efficiency, ergonomics, and architectural aesthetics. From these, a specialist can then make a selection.
“We obtain results, work backward, reach the parameter-setting stage, adjust the parameters to more optimal ones, and move forward,” added Manin.
He clarified that robots not only create basic architectural plans but also integrate furniture, engineering networks, and facade solutions into them.
As noted by the business development director of NTs "Platform," in the next four years, robotic design technologies will be actively used to create projects for standard objects, such as apartment buildings and office buildings. Fully automated design of these objects will reduce development times by 2 to 3 times.
“By 2030, we expect that all projects for standard housing will be completed by machines. They will no longer distract people from creativity, which encompasses all unique, complex objects, especially those built within city limits — in Moscow and St. Petersburg. Architects will continue to create. The same applies to engineers. They will continue to work on complex industrial objects,” the speaker predicted.
At the same time, he emphasized that humans will need to oversee the work of AI and must correctly set tasks. Such new challenges will require a reassessment of approaches to management and work organization. An important issue will be retraining personnel, adapting specialists to new tools and technologies.
In the near future, robotic design will reduce construction costs, increase work speed, and minimize the risk of errors. However, alongside this, new challenges will arise: the need for technology standardization, working with large volumes of data, and finding a balance between automation and preserving jobs.
The advancement of construction technologies has progressed so far that specialists at the Moscow Aviation Institute are contemplating the creation of a lunar 3D printer. Unlike terrestrial models, it will operate using solar energy. A special parabolic mirror with a diameter of 2.5 meters will be installed in the energy concentrator. This mirror will focus sunlight, which will then sinter lunar soil. Following this, layer-by-layer manufacturing of blocks for living modules will commence. If the research is successful, blocks for creating long-term habitable lunar stations could be printed directly from regolith on the Moon, eliminating the need to transport heavy construction materials from Earth.
Currently, MAI is completing mathematical modeling of the lunar concentrator's operation. At the end of 2023 — beginning of 2024, experiments are planned to be conducted on a powerful industrial 3D printer at the Moscow Aviation Institute's "General Engineering Training" base.
Such progressive developments, like the technology for sintering lunar soil, demonstrate that AI can transcend conventional boundaries, opening up possibilities for construction on other planets. On Earth, these technologies continue to assist architects, engineers, and construction companies in creating higher quality, safer, and more accessible buildings.
However, the success of technology implementation depends on the readiness of professionals to adapt to changes, as well as on the development of a new regulatory framework that integrates digital standards. Experts are confident that in the future, humans and machines will work hand in hand, creating a world where innovation and creativity unite for the sake of progress.
The material was prepared with the support of the Ministry of Education and Science of Russia.