Translartion. Region: Russians Fedetion –
Source: Peter the Great St Petersburg Polytechnic University – Peter the Great St Petersburg Polytechnic University –
From predicting the quality of glass and plant selection to accounting for construction projects – this is the range of applications of artificial intelligence technologies that was considered by participants in the latest seminar at SPbPU.
The project managers, Marina Petrochenko, Director of the Institute, and Pavel Nedviga, Senior Lecturer, spoke about the software package “IMPULSE” created at the Polytechnic Institute of Civil Engineering. According to Pavel Nikitich, the package has already been included in the register of Russian software and is successfully used in business for the automated classification of elements of the information model of buildings in accordance with the legislation of the Russian Federation. “IMPULSE” uses artificial intelligence methods and is a constantly learning model that works with an accuracy of up to 98%. Marina Vyacheslavovna emphasized that thanks to this development, many hours of work by specialists have been replaced by three minutes of machine work, and a person has thus been freed up to solve other problems.
Associate Professor of the Higher School of Physics and Materials Technology of the Institute of Mechanical Engineering, Materials and Transport Viktor Klinkov spoke about the developed model for predicting infrared optical systems. Scientists wondered what properties glass of a particular chemical composition would have, and are creating a neural network to collect and process data. The traditional method of determining the composition of glass by trial and error is labor-intensive, time-consuming and risky for industrial use. But ML (machine learning) and AI (artificial intelligence) tools for predicting materials have proven to be successfully applicable to glass, since there is a large volume of data on existing compositions, there are no restrictions on stoichiometry – the composition of glasses can be continuously changed and there is a wide demand for compositions with a special combination of properties. As Viktor Artemovich noted, the created model allows predicting properties and accelerating modeling, selecting an “individual design” for subject areas, and obtaining new knowledge about the nature of glasses at the molecular level.
Director of the Scientific and Educational Center “Nanotechnologies and Coatings” Alexander Semencha supplemented his colleague’s speech by talking about the use of AI in the design of optical devices. Alexander Vyacheslavovich demonstrated a microspectrometer that identifies the properties of materials by making a comparative analysis with the collected database. The device was tested on food products, fabrics, paper and showed almost 100% results.
A wide range of tasks are solved by neural networks in genomic and post-genomic technologies of agrobiology. This was reported by Maria Samsonova, professor of the Higher School of Applied Mathematics and Computational Physics of the Physics and Mechanical Institute. Neural networks trained on the collected data help to predict the properties of plants, determine the best areas for their cultivation, and identify predisposition to diseases. Scientists have conducted many experiments and collected datasets for such plants as flax, chickpeas, rye, etc. For example, using a neural network that extracts and selects the characteristics of different chickpea samples, a panel of markers for breeding this crop was created. According to Maria Georgievna, one of the important tasks today is the development of standards for collecting, processing, and storing information.
Summing up the results of the seminar, Vice-Rector for Research at SPbPU Yuri Fomin drew attention to the fact that over the course of five seminars it was possible to hear reports from almost all the leaders of scientific groups involved in technologies in the field of artificial intelligence, and it is time to expand the circle of participants.
I believe that for the further development of our scientific activities it will be important and useful to listen to the opinions of representatives of various industries on the use of artificial intelligence in practice, to understand their needs. I invite everyone to take part in this conversation, which will take place on February 26, and actively ask questions to our partners whom we will invite, – urged Yuri Vladimirovich.
Please note: This information is raw content directly from the source of the information. It is exactly what the source states and does not reflect the position of MIL-OSI or its clients.