Article
Engineering, Geological
Tien-Thinh Le, Athanasia D. Skentou, Anna Mamou, Panagiotis G. Asteris
Summary: In this research, a series of artificial neural networks were trained and developed to predict the unconfined compressive strength of rock. Compiling a data and site independent database from 367 datasets, the input parameters used were Schmidt hammer number R-n, compressional wave velocity V-p, and effective porosity n(e). The study found that the ANN-ICA model had the highest accuracy.
ROCK MECHANICS AND ROCK ENGINEERING
(2022)
Article
Computer Science, Information Systems
Ana Caren Hernandez-Ruiz, Javier Alejandro Martinez-Nieto, Julio David Buldain-Perez
Summary: The SA-CNN-DC methodology proposed in this paper utilizes neural networks and clustering techniques for automated counting of steel bars, improving accuracy and efficiency in counting. The method offers various advantages in a steel warehouse, such as reducing counting time and resources, ensuring employee safety and productivity, and increasing confidence in inventory management.
Article
Engineering, Geological
Athanasia D. Skentou, Abidhan Bardhan, Anna Mamou, Minas E. Lemonis, Gaurav Kumar, Pijush Samui, Danial J. Armaghani, Panagiotis G. Asteris
Summary: This study examined the use of three artificial neural network (ANN)-based models to predict the unconfined compressive strength (UCS) of granite using three non-destructive test indicators. The ANN-LM model, constructed using the Levenberg-Marquardt algorithm, was determined to be the most accurate. In the validation phase, the ANN-LM model achieved the best predictive performance with R = 0.9607 and RMSE = 14.8272. The developed ANN-LM model outperformed existing models and a graphical user interface (GUI) was developed for easy estimation of UCS using this model.
ROCK MECHANICS AND ROCK ENGINEERING
(2023)
Article
Materials Science, Multidisciplinary
Qian Xie, Manu Suvarna, Jiali Li, Xinzhe Zhu, Jiajia Cai, Xiaonan Wang
Summary: In industrial steel plate production, a deep learning model was developed to predict mechanical properties based on process parameters and steel grade composition. Comparative analysis showed that the accuracy of the deep neural network model was higher than other classic machine learning algorithms. The tuned model was deployed in a real steel plant for online monitoring and control of steel mechanical properties.
MATERIALS & DESIGN
(2021)
Review
Nutrition & Dietetics
Jaroslaw Sak, Magdalena Suchodolska
Summary: Artificial intelligence is increasingly being utilized in biomedical, clinical, and nutritional epidemiology research in the fields of medicine and nutrition science. Different AI methods, such as artificial neural networks, machine learning algorithms, and deep learning algorithms, are applied in various aspects of nutrient studies. AI technology has the potential to revolutionize personalized nutrient supply and monitoring through the development of dietary systems.
Article
Chemistry, Physical
Beata Potrzeszcz-Sut, Agnieszka Dudzik
Summary: This paper presents a technique that combines traditional indentation tests with mapping the shape of the imprint for estimating material parameters and solving the inverse problem through neural networks. The proposed sparse model demonstrates efficient data reconstruction performance and consistently outperforms principal component analysis compression results in testing datasets.
Article
Engineering, Mechanical
Olga Ibragimova, Abhijit Brahme, Waqas Muhammad, Daniel Connolly, Julie Levesque, Kaan Inal
Summary: This study combines convolutional neural networks (CNNs) with the crystal plasticity finite element method (CPFEM) to propose a framework for rapid and accurate prediction of stress and strain in materials. The trained CNN shows excellent agreement with CPFEM simulations and can handle different materials and microstructures.
INTERNATIONAL JOURNAL OF PLASTICITY
(2022)
Article
Business
Maria Teresa Ballestar, Aida Garcia-Lazaro, Jorge Sainz, Ismael Sanz
Summary: This study aims to investigate the impact of companies' adoption of robotics on the workforce and provide decision support. The research identifies factors influencing the degree of adoption of robotics and develops a predictive model based on these factors. The study also conducts a characterization and segmentation analysis of companies that fail to correctly predict the degree of adoption.
JOURNAL OF BUSINESS RESEARCH
(2022)
Article
Engineering, Chemical
Paraskevi Karka, Stavros Papadokonstantakis, Antonis Kokossis
Summary: This research aims to develop a data-science based framework for estimating life cycle assessment (LCA) metrics of bio-based and biorefinery processes in the early design stages. The framework applies advanced analytics such as classification trees and artificial neural networks to improve the robustness and efficiency of LCA estimations.
CHEMICAL ENGINEERING SCIENCE
(2022)
Article
Engineering, Civil
Xianlin Wang, Yuqing Liu, Yuanchun Lu, Xuefeng Li
Summary: A new perforated web connection (PWC) for bridge application, which enhances the shear performance at the steel-concrete interface, was proposed and evaluated through push-out tests and numerical analysis. The results showed that PWC significantly improved the interfacial shear-slip response and transformed the failure mode from brittle to ductile. The parametric analysis indicated that PWC with large web openings had great ultimate shear resistance and sensitivity to other parameters.
ENGINEERING STRUCTURES
(2022)
Article
Energy & Fuels
Changsu Kim, Jiyong Kim
Summary: In this study, four different ANNs were used for predicting the performance of Pt-based catalysts in water gas shift reaction, with the multilayer perceptron model showing the best performance. It was demonstrated how selecting the optimal ANN structure can improve prediction accuracy and reduce computational load.
INTERNATIONAL JOURNAL OF ENERGY RESEARCH
(2022)
Article
Computer Science, Artificial Intelligence
Pablo Sanchez-Fernandez, Luis Gonzaga Baca Ruiz, Maria del Carmen Pegalajar Jimenez
Summary: In this study, different Machine Learning methods were applied to solve the problem of personality identification using a dataset labelled with the MBTI personalities. Comparing several algorithms, it was found that the classification approach outperformed the clustering methods with an average accuracy of around 90%. Finally, the model will be validated with the latest news about COVID-19 and the La Palma Volcano.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Materials Science, Coatings & Films
Damjan Hatic, Xiaoyin Cheng, Thomas Stephani, Markus Rauhut, Jan Gabler, Reinhold Bethke, Hunter King, Hans Hagen
Summary: This paper introduces a novel method for fine-scale thin-film hard coating adhesion classification by developing a set of features for standard parametrization. Research shows that these features are suitable for parametrizing standard's classes, and highlights that ISO standard requires both delamination and cracking features, whereas DIN requires only delamination.
SURFACE & COATINGS TECHNOLOGY
(2021)
Article
Acoustics
Donovan Birky, Joshua Ladd, Ivan Guardiola, Andrew Young
Summary: The study focuses on how data reduction tools can be applied to dynamic structural data, combined with clustering methods and artificial neural networks, to create a dynamic model for predicting the acceleration response of a structural system. The results demonstrate that mathematical morphology tools can reduce the dimensionality of time series data while preserving essential characteristics, and the use of ANNs shows promise as a surrogate model for dynamic response prediction.
JOURNAL OF LOW FREQUENCY NOISE VIBRATION AND ACTIVE CONTROL
(2022)
Article
Engineering, Civil
Mateusz Motyl, Lukasz Madej
Summary: The aim of this study is to develop an automated procedure based on machine learning for identifying pearlite islands in two-phase pearlitic-ferritic steel. Different quality input data and neural network architectures were used to find a method with high precision for recognizing pearlite islands. This approach removes the need for expert knowledge in interpreting image data and achieves low overtraining levels.
ARCHIVES OF CIVIL AND MECHANICAL ENGINEERING
(2022)
Article
Chemistry, Multidisciplinary
Alexey N. Beskopylny, Evgenii M. Shcherban', Sergey A. Stel'makh, Levon R. Mailyan, Besarion Meskhi, Irina Razveeva, Alexey Kozhakin, Diana El'shaeva, Nikita Beskopylny, Gleb Onore
Summary: The creation and training of artificial neural networks enable the identification of patterns and hidden relationships in the production of unique building materials, prediction of mechanical properties, and problem-solving in defect detection and classification.
APPLIED SCIENCES-BASEL
(2023)
Article
Mathematics
Ivan Panfilov, Alexey N. Beskopylny, Besarion Meskhi
Summary: The coronavirus infection SARS-CoV-2 affected around 500 million people in the beginning of 2022. This article presents a mathematical model to study the spread of viral particles in technological transport. By simulating the movement of liquid droplets in a flow, accounting for diffusion and evaporation, the propagation of viral particles is investigated.
Article
Materials Science, Composites
Alexey N. Beskopylny, Evgenii M. Shcherban, Sergey A. Stel'makh, Levon R. Mailyan, Besarion Meskhi, Andrei Chernil'nik, Diana El'shaeva
Summary: The aim of this study is to investigate the effect of heavy concrete manufacturing technology on the resistance of concrete to alternate freezing and thawing in an aggressive environment. The results showed that centrifuged and vibrocentrifuged variotropic concrete have greater resistance and endurance to cycles of alternate freezing and thawing compared to vibrated concrete.
JOURNAL OF COMPOSITES SCIENCE
(2023)
Article
Materials Science, Composites
Alexey N. Beskopylny, Evgenii M. Shcherban', Sergey A. Stel'makh, Levon R. Mailyan, Besarion Meskhi, Valery Varavka, Andrei Chernil'nik, Anastasia Pogrebnyak
Summary: The aim of this study was to develop improved structural foam concrete using fly ash and polypropylene fiber, and optimize the recipe technological parameters. The results showed that replacing cement with 10% to 40% fly ash can reduce the dry density of foam concrete, and samples with 10% fly ash replacement exhibited the best compressive strength, flexural strength, and thermal insulation properties.
JOURNAL OF COMPOSITES SCIENCE
(2023)
Article
Materials Science, Multidisciplinary
Alexey N. Beskopylny, Evgenii M. Shcherban', Sergey A. Stel'makh, Levon R. Mailyan, Besarion Meskhi, Alexandr Evtushenko, Diana El'shaeva, Andrei Chernil'nik
Summary: An urgent and promising direction in building materials science is to improve the quality of non-autoclaved aerated concrete by rational selection of composition and other recipe-technological factors. Complex compositions and technological solutions were explored to modify aerated concrete with various additives and reinforce it with environmentally friendly plant fibers.
Article
Construction & Building Technology
Yasin Onuralp Ozkilic, Ceyhun Aksoylu, Ibrahim Y. Hakeem, Nebi Ozdoner, Ilker Kalkan, Memduh Karalar, Sergey A. Stel'makh, Evgenii M. Shcherban, Alexey N. Beskopylny
Summary: The present study investigated the effects of transverse opening diameters and shear reinforcement ratios on the shear and flexural behavior of RC beams with two web openings of different spans. Twelve RC beams with various opening diameter-to-beam depth ratios and shear reinforcement ratios were tested until failure under four-point bending. The results showed that increasing opening diameter led to more pronounced frame-type shear failure, and the reductions in load capacity and modulus of toughness were more significant in the presence of inadequate shear reinforcement.
Article
Construction & Building Technology
Evgenii M. Shcherban', Sergey A. Stel'makh, Alexey N. Beskopylny, Levon R. Mailyan, Besarion Meskhi, Valery Varavka, Andrei Chernil'nik, Diana Elshaeva, Oxana Ananova
Summary: A current problem in the construction industry is the lack of complex, scientifically based technological materials and design solutions for universal types of building materials, products, and structures, especially in terms of structures operating under conditions of aggressive chloride exposure. The aim of the study was to compare and evaluate the differences in the durability of conventional and variotropic concretes made using three different technologies, vibrating, centrifuging, and vibro-centrifuging, modified with the addition of microsilica, under conditions of cyclic chloride attack. Vibro-centrifuged concrete showed the highest resistance to cyclic aggressive chloride exposure, while the use of microsilica as a modifying additive had a positive effect on the resistance of concrete.
Article
Engineering, Multidisciplinary
Alexey N. Beskopylny, Anton Chepurnenko, Besarion Meskhi, Sergey A. Stel'makh, Evgenii M. Shcherban', Irina Razveeva, Alexey Kozhakin, Kirill Zavolokin, Andrei A. Krasnov
Summary: The article discusses the development of a computer vision algorithm using the CNN YOLOv4 to detect water globules in oil samples and analyze their sizes. The algorithm is trained using an augmented dataset of microphotographs. The accuracy of the model is evaluated and found to be AP@50 = 89% and AP@75 = 78%.
Article
Chemistry, Multidisciplinary
Alexey N. Beskopylny, Evgenii M. M. Shcherban, Sergey A. Stel'makh, Levon R. Mailyan, Besarion Meskhi, Irina Razveeva, Alexey Kozhakin, Diana El'shaeva, Nikita Beskopylny, Gleb Onore
Summary: In recent years, machine vision algorithms have become widely used in industry for visual automatic non-destructive testing. This approach utilizes convolutional neural networks to detect, classify, and segment defects in building materials and structures. Implementing intelligent systems in the early stages of manufacturing can help identify and eliminate defective materials, prevent the spread of defective products, and determine the cause of specific damages.
APPLIED SCIENCES-BASEL
(2023)
Article
Chemistry, Physical
Evgenii M. Shcherban', Alexey N. Beskopylny, Sergey A. Stel'makh, Levon R. Mailyan, Besarion Meskhi, Alexandr A. Shilov, Elena Pimenova, Diana El'shaeva
Summary: There is currently a strong interest in using geopolymer composites as a sustainable alternative for restoring building facades. The study focused on developing geopolymer concrete with improved physical, mechanical, and adhesive properties. By adding ceramic waste powder (PCW) and polyvinyl acetate (PVA) as additives in optimal dosages, the geopolymer concrete showed enhanced strength, lower water absorption, and improved adhesion. The developed compositions are suitable for the restoration of building facades.