NCM 

Publication

White papers and scientific publications

 

Industry 4.0. Predictive Analytics System

Assoc. Prof. Bezobrazov S. PhD., M.Sc. Anfilets S., Mr. Fabrizio G., Mr. Fabrizio F., Mr. Fabrizio L.

Abstract: Industrial Predictive Analytics for Industry 4.0 is a system that predicts and prevents machine failures and breakdown by analysing time-series data (temperature, pressure, vibration etc.) received from sensors embedded in machines and equipment. The system can analyse machine parameters to identify patterns and predict breakdowns before they happen. The core of the proposed system is based on the Artificial Neural Network approach (both Deep and Shallow Neural Networks). Artificial Intelligence and Artificial Neural Networks allow the analyses the huge amounts of data collected from the manufacturing processes and predict what will go wrong, and when. The proposed system works in the paradigm of Industry 4.0 and provides the abilities in the area of Predictive Maintenance. The Industrial Predictive Analytics for Industry 4.0 also contains a decision-making system and support system that significantly increases the level of maintenance.

Keywords: INDUSTRY 4.0, SMART FACTORY, PREDICTIVE MAINTENANCE, ARTIFICIAL INTELLIGENCE


                                                                                                                                                                                               Download white paper  

 


Deep Multilayer Neural Network for Predicting the Winner of Footbal Matches

Anfilets S., Bezobrazov S., Golovko V., Sachenko A et al.

Abstract: In this work, we draw attention to prediction of football (soccer) match winner. We propose the deep multilayer neural network based on elastic net regularization that predicts the winner of the English Premier League football matches. Our main interest is to predict the match result (win, loss or draw). In our experimental study, we prove that using open access limited data such as team shots, shots on target, yellow and red cards, etc. the system has a good prediction accuracy and profitability. The proposed approach should be considered as a basis of Oracle engine for predicting the match outcomes.

Keywords: soccer analytics, forecasting, neural networks, Deep Elastic Net, open access dataset, data preprocessing, artificial intelligent.


                                                                                                                                                                                               Download white paper  

 


Artificial Intelligence for Sport Activity Recognition

Bezobrazov S., Sheleh A., Kislyuk S., Golovko V., Sachenko A et al.

Abstract: This paper presents and explains an implementation of an Artificial Neural Network approach for sport activities (gestures) detection and recognition using PIQ ROBOT device. Tennis was chosen as an example of sports activities. The development of artificial intelligence has given rise to gesture-recognition-based devices. The global gesture recognition market size was valued at USD 6.22 billion in 2017 and it is likely to reach USD 30.6 billion by 2025. This paper starts our ambitious research in the area of artificial neural networks for activity recognition in the sport.

Keywords: gesture recognition; sport activity recognition; artificial intelligence; artificial neural networks; time series processing.


                                                                                                                                                                                               Download white paper  

 


Development of Solar Panels Detector

Golovko V., Kroshchanka A., Bezobrazov S., Sachenko A., Komar M

Abstract: The paper describes the method of detection of roof-installed solar photovoltaic panels in low-quality satellite photos. It is important to receive the geospatial data (such as country, zip code, street and home number) of installed solar panels, because they are connected directly to the local power. It will be helpful to estimate a power capacity and an energy production using the satellite photos. For this purpose, a Convolutional Neural Network was used. For training and testing dataset consists of low-quality Google satellite images was used. The experimental results show a high rate accuracy of detection with low rate incorrect classifications of the proposed approach. The proposed approach has enormous implementation and can be improved in future.

Keywords: convolutional neural network; solar panels detection; satellite photos; geospatial data; power capacity; energy production, artificial intelligence, computer vision.


                                                                                                                                                                                               Download white paper  

 


High Performance Adaptive System for Cyber Attacks Detection

Komar M., Kochan V., Dubchak L., Sachenko A., Golovko V., Bezobrazov S.

Abstract: To increase the security of intrusion detection system, generalized structure of highly performance adaptive system for cyber attacks detection was developed. To improve its robustness, methods of artificial intelligence were proposed. Neural immune detectors were used as the main tool for identifying cyber attacks. These detectors for cyber attacks identification and classification and other vulnerable subsystems were implemented in programmable logic arrays. To provide high performance, the Mamdani fuzzy inference rules were used and relevant subsystem structures were developed.

Keywords: intrusion detection system, high performance adaptive system, cyber attacks, artificial intelligence, neural immune detector.


                                                                                                                                                                                               Download white paper  

 


Cryptography Using Artificial Intelligence

Blackledge J., Bezobrazov S., Tobin P.

Abstract: This paper presents and discusses a method of generating encryption algorithms using neural networks and evolutionary computing. Based on the application of natural noise sources obtained from data that can include atmospheric noise (generated by radio emissions due to lightening, for example), radioactive decay, electronic noise and so on, we `teach' a system to approximate the input noise with the aim of generating an output nonlinear function. This output is then treated as an iterator which is subjected to a range of tests to check for potential cryptographic strength in terms of metric such as a (relatively) large positive Lyapunov exponent, high information entropy, a high cycle length and key diffusion characteristics, for example. This approach provides the potential for generating an unlimited number of unique Pseudo Random Number Generator (PRNG) that can be used on a 1-to-1 basis.

Keywords: coding and encryption; artificial intelligence; multiple algorithms; personalised encryption engines; artificial neural networks; evolutionary computing.


                                                                                                                                                                                               Download white paper