Over the last few years, Deep learning has begun to play an important role in analytics solutions of big data. Deep learning is one of the most active research fields in machine learning community. It has gained unprecedented achievements in fields such as computer vision, natural language processing and speech recognition. The ability of deep learning to extract high-level complex abstractions and data examples, especially unsupervised data from large volume data, makes it attractive a valuable tool for big data analytics. In this paper, we review the deep learning architectures which can be used for big data processing. Next, we focus on the analysis and discussions about the challenges and possible solutions of deep learning for big data analytics. Finally, have been outlined several open issues and research trends.
Journal section: TECHNOLOGICAL BASIS OF “INDUSTRY 4.0”
The peculiarities of carrying out the basic methods of solving the problems of identification, which have become widespread in transport technologies, are considered. Three approaches were tested: a deterministic approach based on the method of selected points, a statistical approach that includes the least-squares method, and – an inductive approach – a method of group accounting of arguments. A number of experiments have been conducted, which is based on structural optimization of the model. In the process of analysis of the obtained models, a rational model of setting and solving identification problems in transport technologies of optimal complexity was obtained. This systematic flow diagram can be successfully applied to solve complex analytical tasks.
Automated production becomes a new reality of the industry, one of the important points in the construction of such production is a more accurate and quick design of technological processes for processing and assembly of parts. Analysis of computer-aided design systems shows the inadequacy of their capabilities to achieve this result. Significant improvement of CAD / CAM systems is required, for which it is necessary to establish close relationships between the part, tool, machine tool, technology, as well as digitize world experience and build an advanced algorithm that independently takes many different decisions, such as grouping part elements into operations and grouping transitions for their processing.
In the last decade, cloud computing has become an extremely important business assistant that offers significantly lower costs than traditional computer-aided resource provision and greater adaptability to business changes. Thus, it helps them to achieve their business goals. The development of cloud computing poses significant risks, which serve as a partial barrier to their use by the businesses. Security and protection of information are considered to be one of the most critical points in the use of cloud computing. Compared to traditional solutions, cloud computing moves application software and databases to large data centers where information and service management are not always secure and reliable. As a result of this, the community and businesses have many concerns and questions about the security of data and applications in cloud infrastructures. In addition, concerns about data security and applications also arise due to the fact that both user data and applications are managed by the cloud provider. Clouds typically use a standardized data and application security architecture, while the demand for consumer and business services is steadily growing and becoming more diverse and with more sophisticated software design that leads to continuous upgrades to data security models in cloud computing. Due to the different methods of implementation in cloud computing service delivery models, the demand for a reasonable level of data protection is of utmost importance. The purpose of this publication is to propose a new information security model, which offers a solution for improving the use of sensitive data by introducing a three-factor authentication – an improvement of preventive control.
The development of augmented reality (AR) technologies and the capabilities of modern mobile devices has led to the emergence of a new class of AR mobile navigation systems. One of the basic requirement of such applications is device localization and precise mapping of the route and other visual objects to observed environment, which requires high accuracy of positioning and orientation.
This paper addresses the problem of localization precision in AR navigators and proposes general concept of improving accuracy of basic GPS-based methods by augmenting them with local device information about its position. Some algorithms and experimental results are provided to demonstrate benefits of the proposed method.
The paper describes the definitions of industry 4.0 and the factors that influence their formation. It describes the areas
that it affects and which are involved in the formation of a new generation industry. Paper also discusses problems that arise in areas related to industry 4.0 when combining them and possible ways to solve them, especially the organization of the production process. Making a focus on properties of software for industrial devices.
RULES FOR CHOOSING COMPONENTS OF A SINGLE DIGITAL PASSPORT FOR THE GENERATION OF CAD DESIGN SOLUTIONSpg(s) 14-15
In previous articles, it was shown that for “Industry 4.0” production one of the important tasks is the collection, storage, analysis and processing of product data and related design and production procedures. This determines the implementation of PDM, ERP, MES and/ or EAM systems at enterprises and the construction of a metasystem based on them – a single digital passport. Its content allows to automate the generation of design decisions for timely management decisions, as well as to propose an approach to solve the problem of minimizing the impact of the human factor for the tasks of Industry 4.0. Therefore, the work presents the results of the development of the rules for choosing the components of a single digital passport for the generation of design solutions based on the principle of building a fuzzy Takagi-Sugeno-Kang network. The results obtained will allow to formulate a criterion for the similarity of the production object and the design and production procedure in the digital environment to the given requirements based on linear approximation methods.
Monitoring and control of Cyber-Physical Systems (CPS) have many challenges related to the heterogeneous environment, the high degree of interaction between the components and the high requirements for functionality and scale. The paper presents an analysis of the state of the art in this area and proposes and analyses different combined approaches for control of CPS based on IEC-61499 standard. The approaches are divided in two groups. The first group of approaches combines IEC-61499 standard with advanced methods of software engineering, such as formal methods and model-driven development, based on UML/SysML. The second group uses other standards as IEC- 61512 and PLCopen to be combined with IEC-61499 standard in order to improve the development lifecycle of cyber-physical systems.
Wet wipe machines are machines for high capacity production at high speeds. Even the short downtimes of the machines in the production area lead to huge losses for manufacturers. However, in these machines; unexpected breakdowns such as belt breakage, bearing damage, sensor motor failures, pneumatic air leaks are frequently encountered. In order to prevent unwanted stoppages, a sensor, actuator and plc based application has been implemented to prevent failure parameters and to detect faultlessly by monitoring the operation parameters. As a result of the studies carried out, our wet wipe machines that we produce have been adapted to the industry 4.0 concept.
New generation of digitized factories in Industry 4.0 is based on a combination of cyber-physical systems (CPS) and digital technologies, to provide an integrated and intelligent synergy in production process, linking the physical, digital and virtual world. Industry 4.0 uses production systems, CPS, CPPS, SCPS and robots in combination with digital technologies IoS, IoP, IoT, IIoT, IoE, CC and etc. The digital infrastructure of a country can be evaluated using different development indices, such as: IDI, EGDI, TII, OSI and etc. The paper is given of trend analyzes of development indices (DI) for using ICT and the Internet., EGDI and TII in Georgia for the period 2003 to 2018.
The goal of this paper is to study algorithm for detection and classification of moving objects based on parameter estimates from their GPS signal shadow. The suggested algorithm uses satellite GPS signals to create radio barriers and to detect moving terrestrial targets. It can be examined as an algorithm for secondary application of the wireless technologies (SAWT). This technology is extremely modern and up-to-date and can find application in Industry 4.0 with regard to the reduction of the electromagnetic radiations. The GPS signals included shadow from several moving vehicles, processed in MATLAB environment in order to obtain the estimations of their parameters. In IBM SPSS Statistics, Scheffe test is used for classifying the moving objects. The results reveal that proposed signal processing of the GPS signal shadow combined with this statistical approach can be successfully applied in practice for object classification.
The article deals with current situation of design of production lines in modern factories and describes possibilities of realization new type of industry lines. The main research deals with proposal of modular production system established from intelligent manufacturing cells. Main positive features of suggested structure are possibility of rapid reconfigurability of the production line, easy maintenance and possibility of simple and fast design of new production lines, thanks to modularity. To achieve these features in the Industry 4.0 ecosystem is necessary to use of modern communication methods and technologies, such as wireless industrial communication and IoT together with cooperative robots. Moreover, the entire system has to be designed as hierarchical, demand synchronized, flexible and is using standardized construct components.