Information Technology and Society

MONITORING MACHINE LEARNING MODEL DRIFT IN PRODUCTION PIPELINES: METHODS, METRICS, AND DEPLOYMENT CONSIDERATIONS

Maryna BAUTINA — Wed, 28 May 2025 00:00:00 +0300

The relevance of the study is determined by the need to ensure the stability and effectiveness of machine learning models in the context of dynamic changes in data. The problem of model drift – changes in the statistical characteristics of input data or relationships between features and the target variable – leads to a decrease in prediction accuracy. Detecting and monitoring drift in real-time is crucial for maintaining the stability of models, particularly in fields such as finance, healthcare, and cybersecurity, where changes in input data or conditions can significantly affect model performance. The aim of the paper is to investigate methods for monitoring model drift, particularly within the integration of CI/CD pipelines, to ensure their stability in real-world conditions. Special attention is paid to types of drift (data drift, concept drift, label drift) and the metrics used for their detection. The research methods include analyzing existing tools for monitoring and detecting changes in model behavior through the example of financial risk forecasting, as well as evaluating the effectiveness of integrating monitoring into CI/CD. The scientific novelty lies in the proposed comprehensive approach to detecting drift and integrating monitoring into production pipelines using advanced tools such as Google Vertex AI, AWS SageMaker, and TensorFlow Extended, which allow automatic response to changes in data. The use of such technologies improves prediction accuracy and reduces errors in realworld environments. The study confirms the importance of integrating drift monitoring into the continuous process of updating and adapting models to maintain their effectiveness in the context of constantly changing data. The conclusions show that integrating drift monitoring systems into CI/CD pipelines significantly improves the stability and effectiveness of models. Timely detection of drift allows for prompt model adjustments, reducing the likelihood of model degradation. It has been found that for achieving model stability, the automation of monitoring is crucial, as it allows for a prompt response to changes without manual intervention. This enhances the system’s efficiency and reduces risks related to the deterioration of prediction quality.

SUPER-RESOLUTION METHODS FOR IMPROVING THE DETAILS OF ULTRASOUND IMAGES

Olga BOYKO, Mykhailo TATARENKO — Wed, 28 May 2025 00:00:00 +0300

The paper investigates the possibilities of using super-resolution methods to improve the spatial detail of ultrasound images in medical diagnostics. The relevance of the topic is due to the limitations of the hardware resolution of traditional ultrasound systems, which limits the accuracy of visualization of small anatomical structures and complicates the clinical diagnosis. The purpose of the article is to provide a comprehensive justification for the effectiveness and feasibility of introducing super-resolution methods into ultrasound diagnostics by systematically analyzing their technical characteristics, mechanisms for improving the quality of image reconstruction, and factors affecting the possibilities of clinical use. The research methodology is based on the generalization of modern approaches to the optimization of model architectures, preparation of input data, pre-training and clinical validation of results. Particular attention is paid to a comparative analysis of the advantages of transformable models, such as SwinIR, in the context of medical visualization. The paper characterizes the main factors of loss of spatial detail in ultrasound images, systematizes the classification of superresolution methods by technical parameters, analyzes organizational barriers to the implementation of algorithms in the clinical environment, and develops reasonable recommendations for the integration of such solutions into medical infrastructure. The study found that neural models of superresolution, in particular, architectures based on convolutional networks and transformers, have a high potential for improving the quality of ultrasound images without the need to upgrade hardware. The key problems associated with the lack of labeled ultrasound datasets, the difficulty of complying with DICOM and PACS standards, and the ethical aspects of medical liability are identified. The paper proves that successful implementation of superresolution methods requires adaptation of algorithms to the specifics of medical systems, gradual integration into existing software solutions, maintaining physician control, and ensuring compliance with regulatory requirements. The scientific novelty is a comprehensive analysis of ways to improve the efficiency of superresolution algorithms in ultrasound imaging and to determine the directions of their optimal use in medical practice.

ANALYSIS OF SENSOR INDICATORS FOR PREDICTIVE MAINTENANCE OF HEAVY INDUSTRY PROCESS LINES

Nadiia BOLIUBASH, Oleksandra FINCOVA — Wed, 28 May 2025 00:00:00 +0300

The article considers the directions of digital transformation in the maintenance of process lines of heavy industry, aimed at predicting potential equipment malfunctions before their complete functional failure. The choice of methods for detecting problematic data coming from sensors of the process line is justified, aimed at reducing the time interval between the occurrence of a problem and its detection. Approaches to ensuring optimal updates of Predictive Maintenance models are considered, which enhance the accuracy of failure prediction. The purpose of the article is to research and identify effective methods for analyzing the quality of data from sensors of process lines in heavy industry to increase the accuracy of failure prediction in predictive maintenance. Methodology. Software and hardware solutions aimed at increasing the accuracy of forecasting faults in process lines based on the RCM methodology have been analyzed. Methods of correlation analysis, descriptive and contextual statistics were used to monitor sensor indicator data. A sequence of stages for detecting problematic data in real-world production conditions in heavy industry is proposed, which provide automated generation of potential sources of faults. The scientific novelty of the research lies in the identification of methods and approaches to analyzing the performance of sensors of process lines in the conditions of Industry 4.0, aimed at increasing the accuracy of predictive maintenance models. Conclusions. Maintaining Predictive Maintenance models that serve heavy industry process lines in real production conditions requires adjusting and updating their parameters based on constant monitoring of the equipment status in real time. To ensure the detection of problematic data, the feasibility of developing an information system, implementing the analysis of sensor indicators data using descriptive statistics, contextual statistics, and correlation analysis methods. The results of implementing the developed information and analytical system, which includes a recommendation system at the stage of forming a data set for analysis, were accompanied by an increase in failure prediction accuracy and a decrease in the time between the occurrence and detection of problems leading to equipment failures. This indicates that operational tasks are being solved with higher accuracy and efficiency.

CURRENT THREATS AND PROTECTION STRATEGIES FOR VISUAL CONTENT IN INTERACTIVE IT PROJECT SYSTEMS

Denys BUKATOV, Yanina KOLODINSKA, Ihor FROLOV, Oleh ROMANENKO — Wed, 28 May 2025 00:00:00 +0300

In the context of rapid digitalization and the accelerated development of artificial intelligence technologies, protecting visual content has become increasingly important. The growing threats of forgery, theft, and unauthorized use of digital images necessitate the improvement of security methods and the development of innovative strategies to counter violations of intellectual property rights. Purpose of the study. The aim of this work is to analyze modern methods of protecting visual content in interactive IT projects, taking into account the rapid development of digital technologies, the increasing risks of digital content forgery, the imperfections of intellectual property regulations, and the influence of artificial intelligence. Methodology. The study employs a comparative analysis of existing digital content protection techniques, including encryption, watermarking, access restrictions, metadata application, and neural network–based approaches. It also examines adversarial attack methods, particularly the Nightshade technology. Special attention is paid to the integration of user interface–level protections and technical restrictions in web environments, ensuring a comprehensive approach to the security of multimedia data. Scientific novelty. The scientific novelty lies in the comprehensive consideration of advanced technologies for protecting visual content from automated collection and use by artificial intelligence systems. The study highlights the critical role of metadata in verifying the authenticity of digital images and suggests modern practices for applying neural networks to detect visual content manipulations, taking into account current cybersecurity challenges. Conclusions. The article emphasizes the importance of multi-level protection of digital content by combining various approaches – from encryption and watermarking to neural network technologies and specialized adversarial methods. It has been found that an effective strategy must encompass both technical and organizational security measures, ensuring flexibility of security systems amid the ever-growing digital threats. Future research directions are defined in the development of new protection algorithms, their integration into IT products, and the optimization of existing methods to enhance resilience to modern challenges.

POSSIBILITIES OF APPLYING ARTIFICIAL INTELLIGENCE TECHNOLOGIES FOR DEVELOPING A FINANCIAL ASSISTANT

Tetiana VAKALIUK, Dmytro ANTONIUK, Serhii DUNIEV, Oleh TALAVER, Illia DOVGALIUK — Wed, 28 May 2025 00:00:00 +0300

Purpose of the work. To explore the possibilities of applying artificial intelligence technologies (machine learning, natural language processing, deep learning, recommendation systems, and computer vision) to create an effective financial assistant capable of analysing user data, forecasting financial indicators, detecting anomalies in transactions, providing personalised recommendations for budget optimisation, and automating financial planning processes. Methodology. The article systematically analyses the current state of development of financial platforms and artificial intelligence technologies. A comparative study of the capabilities of various AI technologies for integration into financial systems was conducted. Financial platforms were classified according to their functional purpose. An analysis of practical examples of the application of artificial intelligence in the financial sector was performed based on scientific publications and industry research. Information synthesis methods identify key directions for developing intelligent financial assistants. Scientific novelty. For the first time, the possibilities of integrating five key artificial intelligence technologies (machine learning, natural language processing, deep learning, recommendation systems, and computer vision) for creating a multifunctional financial assistant have been comprehensively investigated. The specific functional tasks of a financial assistant have developed a classification of applications for each technology. The main problems and challenges of integrating AI into financial platforms have been identified, particularly the need to ensure high accuracy in processing large volumes of data, detecting new types of fraud, personalising services and complying with regulatory requirements. Five key areas for further research have been identified, determining the prospects for developing intelligent financial systems. Conclusions. Integrating artificial intelligence technologies into financial platforms provides process automation, improved data analysis accuracy, personalised recommendations and increased security of financial transactions. Machine learning enables the prediction of financial results, the analysis of payment behaviour, and the detection of anomalies in realtime. Natural language processing automates the creation of personalised reports, the development of intelligent chatbots and the analysis of market sentiment. Deep learning optimises the construction of accurate predictive models and investment portfolios. Recommendation systems provide personalised advice on financial management. Computer vision automates data entry from financial documents. Implementing these technologies contributes to effective financial management, risk reduction, and achieving users’ long-term financial goals.

MAIN METHODS OF MACHINE LEARNING IN DECISION SUPPORT SYSTEMS

Artem VATULA — Wed, 28 May 2025 00:00:00 +0300

The purpose of this paper is a comprehensive review and systematization of the main classes of machine learning (ML) methods utilized in decision support systems (DSS). The research includes a detailed analysis of distinctive features, practical examples of ML algorithms application in various fields (medicine, finance, logistics), and a critical comparison of the advantages and disadvantages of each method to identify optimal conditions for their integration into DSS and to enhance decision-making efficiency. The research methodology is based on a comprehensive analysis of scientific publications devoted to the use of ML in DSS, employing classification methods of algorithms according to the type of learning (supervised, unsupervised, semisupervised, reinforcement learning, deep learning). The study utilizes methods of analysis, synthesis, comparative evaluation of algorithm capabilities, and generalization of practical implementation results. Additionally, the research considers the quality of input data used for training models and addresses potential ethical and technological constraints related to the integration of these methods. The scientific novelty of this research lies in the first-time holistic and systematized assessment of various categories of ML methods within DSS, providing an in-depth analysis of their advantages, limitations, the most suitable areas of application, and the conditions determining their integration effectiveness. The paper generalizes and compares the outcomes of previous studies, forming a comprehensive vision of prospective directions for the development of intelligent DSS, taking into account contemporary technological and ethical challenges. The conclusions of the research confirm that the integration of ML methods significantly enhances the quality and validity of managerial decisions by automating the analysis of large data volumes and adapting recommendations based on accumulated experience. It is established that selecting a specific class of methods should depend on the nature of the task, the volume of available labeled data, the requirements for model interpretability, and allowable resource expenditures. Further research perspectives involve developing methodological guidelines for the optimal selection of methods for various types of DSS, improving algorithm transparency and interpretability, and creating effective mechanisms to ensure their ethical use, reliability, and safety in practical applications.

PLANT MONITORING SYSTEM: AI SYSTEM FOR INTELLIGENT PLANT MONITORING

Oleksiy HRACHOV — Wed, 28 May 2025 00:00:00 +0300

Modern agriculture faces numerous challenges, such as water scarcity, climate change, and the need for accurate crop monitoring. Traditional monitoring methods often do not provide sufficient accuracy, which can lead to reduced yields and inefficient use of resources. The aim of the work is to develop and evaluate the effectiveness of the SmartPlant AI intelligent system for automated plant monitoring in order to optimize resource use, increase yields, and reduce the impact of the human factor. Methodology. The SmartPlant AI system combines IoT devices, cloud technologies, and artificial intelligence algorithms to analyze critical environmental parameters in real time. Sensors are used to measure soil moisture, temperature, lighting, and nutrient composition. ESP8266, ESP32, Raspberry Pi, and Arduino controllers transmit this data to the cloud via Wi- Fi or LoRaWAN using MQTT or HTTP API. Machine learning algorithms predict threats and generate recommendations, in particular, neural networks analyze the relationships between environmental parameters, regression models assess their impact on plant growth, and decision trees optimize irrigation and fertilization. The scientific novelty. The proposed system provides integration of IoT, cloud computing, and artificial intelligence for automated monitoring of plant health, which allows to increase the accuracy of agricultural technologies. The use of machine learning algorithms in the data processing process makes it possible to predict threats and adapt plant care to current conditions. Conclusions. The results of testing on the example of growing tomatoes showed that the SmartPlant AI system reduces water consumption by 31%, stabilizes soil moisture levels, and promotes faster plant growth by 34% compared to traditional methods. Automation of the plant care process allows to reduce the impact of the human factor, increase the efficiency of agricultural production, and minimize the risks associated with climate change. Further research may include integrating additional sensors, using drones and satellite data for even more precise management of agricultural processes.

DEVELOPMENT OF A SOFTWARE PLATFORM FOR CONTROLLING A WIFI-ENABLED MANIPULATOR USING A CAMERA

Denys DUBOVYK, Tetiana DUBOVYK, Vadym MATUS — Wed, 28 May 2025 00:00:00 +0300

The purpose of this work is to create a convenient and effective program interface for controlling the manipulator, establishing communication with the camera to provide visual feedback, as well as developing a modular system that can accommodate various plugins. Methodology. The article discusses the key features of the software platform and its architecture, focusing on the interaction between the manipulator, camera, and various plugins. The manipulator, in its main function as an actuator, is equipped with clamps for work and a chamber above them (a kinematic scheme is used, a problem of reverse kinematics). The smartphone acts as a server (connected via Wi-Fi), sends commands to the manipulator. The program on the server has a modular structure: at any given time it uses only one plugin, which receives data from the manipulator through the interface, analyzes it and, depending on the functionality of the plugin, provides the appropriate commands to the manipulator. The program is developed in the Kotlin programming language using the Jetpack Compose graphics framework. It allows you to import the program to various platforms such as Linux, Mac OS, and Windows. Currently, the app only works on the Android platform. Scientific novelty. The importance of using input from a video camera in robot control, allowing tasks such as object recognition, tracking and improving decision-making, technical aspects of the application, including data exchange protocols for the manipulator and camera, plugin architecture, and user interface. Features of the modular software platform for controlling a manipulator with a video stream. Modularity, flexibility and real-time video input contribute to its practicality and relevance in the field of robotics. The practical significance of the Modular Manipulator Control Software Platform with Video Stream Implementation on Android is designed to simplify the control of robotic manipulators through a single software interface and real-time video integration to improve control and decision-making This work lies in its potential application in various industries such as manufacturing, healthcare, and automation. The modular nature of the platform simplifies integration of specific functions, making it suitable for a wide range of tasks in the field of robotics. The relevance of the platform lies in the growing demand for flexible robotic systems that can be adapted to different applications. By providing modularity and extensibility, the platform allows developers to create and integrate specific plugins for different tasks without having to reprogram the entire control system. Conclusions. The main results of the research indicate that the modular software platform successfully provides a single interface for manipulator control and real-time video stream integration. The platform’s flexibility allows users to adapt the system to different tasks, making it a valuable tool for researchers and developers in the field of robotics.

DEVELOPMENT OF A GROUND-BASED CONTROLLED OFF-ROAD DRONE WITH RADIO CONTROL

Tetiana DUBOVYK, Daniil Ivanitskiy, Ihor LEVCHUK, Hanna HUZ, Oleksandr Romanchuk — Wed, 28 May 2025 00:00:00 +0300

The purpose of this work is to develop a ground-based controlled off-road drone with radio control, which can work effectively on rough terrain. Methodology. For programming, a simplified version of C++ is used, development can be carried out both using the free Arduino IDE environment and using arbitrary C/C++ tools. For programming and communicating with a computer, you will need a USB cable, and for autonomous operation, you will need a 7.5-12V power supply 5.5*2.1mm connector. Practical importance is the rapid transmission of information: radio communication allows you to quickly transmit messages over long distances, in particular where other methods of communication may not be reliable enough or even impossible. Thanks to the use of radio waves, information is transmitted instantaneously, which is extremely important for the work of emergency services, coordination of operations or the provision of assistance in real time. Radio communication plays an important role in military communications, providing communication between units in the field, coordinating operations and transmitting commands in real time. The military uses special secure communication channels that can function in conditions of interference and attempts to jam the signal, which ensures the safety and effectiveness of military operations. Scientific novelty. The drone has its own encryption systems that are used for confidential communications. Regulators set obligations for telecom operators, which must ensure the protection of user data and ensure its inviolability in their networks. The relevance of the topic lies in the constant growth of demand for autonomous and semi-autonomous vehicles capable of performing tasks in difficult conditions. The proposed project is aimed at solving problems related to increasing the crosscountry ability and ease of control of ground drones Conclusions. The design of the structure, the selection of components, the creation of an electronic part, control software have been implemented. The results achieved demonstrate the possibility of using the developed drone to perform tasks in difficult weather conditions or in the field. Increased cross-country ability is provided by a well-thought-out chassis design and powerful drives, and ease of control is achieved thanks to an optimized radio communication system. The drone has its own encryption systems that are used for confidential communications.

DISCUSSING PERSPECTIVES OF DEVELOPMENT OF AN OFFLINE UKRAINIAN-SPEAKING LLM-BASED ASSISTANT INTEGRATED WITH SPEECH SYNTHESIS & RECOGNITION TECHNOLOGIES

Vladyslav IVANOV, Lidiia HOBYR, Tetiana VAVRYK — Wed, 28 May 2025 00:00:00 +0300

The purpose of the article is to analyze modern approaches and technical possibilities for implementing a fullfledged Ukrainian-speaking voice assistant to meet the need for autonomy, confidentiality, personalization and to ensure the flexibility of customization to the specific requirements of target consumers. The paper considers the problems of existing solutions on examples of well-known cloud platforms and emphasizes the need to develop independent and autonomous analogs. An analysis of known open-source projects for speech recognition, text-to-speech and human-like speech synthesis was conducted to identify those that provide high quality processing at relatively low resource costs, are capable of working with different languages, including Ukrainian, and can be used to implement a demo application. Methods and techniques for reducing the hardware requirements of the end system to ensure efficient operation of the system in resource-limited environments are analyzed. Particular attention was paid to optimizing the performance of mathematical operations of model inference by using the hardware acceleration capabilities of individual computing platforms. In addition, the peculiarities of integrating each component into a single system based on a microservice architecture with the ability to adapt these tools to the specific needs of users are considered. The scientific novelty lies in the systematization of information on technologies and tools suitable for creating such assistants without using cloud services and with a combination of various optimization techniques for deploying such systems on consumer-level devices. The article proves that if the components are combined with minimal interaction delays, performance and processing quality are improved without a significant increase in resource requirements, and the system is properly implemented to combine all components, a competitive autonomous voice assistant capable of being integrated into any system thanks to an open platform can be implemented.

RESEARCH ON THE RELEVANCE OF IMPLEMENTING THE INTERNET OF THINGS AS A COMPONENT OF AUTOMATION IN THE CHEMICAL INDUSTRY

Artem KAMENSKYI, Vira BABENKO — Wed, 28 May 2025 00:00:00 +0300

The paper considers the results of research into the relevance of introducing such innovative technology as the Internet of Things (IoT) into the chemical industry. Modern chemical enterprises pay special attention to automation and control tools. This is important not only from the point of view of obtaining high-quality final products, but also for maintaining specified technological modes, analyzing the condition of equipment and production safety in general. Therefore, most large enterprises that produce mineral fertilizers, chemical reagents, etc. use SCADA systems for automation and monitoring. The rapid development of technologies has made it possible to introduce an IoT monitoring system into various areas of business, which is investigated in the article. The paper provides a comparative analysis of the differences in controlling production processes using SCADA systems and the modern innovative concept of IoT. The purpose of the article is a comprehensive study of the possibilities and prospects for the implementation of Internet of Things (IoT) technologies in chemical production by applying methods of search, analysis, comparison and generalization of data from information sources. Methodology. To achieve the set goal, a systematic analysis of scientific publications, articles and technical reports was carried out, devoted to the features of the Internet of Things (IoT) architecture, the practice of its implementation in various industries, as well as the advantages and challenges associated with the use of this technology. This approach allowed us to form a holistic picture of the potential of IoT in the field of chemical production and to objectively assess the relevance of its implementation in the context of modern Ukrainian realities. The method of comparing control technologies used in the work makes it possible to assess the similarities and differences of SCADA and IoT. Scientific novelty. The article presents a comparative characteristic of control technologies according to the following parameters: purpose and functions, architecture, data types, management and control, scalability, application in the chemical industry. Special attention is paid to assessing the feasibility of integrating IoT into existing automated production process control systems. Conclusions. Each of the control and management, monitoring and data processing systems (SCADA and IoT) has its own advantages and disadvantages. It has been determined that for the high-quality operation of a modern chemical enterprise it is advisable to apply the advantages of both technologies, creating a high-quality environment optimized for both process automation and data analysis.

MODERN RISK MANAGEMENT MODELS WITH A SYSTEMIC APPROACH

Andrii KOLIADA — Wed, 28 May 2025 00:00:00 +0300

The article explores the essence of risk management based on system analysis. The main principles and advantages of system analysis in risk management are identified. It is established that the use of a system model as the only source of truth for risk and reliability analysis leads to a consistent and complete assessment. The main risk management models are investigated and the components of system analysis in the risk management model are identified. The areas of application of the considered risk management models are presented. The objective of the study is to analyze risk management models with a systemic approach. Methodology. The research methodology includes an analysis of scientific literature, its critical understanding, synthesis. The use of such a methodology will provide a comprehensive and in-depth analysis of the issue, will contribute to the development of an effective risk management model that takes into account the specifics and dynamics of the modern business environment. The scientific novelty lies in the selection of the most adapted and flexible risk management models that take into account the systemic interrelationships of risks. Conclusions. This paper concludes that risk management models based on systems analysis involve identifying key elements of risk management, establishing relationships between them, and developing mathematical and logical processes for forecasting and managing risks. It has been established that risk management models contribute to the organization's flexibility, allowing it to respond quickly to changes in the external environment and internal challenges, which is a key factor in survival and development in conditions of uncertainty. The use of a systems approach involves the use of advanced methods, such as modeling of loss distributions, regression analysis, Bayesian networks, and fuzzy logic, which allows for more accurate risk assessment and forecasting.

SPECIAL TECHNIQUE FOR MANAGING DYNAMIC PROCESSES OF DEVELOPMENT OF INTELLECTUAL PRODUCTS PRODUCTION ACCORDING TO THE NEEDS OF SOCIAL LIFE CYCLES

Olena KOMISARENKO, Georhii BARANOV, Daria METELSKA — Wed, 28 May 2025 00:00:00 +0300

This article presents the results of a modern analysis of the problems of development of organizational management engineering of special technical means in the conditions of special operations that critically affect the life cycles of the relevant technological and natural regions. The purpose of the work is to develop a special technique for managing transitional processes in emergency social conflicts and extreme natural phenomena in order to make a decision on the use of control processes after ergatic modeling of a complex dynamic system based on digitalized information and intelligent technologies. Methodology. The processes of special control technology are initiated by a computational base model. It has standard program modules. The algorithms holistically reflect the real parameters of external disturbances, threats, interference, and intra-network interdependent connections between the defining fragments of the control object, which are the goals of each problem-situational task. Results. The operational synthesis of the computational basic model, according to the requirements of an accurate description, combines the specific functions of the consequential response of elements of typical program modules of a flow complex dynamic system and the source knowledge of the factors of the distributed large-scale influence of the external environment of the problem-situational task. Scientific novelty. This new computational basic model enables various experiments on mathematical models. The innovative law of synergistic process control adequately represents real parameters: A – set by the researcher; B – emergency attacks; C – sensitive objects of a streaming complex dynamic system; D – consequences and optimal parameters of guaranteed adaptive control. Conclusions. The experience of ergatic modeling and spatio-temporal situational awareness (STSA) in thesauri defines the essence, feature, specificity of the functioning of product production in the needs of a variety of procedures to prevent destruction, accidents, and disasters (Vcas Vehicle Collision Avoidance System).

INTELLIGENT PLATFORM FOR OPTIMIZING MANAGEMENT OF AGRO-INDUSTRIAL ECONOMY BASED ON CALS TECHNOLOGIES

Wed, 28 May 2025 00:00:00 +0300

Research objective. Development of an intelligent platform for optimizing the management of the agroindustrial economy, which is based on CALS technologies and integrates modern approaches to the automation of data collection, processing and analysis. The research is aimed at creating a system that provides operational forecasting of production processes, optimization of management decisions and adaptation to rapidly changing market conditions, which is a relevant task in the context of the digital transformation of the agricultural sector. Methodology. The research used a comprehensive approach, including mathematical modeling of system dynamics using differential equations, development of an expert forecasting module based on artificial neural networks, and development of algorithms for optimizing management decisions. The platform architecture includes integrated data management, a configuration module for specific farm conditions, and a decision support unit that ensures cyclic adaptation of the system to external and internal influences. Scientific novelty. A new approach to managing the agro-industrial economy is proposed, based on the synergy of CALS technologies with modern information systems. The developed mathematical model allows not only to predict changes in key indicators (production, product quality, market demand), but also to adjust management decisions in real time. The innovation of the platform lies in the use of hybrid methods of data analysis and the integration of machine learning algorithms to support optimal decision-making. Conclusions. The results obtained demonstrate the effectiveness of the developed platform, which allows adaptively responding to changes in market conditions and increasing the competitiveness of agro-industrial enterprises. Numerical modeling confirmed the possibility of optimizing production processes through cyclical adjustment of investments, which positively affects the volume of production, product quality and the level of market demand. The proposed system has high potential for practical implementation in the agricultural sector, and also opens up prospects for further research in the field of digitalization of economic process management.

USING IOT DEVICES TO MONITOR PLANT HEALTH IN AGRICULTURE

Wed, 28 May 2025 00:00:00 +0300

The article considers the process of developing an intelligent IoT device for monitoring plant health in agriculture. The system architecture, neural network training methodology, and real-time image analysis mechanism are described. The purpose of this study is to develop and experimentally verify an intelligent IoT device for monitoring plant health, capable of detecting signs of diseases and the presence of pests using convolutional neural networks. Research methodology. The study uses a comprehensive approach that combines hardware and software into a single IoT plant monitoring system. The YOLOv9 convolutional neural network is used to analyze plant images. The scientific novelty lies in the creation and testing of a prototype of an autonomous intelligent device for monitoring plant health. A system architecture is proposed that allows local detection of signs of diseases and pests without connecting to external servers, which reduces delays and increases reliability in field conditions. Conclusions. As a result of the research, a prototype of an intelligent IoT device for monitoring the condition of plants in agriculture was successfully developed and tested. The experiments conducted proved the effectiveness of the device in a real environment.

CHALLENGES AND PROSPECTS FOR THE DEVELOPMENT OF ARTIFICIAL INTELLIGENCE IN THE REPUBLIC OF IRELAND

Marianna MARUSYNETS — Wed, 28 May 2025 00:00:00 +0300

The article presents an analysis of trends in the development of artificial intelligence (AI) and technology startups in the Republic of Ireland in recent years. Particular attention is paid to how the Irish state has become the new European headquarters of such technology giants as Microsoft, Google and Facebook. It is noted that after the UK's exit from the EU, international brands began to move their European headquarters from the UK to the Republic of Ireland. The article emphasises the importance of effective interaction between academia and industry, which allows Ireland to take a leading position in the field of artificial intelligence. The article emphasises the growing demand for AI technologies in various sectors of the economy and the prospects for their development in the country. It is argued that in the modern world, AI technologies are playing an increasingly important role in various spheres of life, from automating production processes to improving the quality of life. However, despite the widespread use of AI, there is still no single definition of this concept. In this context, an attempt is made to analyse different approaches to the definition of AI proposed by various experts and organisations, as well as to consider the main classifications and applications of AI. The purpose of the study is to provide a comprehensive and integrated analysis of trends, potential and prospects for the development of artificial intelligence in the Republic of Ireland. Scientific novelty. The scientific novelty of the study lies in the comprehensive analysis of AI development in the Republic of Ireland, as well as in the substantiation that the Irish state has a rich history of technological innovation, and the government is actively investing in the development of AI, which contributes to the development of the country's technology sector. Irish specialists have a high level of education and English language skills, which facilitates interaction with companies around the world. AI is developing rapidly and can significantly affect the life and work of the Irish, bringing both benefits and risks. The Republic of Ireland is actively working to create a fair and just AI system, successfully addressing complex ethical issues and establishing strict rules. The ethical principles of AI in the Republic of Ireland include transparency, fairness, non-discrimination, privacy, and security. The study and systematisation of official documents of the Republic of Ireland's state authorities, as well as the analysis of relevant factual data, identified key areas, potential and prospects for further AI development in the Republic of Ireland. Methodology. The theoretical and methodological basis of the study includes general scientific methods of analysis, synthesis, induction and deduction, as well as the systemic method, the method of structural and functional analysis and legal methods in cybersecurity. In order to understand the essence of AI development in the Republic of Ireland, as well as to identify the prospects and challenges on the way of its evolution, a systematic approach was applied. It allows us to consider AI as an integral complex system consisting of subsystems and elements that are in constant interaction. The study of the extensive legal framework within the framework of the regulation of the personal data and information protection system has led to the use of an interdisciplinary approach. As part of the interdisciplinary approach, legal methods in cybersecurity were applied, including the development of a legal framework governing relations related to the creation, storage, exchange and accessibility of information, as well as the creation of methodological recommendations, compliance with which will ensure the safety of personal data and confidential information. Conclusion. The study proves that the regulatory framework of the Republic of Ireland is aimed at balancing innovation and ethical standards, with the application of the Data Protection Act 2018 and the General Data Protection Regulation 2018 to AI programs that use personal data. In the EU, there has been a shift in AI regulation with the adoption of the European Union Artificial Intelligence Act 2024. Its impact on the Republic of Ireland remains to be seen, but it will have significant implications for the country's approach to AI in the future. The Irish state's AI compliance strategies are based on several principles: establishing a robust regulatory framework, fostering a culture of responsibility, proactive risk analysis, cooperation, and adaptation to technological advances.

AUTOMATED CONTROL SYSTEMS FOR UNMANNED PLATFORMS FOR AGRICULTURAL DATA COLLECTION: ARCHITECTURE, SOFTWARE, DEVELOPMENT PROSPECTS

Olena MARCHENKO — Wed, 28 May 2025 00:00:00 +0300

The article is devoted to the study of the architectural principles and software and hardware of automated control systems for unmanned platforms for collecting, processing and analyzing data in agriculture. Particular attention is paid to the use of UAVs for precision agriculture, which provides monitoring of agricultural landscapes and optimization of agricultural processes. Purpose of the study. The purpose of the work is to analyze and evaluate architectural solutions and software and hardware for the effective functioning of automated UAV control systems aimed at monitoring and analyzing agro-technological parameters in agriculture. Materials and methods. The study used UAVs equipped with RGB cameras, multispectral and infrared sensors, thermal imagers, laser scanners and radar systems. The methods of photogrammetry, analysis of vegetation indices (NDVI, GNDVI, ExG, ExGR), convolutional neural networks (CNN) and deep learning for image processing were applied. Training samples were formed using data generators, and processing was carried out using specialized software integrated with GIS and cloud computing. Scientific results obtained. A prototype of an automated system based on UAVs was developed, capable of detecting phytopathological changes, assessing the state of soils and vegetation with an accuracy of spatial extension of 1–10 cm. A neural network model (InceptionV2) was implemented, which provides high-precision image classification, in particular for diagnosing plant diseases. The use of vegetation indices and heat maps made it possible to create cartograms of crop conditions and predict yield. The system demonstrated effectiveness in monitoring soil moisture, plant stress, and soil texture, contributing to the adoption of informed agrotechnological decisions. Prospects for further research. Further research involves improving sensor systems, implementing autonomous UAV swarms, integrating with IoT, and developing artificial intelligence algorithms for predicting agricultural parameters. It is planned to scale solutions for large agricultural areas and increase economic efficiency by optimizing monitoring and management costs.

INTEGRATION OF ARTIFICIAL INTELLIGENCE INTO THE EDUCATIONAL PROCESS FOR TEACHING THE DISCIPLINE «FUNDAMENTALS OF PROGRAMMING IN C LANGUAGE»

Iryna MYKHAILIYK, Olha MURAVA — Wed, 28 May 2025 00:00:00 +0300

The article explores the possibilities of integrating artificial intelligence (AI) into the educational process for the course «Fundamentals of Programming in C». The aim of the study is to examine ways of integrating AI into education to improve the effectiveness of learning programming in C. Given the key role of the C language in IT education, mastering it is an essential step for students who aspire to develop programming skills and work with modern languages such as C++, Java, or Python. Additionally, knowledge of C is crucial for professional activities in fields related to operating system development, drivers, and embedded systems. The research methodology is based on the analysis of modern scientific sources that discuss the application of AI in education. The primary focus is on algorithms for automating code verification, methods for analyzing algorithm efficiency, and personalized approaches to student learning. The article examines AI integration mechanisms in education, including automatic error detection and analysis in code, algorithm optimization, test task generation, and code augmentation. The scientific novelty of the study lies in its practical approach to integrating AI into C language education. The use of intelligent systems enables the adaptation of the educational process to the students' level of preparation, fostering a deeper understanding of programming principles and enhancing their competence in working with algorithms and data structures. Special emphasis is placed on student interaction with AI, where algorithms analyze errors, suggest corrections, and provide instant feedback, strengthening the learning effect. Another important aspect of the study is the evaluation of the effectiveness of an interactive learning approach, implemented through adaptive algorithms and automated learning platforms. The article’s conclusions highlight the promising prospects of AI usage in education. The key benefits of implementing such technologies include adaptive learning, quick access to feedback, the ability to create variable tasks, and the development of students' practical skills. AI integration into C language education not only enhances material comprehension efficiency but also makes the learning process more interactive and engaging for students.

THE USE OF ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING IN SCIENTOMETRIC SYSTEMS AND DATABASES

Viktor OBODIAK, Mikhailo OTROSHKENKO — Wed, 28 May 2025 00:00:00 +0300

In today's scientific environment, the use of innovative technologies in scientific publication management systems, in particular artificial intelligence and machine learning, plays a key role in improving the efficiency and accuracy of processing and analysing large amounts of scientific information. This study aims to explore the potential and benefits of using artificial intelligence and machine learning in scientific publication management systems. The main purpose of this study is to unlock the potential of artificial intelligence and machine learning to improve the efficiency of scientific publication management. An analysis of recent research and publications has shown that the use of artificial intelligence allows automating the processes of classifying and analysing scientific publications, which helps to identify new scientific trends and increases the speed of decision-making. Machine learning, on the other hand, provides the ability to create predictive models that help forecast the development of scientific disciplines and determine their impact. The article analyses modern approaches to managing scientific publications and explores the possibilities of using artificial intelligence and machine learning to optimise them. The article also considers the problems of accuracy and objectivity of scientific activity assessment, which can be solved with the help of innovative technologies. The scientific novelty of this work is to study and analyse the use of artificial intelligence and machine learning in scientific publication management systems with a focus on scientometric systems and databases. The results of the analysis indicate that the integration of innovative technologies into scientific publication management systems will significantly improve the quality and speed of processing scientific information, as well as contribute to an objective assessment of scientific activity. It is suggested that these results should be taken into account when developing and improving scientific publication management systems to ensure a more efficient and innovative approach to scientific activity.

EXPLORING THE CAPABILITIES OF AUTOMATED TEXT SENTIMENT ANALYSIS USING MODERN LARGE LANGUAGE MODELS

Dmytro PAVLIUK, Oleh BAIBUZ — Wed, 28 May 2025 00:00:00 +0300

The objective of the study. This article explores the possibilities of automated analysis of political comments using modern large language models (LLM). The aim is to develop a software solution that classifies textual comments into two levels: by emotional tone (positive, negative, neutral) and by the target object of the reaction (event, author, publication style, community). The effectiveness of using LLM for sentiment analysis of political comments based on data from Telegram channels is evaluated. Methodology. To achieve the goal, a software prototype was developed that performs automatic text analysis. The prototype uses two classification dimensions: emotional tone and target object of reaction, taking into account the specifics of the political context. The input data consists of textual posts from Telegram channels and corresponding user comments, and the classification results are achieved using LLM with a few-shot learning approach. Scientific novelty. The developed prototype allows for multidimensional classification of texts, which is an uncommon approach in the study of political discourse, where it is important not only to determine the overall tone of the comment but also to identify who or what the reaction is directed towards. The research also offers strategies to improve classification results, including the integration of dynamic instructions and localization of training on Ukrainian-language data, which could be an important step in enhancing the effectiveness of using LLM for political content in Ukraine. Conclusions. The results of the research showed that LLMs have significant potential for performing multidimensional classification of political comments. However, limitations were identified, particularly in detecting sarcasm and irony, as well as in working with local specific contexts. Proposed improvement strategies, such as adapting the model to Ukrainian-language data and using dynamic prompts, allow for improved accuracy of results. The research highlights the need to adapt LLMs to the political context, especially for content moderation and sociological research. Future research should focus on collecting larger and more balanced datasets for more relevant and generalized results in the operation of the developed software.

ASPECTS OF USING GENERATIVE NEURAL NETWORKS FOR DESIGN PROJECTS

Dmytro PETRYNA, Olena KORNUTA, Sofiia MELNYK — Wed, 28 May 2025 00:00:00 +0300

This article explores the key aspects of using neural networks in architectural and interior design. The advantages of generative adversarial networks (GANs) in this field are discussed. Based on prompts, various images of a mountain chalet and its interior design were generated. Seamless textures were also created using the Midjourney neural network. Using the same neural network, furniture textures and posters were produced for 3D visualization of an apartment in SketchUp. This work aims to provide a comprehensive study of neural networks' application in design projects. To achieve this, it is necessary to analyze the current state of research in this field, identify the main development directions, and identify existing issues. The study describes various ways of using neural networks in architecture and interior design, evaluates the potential benefits and risks of their application, and develops recommendations for the effective use of neural networks in architectural and interior design practice. Methodology. The article uses theoretical methods of literature review, testing, and comprehensive analysis of the capabilities of modern neural networks, as well as empirical methods for studying the use of neural networks in design projects. Scientific novelty lies in the adaptation of modern approaches to the use of neural networks for the generation of architectural projects and interior design. Conclusions. The use of neural networks in architectural design as well as in interior and exterior design is becoming increasingly widespread. The development of generative image creation is progressing rapidly and is being actively applied in these fields. AI-based editing tools are being integrated into more and more well-known design programs and require preliminary analysis and study to develop specific recommendations for their use. This work provides a detailed analysis of future development directions of AI-based networks and proposes practical tools for their implementation.

FORMAL TESTING OF REALVALUED FUNCTIONS

Oleksii PISKUNOV, Natalia TUPKO, Alexander VASILIEV, Nadiia TOPIKHA — Wed, 28 May 2025 00:00:00 +0300

The article discusses the algebraic approach to software design and testing. The aim of the study is to apply methods of formal software design to the development of functions for calculating the length of a meridian arc. Although formal software development methods have not yet been widely adopted in practical activities, their application in mathematical research seems reasonable due to the formal nature of programming languages. Research methods: the research utilizes the basic principles of the RAISE formal development method, which allows the use of formal logic. Specifically, the study explores the use of software development tools provided by the RAISE Method Group to specify functions for calculating the length of a meridian arc and to define conditions for creating a test driver for these functions at an early programming stage. The testing conditions are specified in the form of axioms using the abstract applicative approach. Scientific novelty the novelty of the study lies in demonstrating that the RAISE Specification Language (RSL) and specialized utilities are sufficiently convenient tools for small research teams that lack specialized skills in formal software development. The focus is on mathematical research, particularly on developing complex real-valued functions. In addition to the primary goal, seven coefficients (C0, C2, C4, C6, D2, D4, D6) of the Maclaurin series expansion for the considered functions were clarified. Formulas and calculations for the next two coefficients (C8, D8) were also performed. A practical method for developing generalized test drivers based on inheriting a class from a mock class was proposed, utilizing. NET features such as partial classes. Overall, the article aims to reduce the gap between the theoretical advantages of formal development methods and their insufficient practical application. Conclusions: algebraic design and testing are based on mathematical principles, enabling the following: avoiding ambiguity in functionality descriptions; ensuring precision and clarity in software requirements formulation; automating the generation of test cases and verification processes, thereby enhancing reliability; identifying and fixing errors at the development stage; and accelerating software development.

THE ROLE OF CONTAINERIZATION AND OS-LEVEL VIRTUALIZATION IN THE DEVELOPMENT OF CLOUD-NATIVE APPLICATIONS

Artem PODVIZHENKO, Anton PEREVERZIEV, Leonid LYTVYNENKO, Oleh SKLIARENKO — Wed, 28 May 2025 00:00:00 +0300

The aim of this study is to examine the impact of modern virtualization and containerization technologies on the transformation of approaches to the design, development, and operation of information systems in cloud environments, particularly in the context of the shift from monolithic to microservices architecture. Methodology. The paper presents an in-depth analysis of the fundamental role of virtualization and containerization technologies in the transformation of cloud computing. It explores the historical development of virtualization, the influence of hypervisors, requirements for guest operating systems, and the benefits of containerization with a focus on tools such as Docker and Kubernetes. The architecture of containerized environments is analyzed, along with their integration with cloud infrastructure and the emerging requirements for base operating systems, including the use of immutable systems. Scientific Novelty. The study summarizes the modern approach to designing operating systems for containerized cloud environments. It particularly emphasizes the evolution of OS requirements, the development of specialized distributions, changes in security models, and the role of Kubernetes as a distributed cloud operating system. Future trends in OS development are outlined, including the growing role of intelligent management in cloud environments. Conclusions. Understanding the role and impact of virtualization and containerization technologies is critically important for the effective design, development, and maintenance of cloud-oriented applications. The future evolution of operating systems will move toward enhanced security, optimized hardware interaction, and the implementation of intelligent management mechanisms to ensure stability and scalability in highly dynamic cloud environments.

ANALYSIS OF THE IMPACT OF UNCERTAINTY ON SOLVING THE INVESTMENT ALLOCATION PROBLEM IN THE FIELD OF ECONOMIC ACTIVITY

Olena Podkovalihina — Wed, 28 May 2025 00:00:00 +0300

The distribution of investments in different sectors of the economy determines their future development. In conditions of economic instability, the profitability of investments becomes unpredictable, which complicates decision-making. Therefore, it is important to assess the consequences of different investment allocation strategies, exclude irrational options and determine the optimal ones. Research objective. To investigate the impact of statistical uncertainty on the optimal allocation of investments in the field of economic activity and develop tools for making effective investment decisions in unstable conditions. Methodology. Methods of exhaustive search methods, statistical analysis and multi-criteria optimization were used. The mathematical model of the investment allocation problem was implemented in the Python environment taking into account the uniform distribution of profits. Scientific novelty. A generalized statement of the investment allocation problem under conditions of uncertainty was formulated. An algorithm and software implementation were developed for calculating and analyzing optimal investment strategies taking into account variable profitability parameters. The proposed algorithm allows: to estimate future profits taking into account economic instability; determine a rational distribution of investments for different scenarios of profit dynamics; reduce the risks of losing investors or attracting new ones due to substantiated forecasts. The distribution of investments in agriculture with variable profits (10% and 40%) was studied, optimal solutions were analyzed according to various criteria. The feasibility of investing in the selected sector was determined depending on the level of uncertainty and external factors. Conclusions. The proposed approach allows predicting the profitability of investments taking into account economic instability. Criteria for optimal resource allocation that reduce investor risks were determined. The results can be used: by investors for decision-making under uncertainty; by government agencies to support strategically important industries; by analysts to model economic scenarios. Prospects for further research: expanding the model to other types of profit distributions (normal, Poisson); taking into account additional risk factors (for example, currency fluctuations); integrating machine learning to predict profit dynamics. This study highlights the importance of uncertainty analysis in investment decisions and offers a practical tool for their optimization.

MODELS AND METHODS FOR DESIGNING INTELLIGENT HIERARCHICAL CONTROL SYSTEMS IN CHEMICAL ENGINEERING PROCESSES

Ihor SHELEHOV, Dmytro PRYLEPA, Vladyslav AVLASOVYCH, Oleksii KOLISNYK — Wed, 28 May 2025 00:00:00 +0300

The article examines the development of an intelligent hierarchical control system for chemical-technological processes based on the information-extreme intelligent technology (IEIT). It describes the process of creating a mathematical model and algorithms for constructing hierarchical classifiers, which enhance the efficiency of managing complex chemicaltechnological processes at industrial enterprises. The approaches to system optimization, grounded in functional-statistical testing, along with the methods for training and adapting classifiers during operation, enable the creation of a flexible and adaptive system capable of effectively responding to changing conditions within the technological process. The aim of the work is to improve technological control and process automation at chemical enterprises through the implementation of intelligent systems incorporating elements of machine learning. The research methodology involves the use of mathematical models and algorithms that ensure precise tuning and control of technological process parameters, as well as the evaluation of the proposed system’s effectiveness using real-world data. A promising direction for further development lies in the implementation of hierarchical classifiers, which have proven to be an effective means of enhancing the reliability of decisions related to the assessment of the functional state of processes of various natures. The scientific novelty of this work lies in the application of hierarchical models and methods of information-extreme intelligent technology for the design of automated, machine learning-capable control systems. According to the calculations, the use of this approach enables functional efficiency in managing production processes in real-time.

INTEGRATION OF GENERAL ARTIFICIAL INTELLIGENCE SAFETY ASSESSMENT METHODS INTO A TARGETED SOFTWARE DEVELOPMENT RISK MANAGEMENT MODEL

Oleksandr PSAROV — Wed, 28 May 2025 00:00:00 +0300

The subject of this article is the comprehensive study and development of approaches to integrating General Artificial Intelligence safety assessment methods into the risk management process in software development. Particular attention is paid to analyzing potential threats that arise when using Artificial General Intelligence in modern software products and examining methods for minimizing these risks. The article presents practical solutions, including the use of the Analytical Hierarchy Process to prioritize risks and genetic algorithms to find optimal risk mitigation strategies. The main principles of Analytical Hierarchy Process application in the context of software risk management are described in detail, including the construction of hierarchies, risk assessment, and priority matrix formation. Genetic algorithms are used to identify optimal solutions based on Analytical Hierarchy Process data, providing adaptability and precision in creating secure software. Calculation examples demonstrate how combining these methods contributes to more effective risk management. The aim of the article is to develop tools for assessing and managing risks associated with the use of Artificial General Intelligence in software development processes and to ensure the safety of these processes by integrating advanced algorithmic solutions. Conclusions. The integration of the Analytical Hierarchy Process and genetic algorithms enables the creation of an effective risk management model in software development involving Artificial General Intelligence, minimizing threats and ensuring high system reliability.

MODELING OF AN INFORMATION AND ANALYTICAL SYSTEM FOR MONITORING GENDER EQUALITY IN HIGHER EDUCATION INSTITUTIONS

Svitlana RZAEVA, Daryna CHERNYSHOVA — Wed, 28 May 2025 00:00:00 +0300

In the modern world, the issue of gender equality is becoming increasingly important. In particular, in higher education institutions, gender equality is a fundamental aspect that affects the quality of education, the motivation of scientific and pedagogical staff, as well as the general moral and psychological climate in institutions. One of the main challenges facing modern universities is to ensure equal opportunities for all employees regardless of their gender. This includes equal access to academic and professional development, equal working conditions, fair distribution of resources and opportunities for career growth. Objective development of an information and analytical system for monitoring gender equality in higher education institutions using UML diagrams, which allows to increase the efficiency of gender equality assessment among the teaching and research staff. Methodology use of uml (unified modeling language) for designing the structure and functionality of the iamgm. Application of formalization and algorithmization methods to analyze, refactor and transform uml diagrams and determine the structural complexity of the system through the number of classes, interfaces and connections. Visualization of system components and identification of key classes (candidate, vacancy, selection criteria, etc.). Scientific novelty a formal approach to the design of iasmg with regard to gender neutrality is proposed. An objective function for optimizing the structure of uml-class diagrams is introduced. A set of semantically equivalent transformations of uml-diagrams is defined to reduce gender bias and the need to use software for factoring uml-diagrams in multi-institutional systems is substantiated. The implementation of such systems is important not only for ensuring fairness and equal opportunities, but also for increasing the overall efficiency of higher education institutions. Studies show that gender inequality can lead to lower labor productivity, reduced employee engagement, increased stress and conflicts in the work environment. Universities that neglect gender equality issues risk losing competitiveness and high-quality personnel. Another important aspect is the impact of information technology on the process of managing gender equality. In particular, an important tool in the process of developing and implementing such systems is UML (Unified Modeling Language). The use of UML diagrams allows you to create standardized models that are easily understood by all participants in the development process, which increases the quality of the system and reduces the risk of errors. The implementation of such models is important in conditions where the gender equality monitoring system must be integrated into the overall university management structure and must meet the requirements of various stakeholders. Conclusions uml diagrams are an effective tool in the development of the iamgm architecture. It is necessary to focus on visualization of analytical information on gender equality. The research prospects are to scale the system to process large amounts of data, as well as to adapt the iamgm to changes in the requirements and needs of users.

GENERATING A SEQUENCE OF ASYMMETRIC CET OPERATIONS WITH AN ACCURACY OF UP TO THE PERMUTATION OF THE SECOND OPERAND

Volodymyr RUDNYTSKYI, Vira BABENKO, Serhii RUDNYTSKYI, Tymofii KOROTKYI — Wed, 28 May 2025 00:00:00 +0300

The article is devoted to the construction of asymmetric CET operations suitable for cryptographic information transformation, namely CET encryption. The use of CET operations in the construction of streaming ciphers guarantees bidirectional transmission of information between subscribers. The specific feature of asymmetric CET operations is that they provide different ciphers in the process of encrypting the same information using the same key sequences. To improve stream ciphers, it is of great importance to be able to use several asymmetric two-operand CET operations with the same properties instead of one. That is why the problems of synthesizing groups of CET operations with permutation accuracy are extremely relevant. After all, their solution will allow us to build generators of pseudo-random sequences of CET operations with the same properties to ensure the construction of new generation of cryptographic algorithms of stream encryption. The purpose of the article is to study the possibility of using the method of generating groups of asymmetric CET operations with an accuracy of up to the permutation of the second operand to build low-resource asymmetric stream encryption systems. Methodology. The paper shows that it is expedient to generate groups of operations of strictly stable coding on the basis of modifying the CET operation of strictly stable coding with an accuracy of permutation of the second operand. To synthesize models of CET operations, both direct and reverse, one of the ways to implement their construction method is presented, which consists in generating a sequence of asymmetric two-operand CET operations with an accuracy of permutation of the second operand. The correctness of the obtained models for the synthesis of groups of direct and reverse asymmetric two-operand CET operations is verified on specific examples, which is confirmed by the results of a computational experiment for 2Ci-quantum asymmetric two-operand CET operations. In the course of the study, the methods and principles of information and coding theory, Boolean algebra, discrete mathematics, the theory of synthesis and modeling of operations, and cryptography were used. Scientific novelty. The scientific novelty of the work is to analyze the practical aspects of using groups of asymmetric CET operations with an accuracy of up to the permutation of the second operand to build low-resource cryptography algorithms by ensuring the generation of direct and reverse asymmetric two-operand CET operations by using only direct one-operand CET operations, which significantly reduces the complexity of the cryptographic algorithm and the complexity of the cryptographic system implementation. Conclusion. As a result of the study, the following was found: all asymmetric CET operations belonging to the synthesized group with an accuracy of up to the permutation of the second operand have the same cryptographic properties, since they implement the same sets of substitution tables in the same way by chance; the ability to generate groups of forward and reverse asymmetric CET operations with an accuracy of up to the permutation of the second operand on the basis of any two-operand CET operation will provide both encryption and decryption of information for building streaming encryption systems.

DEVELOPMENT OF A SPECIALIZED COMPUTER SYSTEM BASED ON ARDUINO UNO «SMART» CAR PARKING

Andrii SERHACHOV, Denys DUBOVYK, Tetiana DUBOVYK — Wed, 28 May 2025 00:00:00 +0300

This article presents the development of a specialized computer system based on Arduino Uno, which will include several important components to create an efficient and convenient parking tool. The system will include a web page where users can book a free parking space. This will reduce the time spent searching for a free space, as well as reduce the likelihood of improper parking. The goal of this project is to develop and implement a smart parking meter that will be able to automate the processes of booking, payment and control of parking spaces, as well as improve the efficiency of the use of available parking resources. The practical significance of this development lies in the automatic control of parking spaces. To do this, powerful machine learning algorithms with automatic text recognition on license plates are used, allowing for quick and accurate identification of vehicles in real time. Methodology. Based on TensorFlow technology, a camera is used that will read car license plates. Thanks to this, the parking meter will be able to automatically record the car that has taken a seat and update the status of space availability in the system in real time. A database will be created to store and process information about reservations, seat availability and user transactions, which will ensure reliable data storage and provide convenient access to information. Arduino Uno, due to its affordability, simplicity and wide possibilities for integration with various sensors and modules, is an ideal choice for the implementation of such a project. The integration of a web page for booking, license plate recognition with TensorFlow and a database for processing information will allow you to create an effective system that will significantly improve the parking process. The relevance of such a development as "smart" car parking lies in the growing demand for flexible robotic systems with artificial intelligence that can be adapted to different applications.

ANALYSIS OF THE CURRENT STATE AND APPLICATIONS OF SPATIAL COMPUTING ACROSS VARIOUS FIELDS

Taras SKYTSKYI — Wed, 28 May 2025 00:00:00 +0300

The article explores spatial computing, an advanced technology that enables the integration of digital information and virtual content into the physical world, opening new possibilities for interaction and data analysis. The article explores the history of spatial computing development, starting from the 1960s when Ivan Sutherland introduced the concept of Sketchpad, one of the first graphical interfaces. A significant technological breakthrough occurred in the 1980s with the development of virtual reality (VR) and immersive devices, and in the 2010s, with the widespread adoption of augmented reality (AR) in mobile devices. This article defines the key concepts and terminology of spatial computing, including virtual reality, augmented reality, mixed reality, and extended reality (VR, AR, MR, XR). The study examines core technologies that enable the integration of digital content into the physical environment, such as computer vision, motion tracking systems, and spatial data visualization. Key devices that contributed to the industry’s advancement, including Microsoft HoloLens and Apple Vision Pro, are also highlighted. The application of spatial computing is analyzed separately across various fields, including education, medicine, industry, the military sector, and tourism. In education, spatial computing facilitates the creation of interactive learning environments, particularly for exploring natural sciences and technical disciplines. In medicine, this technology is used for simulating surgical procedures, treating phantom limb pain, and rehabilitating patients. In manufacturing and design, spatial computing aids in creating 3D prototypes, while in the military sector, it is utilized for training soldiers and strategic mission planning. The article also addresses the challenges associated with spatial computing, such as high technical requirements, cybersickness, limited field of view, significant cognitive load on users, and the high cost of equipment. The conclusions emphasize the necessity of further research to reduce technology costs, develop effective interaction algorithms for spatial environments, and integrate machine learning for processing large volumes of data. Spatial computing has the potential to fundamentally transform how people interact with the digital world, but overcoming technological and social barriers is essential for its full implementation.

ARTIFICIAL INTELLIGENCE: HISTORY AND FORMATION

Serhiy STASEVYCH, Olena HOLODOVSKA — Wed, 28 May 2025 00:00:00 +0300

The article is devoted to the use of artificial intelligence in various sectors of the country's economy. The history of the emergence of the term «artificial intelligence» and its introduction into scientific circulation in research aimed at creating intelligent machines is considered. The initiator of the scientific event dedicated to the problems of creating intelligent machines was John McCarthy (associate professor of mathematics at Dartmouth College). In 1956, a landmark seminar was held in Dartmouth, where ten scientists in the field of mathematics and computer science gathered to discuss the possibility of creating electronic computing machines capable of learning, solving problems, and demonstrating other forms of intelligence. The purpose of the work. An analysis of various types of artificial intelligence used in modern sectors of the economy is carried out: analytical, functional, interactive, textual and visual. To solve real problems in various sectors of the economy and social life and build a model based on artificial intelligence, various technologies and algorithms are used. The advantages and disadvantages of using artificial intelligence in industry are presented. Methodology. The main goal of using artificial intelligence is to try to teach computers and machines to perform the cognitive functions of the human brain – the ability to understand, learn, study, realize, perceive and process external information, and as a result, to provide them with such human functions as problem solving, decision-making, perception and understanding of human communication. Scientific novelty. An analysis of the history of the emergence, development and application of artificial intelligence at the current stage of development of the world economy is carried out. Conclusions. Artificial intelligence is already used in various sectors of the economic and social life of the state and continues to actively develop further, conquering more and more processes in the field of application: production processes, agriculture, environmental protection, medicine, social relations, safety and security.

THE USE OF A COLLECTIVE OF MOBILE AGENTS FOR THE EXPLORATION OF UNDIRECTED GRAPHS

Andrii STOPKIN — Wed, 28 May 2025 00:00:00 +0300

The study proposed in this work is devoted to the problem of graph exploration by a collective of agents consisting of one stationary agent, which performs the necessary computations, and two mobile agents, which move through the graph and collect information about its structure. The purpose of this work is to develop a new efficient algorithm for the exploration of finite undirected graphs without loops and multiple edges by a collective of agents. The paper proposes the following methodology to achieve this goal: utilizing a collective of three agents. Two of them are agent-researchers, which can move through the graph, read labels on graph elements, and place or remove these labels. The agent-researchers have finite memory that does not depend on the number of nodes in the explored graph and use two colors each (three distinct colors in total) for graph exploration. During the process, the agent-researchers send messages to the third agent, the agent-experimenter, which is a stationary agent that stores the results of the agent-researchers' operations. Based on this information, the agent-experimenter gradually constructs a representation of the explored graph in its memory in the form of edge and node lists. The paper describes the working modes of the agent-researchers in detail. It also thoroughly examines the algorithm for processing the messages received by the agent-experimenter from the agent-researchers, which serves as the basis for constructing a map of the explored graph. The study analyzes the time, space, and communication complexities of the developed algorithm and evaluates the number of edge transitions required for agents to explore the graph. The scientific novelty lies in the development of an efficient algorithm for graph exploration by a collective of agents. Conclusions. The paper proposes a new graph exploration algorithm that has quadratic time, space, and communication complexities, with the number of edge transitions performed by the agent-researchers being estimated as On2 , where n is the number of nodes in the explored graph. The functioning of the proposed exploration algorithm is based on the depth-first traversal method.

ANALYSIS OF EXISTING METHODS FOR DETERMINING THE TRAJECTORIES OF ATMOSPHERIC AIRCRAFT FOR THE POSSIBILITY OF THEIR IMPROVEMENT

Oksana TYMOSHCHUK, Emil LUTFALIEV — Wed, 28 May 2025 00:00:00 +0300

The article provides a structured analysis of current approaches to flight trajectory planning for atmospheric unmanned aerial vehicles (UAVs), emphasizing the conceptual interplay between «trajectory», «route», and «navigation» within autonomous control systems. The research generalizes interdisciplinary interpretations of these terms, drawing from aviation, control theory, cybernetics, and artificial intelligence, to clarify their impact on navigation logic under uncertainty. A comparative review of planning methods is presented, including classical kinematic models, Kalman filtering, heuristic algorithms (e.g., potential fields, swarm optimization), and AI-driven techniques such as neural networks and reinforcement learning. These approaches are evaluated based on accuracy, adaptability, computational cost, and energy efficiency. Findings show that classical methods offer precision and predictability in structured settings, but struggle to adapt to dynamic environments. AI methods demonstrate higher adaptability and learning potential but require significant computational resources and training datasets. The article outlines key limitations typical for complex missions – such as unpredictable weather, urban obstacles, and real-time decision-making across multiple agents – and proposes directions for improvement. Suggested solutions include the integration of online learning, the use of generative AI models (e.g., GPT, diffusion models), and the implementation of multi-objective optimization techniques tailored to operational constraints. The study introduces a hybrid trajectory planning architecture combining deterministic models with intelligent adaptive layers that support autonomous control in changing environments. The proposed framework enhances UAV mission flexibility, reliability, and autonomy, contributing to the development of intelligent decision-support systems for next-generation aerial robotics.

PEDESTRIAN SPEED ESTIMATION METHOD APPLICATION FOR STATISTICAL ANALYSIS OF CROWD DYNAMICS

Maksym TISHKOV, Oksana TYMOSHCHUK — Wed, 28 May 2025 00:00:00 +0300

This paper addresses the problem of pedestrian detection, tracking, and gait speed estimation based on video footage from a surveillance camera positioned above the walking area. Such a configuration is typical for surveillance systems in public spaces. The method proposed by the authors does not require any specialized equipment, making it suitable for a wide range of real-world scenarios. The aim of this study is to evaluate the previously developed method for pedestrian speed estimation in a real-life case – video surveillance in a school corridor. The research seeks to assess the effectiveness and reliability of the algorithm under conditions that differ from a controlled laboratory environment. Research methodology. To test the method, a video recording was conducted in a school corridor, where a surveillance camera captured pedestrian flows over a period of 6.5 hours. Using computer vision algorithms, pedestrians were detected, their trajectories tracked, and walking speeds estimated. In total, 1,841 trajectories were extracted, with an average walking speed of 1.15 m/s. The collected data enabled a statistical analysis that revealed patterns of pedestrian traffic under both normal and emergency conditions. Scientific novelty. The study presents the first real-world validation of the previously proposed method, implemented without any prior environmental preparation or the use of additional sensors. A distinguishing feature of the experiment is the consideration of external factors, particularly regular air raid alerts occurring during wartime in Ukraine. Their impact on pedestrian behavior was identified and reflected in the analytical results. This approach contributes to a deeper understanding of human adaptive behavior in stressful situations. Conclusions. The conducted study confirmed the effectiveness of the previously developed method for pedestrian detection, tracking, and speed estimation in real-world conditions. The obtained results can be used to further improve decision support systems in the fields of safety, evacuation planning, and crowd behavior analysis. The use of real data and the inclusion of a crisis context significantly enhance the practical value of the proposed methodology.

WEBSITE DEVELOPMENT FOR MANAGING YOUR SCHEDULE AND TASKS

Tetyana CHILIKINA, Hanna KHARNAUKHOVA, Oksana ORIKHIVSKA, Yana ZAHORUIKO — Wed, 28 May 2025 00:00:00 +0300

In today's dynamic world, effective time management is critically important, but existing digital planners often do not meet user needs due to complexity or limited functionality. The purpose of the work is to create a web application that combines an intuitive interface, effective management of goals and tasks, and adaptation to the individual needs of the user. Methodology. Analytical approach to studying user needs and behavior patterns, UX/UI design methods, application of modern frameworks and programming languages (TypeScript, React, Next.js, NestJS), as well as use of principles of adaptive layout, client-server architecture and MongoDB databases. The scientific novelty lies in the structured integration of the aforementioned approaches into the functionality of one web application: limiting the user to 5 active goals, which ensures focus; the ability to analyze achievements through the statistics section; adaptability of the interface with support for Ukrainian and English languages. The web application is implemented based on a modular architecture with a clear division of responsibilities between modules (auth, user, tasks, goals, statistics). The client part is built according to the principles of features based architecture, which increases the scalability of the project. A tool has been developed that can be used by a wide range of users – students, professionals, freelancers, creative people – to effectively manage personal time and achieve strategic life or professional goals. Conclusions. The digital planner proposed in this work has a number of key advantages, namely: goal-oriented, flexible planning, minimalist interface, progress analytics, and adaptability.

ENHANCING THE SECURITY OF INTERBANK PAYMENTS WITH, A COMPREHENSIVE CRYPTOGRAPHIC ARCHITECTURE

Liudmyla SHPORT — Wed, 28 May 2025 00:00:00 +0300

In this article, the author discusses the problem of ensuring the security of electronic interbank payments in the context of growing cyber threats. Existing security systems, in particular SWIFT and SEPA, are analyzed and their limitations are identified. The purpose of this article is to develop a comprehensive cryptographic architecture that can significantly improve the security of electronic interbank payments by eliminating identified vulnerabilities and taking into account modern cyber threats. Methodology. The potential of integrated, systematic scientific approaches was used. Their combination allows for a comprehensive analysis of technical and pedagogical literature, including the study of monographic scientific publications and dissertation research covering the chosen topic. The scientific novelty lies in the development of a comprehensive cryptographic architecture based on encryption, authentication, intrusion detection, and security audit modules. Conclusions. The study found that modern interbank payment systems, such as SWIFT and SEPA, use a comprehensive approach to security, including organizational and technical measures. SWIFT uses asymmetric cryptography, hardware security modules, and two-factor authentication. SEPA uses EMV standards for card transactions and TLS/SSL protocols for online payments. It has been determined that to effectively counter modern cyber threats, a comprehensive cryptographic architecture for interbank payments is needed that meets strict requirements for security, performance, scalability, and interoperability. The author's proposed comprehensive security architecture for interbank payments consists of interconnected modules that provide multi-level protection. The modeling and experimental evaluation of the effectiveness of the proposed architecture was carried out, which confirmed its ability to provide a high level of security, performance and resistance to attacks.

ASSESSMENT OF THE PROBABILITY OF CONNECTION BETWEEN STRUCTURES OF INFORMATION SYSTEMS ON DIFFERENT GRAPH MODELS

Oleg BARABASH, Andrii MUSIENKO, Olha SVYNCHUK, Oleksii DUDKIN — Wed, 28 May 2025 00:00:00 +0300

Functional stability of information systems is a key factor in their reliability, especially for critical infrastructure, cloud services, and distributed networks. The failure of individual components can lead to significant economic and technical losses. Assessing the probability of system connectivity based on graph models allows predicting its resilience to failures. The aim of this paper is to analyze and compare different methods for assessing the functional stability of information systems based on graph models, as well as to develop software for their modeling. Methodology. The paper examines the main methods for assessing the functional stability of information systems and conducts a comparative analysis of exact and approximate calculation methods. A software tool for visualizing and analyzing the functional stability of network structures has also been developed, determining the impact of various system parameters on overall stability. Scientific novelty. The scientific contribution of this study lies in the development of an approach to assessing the functional stability of information systems based on graph models, comparing exact and approximate methods, creating a software tool for modeling and analyzing structural connectivity, and identifying key factors affecting system stability. Conclusion. The paper analyzes methods for assessing the functional stability of information systems based on graph models. Exact and approximate approaches, including the full enumeration method, minimal path method, Ezary-Proshan method, and Litvak-Ushakov method, have been compared. It has been determined that the choice of method depends on the complexity of the system and the accuracy requirements. The developed software enables modeling and analyzing the functional stability of various graph structures. The obtained results can be used to enhance the reliability and continuous operation of information systems.