INNOVATIVE TECHNOLOGIES AND SCIENTIFIC SOLUTIONS FOR INDUSTRIES

Investigation of the damping properties of 3D-printed liners with controlled perforation

2026-04-01T00:19:35+03:00

This study focuses on thermoplastic polyurethane (TPU) liners with controlled internal perforation intended for use in the “residual limb–liner–socket” system of lower-limb prostheses. The influence of material stiffness, hole geometry, and degree of perforation on vibration damping efficiency under impact–dynamic loading is investigated. The aim of the work is the experimental determination and optimization of the damping characteristics of TPU liners by varying material stiffness, hole shape, and perforation percentage using the free-decay vibration method. Tasks: to analyze the functional role of the liner as a damping element in the prosthetic system; to fabricate a series of 3D-printed TPU specimens with different stiffness levels, hole geometries, and perforation degrees; to implement an impact-based method for measuring free damped vibrations for each specimen; to determine the damping ratio and vibration attenuation percentage; to process experimental data in order to identify optimal combinations of material and geometric parameters; and to establish the relationships between material stiffness, perforation level, hole pattern, and the damping properties of liners. Results: a method for evaluating the damping characteristics of 3D-printed TPU liners with controlled internal structures was implemented and experimentally validated. A nonlinear dependence of vibration damping efficiency on the degree of perforation was identified, along with a systematic decrease in damping as TPU stiffness increased. It was shown that hexagonal hole geometry provides a more uniform deformation distribution and slightly higher damping efficiency compared to rhombic and triangular patterns. The obtained relationships enable targeted design of damping liners with an optimal balance between stiffness and vibration attenuation capacity. These findings support the solution of the following practical challenges: reduction of impact loads by decreasing peak dynamic forces transmitted from the prosthetic socket to the soft tissues of the residual limb; pressure redistribution through the formation of more uniform contact stresses at the skin–liner interface; improved user comfort by reducing vibration, pain, and skin irritation during walking; individual optimization through personalized selection of liner geometry and material based on body parameters, activity level, and tissue condition; and engineering design of structurally optimized components for biomedical applications. Conclusions: the study experimentally confirms the feasibility of controlling the damping properties of TPU liners by adjusting material stiffness, degree of perforation, and hole geometry. An optimal perforation range was identified for each TPU stiffness level, as well as the advantages of hexagonal perforation in terms of deformation uniformity and vibration damping efficiency.

Model and method of automated high-precision measurement of three-dimensional objects based on computed tomography data

2026-03-31T22:24:08+03:00

The subject of study is the linear dimensions of three-dimensional objects based on computed tomography data, specifically metallic foreign bodies in human tissues and organs. The purpose of the study is to develop a mathematical model and a method for automated, high-precision measurement of the linear dimensions of three-dimensional objects based on computed tomography results, with implementation in the form of software. Objectives: to formulate a mathematical model for representing a three-dimensional object in voxel space; to develop a method for segmenting metal fragments based on computed tomography results; to propose a method for spatial alignment of segmented objects based on principal component analysis; develop a method for determining the maximum linear dimension of an object in a new coordinate system; implement the proposed model and method as software and experimentally verify the measurement accuracy. Research methods: analysis of tomographic parameters, threshold segmentation with adaptive selection of threshold values, wave-based algorithm for finding connected components, principal component analysis to determine the object’s orientation, voxel modeling, and calculation of Euclidean distances between boundary points of a three-dimensional object. Results. A mathematical model for representing a three-dimensional object and a method for automated high-precision determination of its maximum linear dimension are proposed. The method was implemented as a software module and tested on 72 samples of metal fragments of six types embedded in the biological tissues of pig organs. The average deviation does not exceed 3%, and in the most complex cases remains within 5%, which indicates the high accuracy and stability of the proposed approach. Conclusions: The developed model and method ensure automated and objective determination of the linear dimensions of foreign bodies based on computed tomography data without operator intervention. The proposed software can be used in military medicine, forensic medical examination, disaster medicine, and healthcare facilities where the speed and reliability of diagnostic decisions are critically important.

Explanation Detected Brain Tumours in MRI Images Using YOLOv8 with LIME-Based Interpretation

2026-03-31T22:33:24+03:00

Relevance. Precise identification of brain tumours in magnetic resonance imaging (MRI) is a critical task in medical image analysis. Although deep learning–based object detectors achieve high localisation accuracy, their limited transparency restricts trust and routine adoption in clinical practice, highlighting the need for explainable artificial intelligence (XAI) approaches. Object of research. The object of this research is the automated detection of brain tumours in MRI scans using convolutional neural network – based object detection models. Subject of research. The subject of the research is the integration of YOLOv8 object detection models with the Local Interpretable Model-Agnostic Explanations (LIME) method to interpret individual detection outputs in medical imaging. Purpose. The aim of this paper is to develop and evaluate an explainable framework for brain tumour detection in MRI images by integrating YOLOv8-based object detection with LIME-based interpretation and by quantitatively assessing the quality of the generated explanations. Methods. Two YOLOv8 variants (YOLOv8n and YOLOv8s) were trained and evaluated on a publicly available MRI dataset containing glioma, meningioma, and pituitary tumour classes. LIME was applied to generate superpixel-based, box-conditioned local explanations for individual detections. Detection performance was assessed using precision, recall, mAP@50, and mAP@50–95. Explanation quality was quantitatively evaluated using stability, sparsity, maximum superpixel weight, and entropy metrics. Results. Experimental results demonstrate that both YOLOv8 models achieve high detection performance, with YOLOv8s providing slightly improved accuracy. LIME successfully highlights image regions that most influence model decisions, and the proposed quantitative metrics confirm that the generated explanations are stable, informative, and aligned with clinically relevant tumour regions. Conclusions. The proposed framework provides a practical approach for combining accurate tumour localisation with interpretable and quantitatively validated explanations, supporting reliability-oriented evaluation of AI-based clinical decision-support systems.

Modeling of adaptive UAV route control based on reinforcement learning algorithms

2026-03-31T22:48:22+03:00

Subject matter is the reward function, action policy, and learning dynamics of the Proximal Policy Optimization (PPO) algorithm in the task of adaptive UAV navigation under dynamic airspace conditions and limited energy resources. Goal is to develop a simulation environment and a modified Proximal Policy Optimization (PPO) model for adaptive route management of a single UAV in 2D and 3D environments, considering the distance to the target, collision risk, and energy consumption. Tasks: to develop 2D and 3D simulation environments with different obstacle configurations and UAV motion parameters; to formulate a combined PPO reward function that incorporates distance to the target, collisions, and energy consumption; to implement and train PPO, DQN, and A2C algorithms in standardized navigation scenarios; to perform a comparative analysis of algorithm performance using key metrics: path length, number of collisions, reward, and energy consumption; to conduct statistical validation of the results using the t-test and confidence intervals; to analyze the influence of PPO hyperparameters on policy stability and learning convergence in 2D and 3D environments. Methods: deep reinforcement learning algorithms (PPO, DQN, A2C); two simulation models (2D and 3D) with randomly generated static obstacles were developed; a combined reward function was formulated, integrating distance-to-target progress, collision penalties, and an energy-related component; model performance was evaluated using average reward, path length, number of collisions, and total energy expenditure; statistical significance was assessed using the t-test and 95% confidence intervals. Results: The modified PPO model reduced the number of collisions in the 2D environment by 94,8% and shortened the route length by 94,3% compared to the baseline PPO, while exhibiting higher energy consumption due to more complex avoidance maneuvers. In the 3D environment, similar trends were confirmed, including improved navigation safety, more stable policy behavior, and statistically significant improvements across key metrics (p < 0,05). Conclusions: A unified 2D/3D simulation environment for adaptive UAV routing and a modified PPO model with a combined reward function were developed. In the 2D environment, the model achieved a ≈94,8% reduction in collisions, a ≈94,3% reduction in path length, and a ≈92,5% increase in average reward compared to the baseline PPO. In the 3D environment, analogous improvements and statistically significant gains (p < 0,05) were obtained. A relationship between avoidance aggressiveness and energy consumption was identified, enabling selection of an optimal policy for BVLOS scenarios.

Evaluation of the effecAtiveness of the enchanced greedyand delta-debugging test case optimization algorithms for C++ libraries

2026-03-31T22:56:00+03:00

Efficient optimization of test cases (TCs) is a necessary condition for improving the efficiency of testing C++ libraries. The subject of the research is test case optimization algorithms for C++ libraries. The purpose of this work is to improve the efficiency of C++ libraries testing by enhancing classical algorithms for greedy TC optimization and delta-debugging TC minimization. Goals. To improve the greedy TC optimization algorithm and eliminate its determinism, ensuring effective TC compression while preserving branch code coverage. To improve the delta-debugging TC minimization algorithm under conditions where redundant actions are sparsely distributed within TCs. To evaluate the effectiveness of the proposed algorithms in comparison with the classical algorithms. Methods. The study applies a greedy TC optimization algorithm, a delta-debugging TC minimization algorithm, mathematical modeling, and statistical analysis methods. The effectiveness of the improved algorithms is evaluated based on statistical analysis of 100 simulation runs for each algorithm on two open-source C++ libraries of different structural complexity. Results. The evaluation results indicate that the improved algorithms provide preservation (and possible increase) of branch coverage with coverage retention coefficient of up to 1.058, increase the TC compression ratio up to 0.86, and reduce the TS execution time by 1.5–2.5× compared to the classical algorithms. Conclusions. The improved algorithms significantly reduce the testing time of C++ libraries without loss of branch coverage. An improved greedy TC optimization algorithm is proposed, in which the selection of TC actions accounts for their future utility estimated by the expected increase in branch coverage. This approach removes the determinism of the classical greedy algorithm, increases the informativeness of TCs, and provides a balance between preserving branch coverage and TC compression. An improved delta-debugging algorithm is proposed that performs group removal of non-unique TC actions according to their contribution to branch coverage, which makes it possible to substantially reduce the length of TCs without losing testing effectiveness.

Predicting Risks of Cardiovascular Disease on Small Datasets using Feature Engineering

2026-03-31T23:10:27+03:00

Relevance. Cardiovascular diseases remain a leading cause of mortality globally, creating a high demand for automated diagnostic systems. However, developing reliable machine learning models for electrocardiogram (ECG) analysis is often hindered by the availability of only small-scale and imbalanced datasets, which limits the effectiveness of deep learning approaches. The object of research is the process of automated processing and classification of electrocardiographic signals for diagnostic purposes. The subject of the research includes methods of beat-centric feature extraction, patient-level aggregation strategies, and machine learning algorithms for cardiovascular risk prediction. The purpose of this paper is to develop and evaluate a reliable classification framework, optimized for small datasets, that increases prediction accuracy by leveraging patient-level feature aggregation and explainable machine learning models. To achieve this goal, the following tasks were solved: 1) implementation of a robust preprocessing pipeline using a refined Pan-Tompkins algorithm for precise beat-centric segmentation; 2) development of a statistical feature aggregation strategy to mitigate local signal variability; and 3) optimization and validation of a Random Forest classifier. The methodology employed includes digital signal processing (Butterworth filtering), advanced feature engineering (HRV, Wavelets analysis), and rigorous 10-fold Stratified Cross-Validation to ensure generalization on limited data. Research results. The study proposes a pipeline initiating with standard signal preprocessing, followed by precise R-peak detection and beat-centric segmentation. Physiological features (HRV, wavelet, morphological) are then extracted from individual segments and statistically aggregated at the patient level. Experiments on a dataset of 164 subjects demonstrated that the proposed patient-level aggregation strategy significantly outperformed traditional segment-based analysis. The final Random Forest model achieved an ROC-AUC score of 0.84. Feature importance analysis confirmed the critical role of Heart Rate Variability (HRV) metrics, particularly SDNN and RMSSD, in differentiating between healthy and high-risk subjects.

Adaptive Resource Allocation Method for the Mobile Fog Layer of High-Density Industrial Internet of Things in Industry 5.0 Networks

2026-03-31T23:25:03+03:00

Relevance of the article. The modern concept of Industry 4.0 laid the foundation for complete digitalization through the industrial Internet of Things. However, the transition to Industry 5.0 requires greater flexibility and resilience of systems. High-density mobile industrial IoT with a fog layer is a critical element of this transformation, as it provides not only automation but also the adaptability of production to human needs and environmental standards. The object of study is the process of pre-processing transactions of the HDIoT edge layer. The main hypothesis of the study: the implementation of a new adaptive method of resource allocation for mobile devices of fog clusters will reduce the average pre-processing time of transactions of the HDIoT edge layer. The goal of the work is to reduce the average time a transaction of the HDIoT peripheral layer spends in the fog layer by developing an adaptive method for distributing the resources of mobile devices in fog clusters. Research objectives: to identify the architectural features of fog computing in HDIoT networks; to create a mathematical model of the process of optimal resource allocation for mobile cluster devices in the fog layer; to formalize a multi-agent approach to cluster resource allocation; to develop and investigate a theoretical game model for managing the resources of a mobile fog cluster of a multi-layer IoT. Methods used: multi-agent approach, game theory, in particular, optimization of a cooperative stochastic game, computer modeling. Results. An adaptive method for distributing resources of mobile devices in fog clusters has been developed. Within the framework of the method, the architecture of a mobile fog cluster has been proposed and a mathematical model of the process of optimal distribution of its resources has been created. In addition, a multi-agent approach is used to find an approximate solution to the formulated two-parameter nonlinear optimization problem, and a game-theoretical approach is implemented to reduce computational complexity and accelerate the search for an approximate solution. Conclusion. As a result of applying the developed method, the average time a transaction of the peripheral layer of a high-density IoT spends in the fog layer has been reduced, which, given the high density of mobile devices, has made it possible to meet QoS requirements.

Architecture of cyberphysical systems for UAV-based late-fusion defect detection in photovoltaic modules

2026-03-31T23:38:06+03:00

The subject of research is the architectural advancement of inspection systems for large-scale solar power plants. As global solar infrastructure expands, reliance on manual or offline analytical methods creates significant operational bottlenecks. The goal of research is to improve the operational utility of unmanned aerial vehicle (UAV)-based photovoltaic module inspection by developing a cyber-physical system (CPS) architecture. It integrates onboard deep learning, edge nodes, cloud analytics, and Supervisory Control and Data Acquisition (SCADA)-aware decision-making into a single coordinated workflow. The tasks of research: 1) formalise a multi-tiered CPS architecture (UAV-edge-cloud) and define interfaces for data, geo-tags, and alarms; 2) develop and validate an onboard thermographic detection pipeline with palette-aware fusion; and 3) integrate detection results with a SCADA-aware logic layer for hazard inference and fire risk mitigation. The methods of research: Computer vision and deep learning (YOLOv11) are used for onboard defect segmentation. Model ensembling via a late-fusion strategy for M2 and M3 thermal palettes mitigates domain shift. RTK-supported spatial clustering algorithms ensure precise geo-indexing and deduplication, and deterministic Boolean logic assesses fire risks based on bypass diode states. The results obtained with five-fold cross-validation shows the proposed architecture significantly outperforms single-modality baselines. The onboard YOLOv11 model achieved a macro mAP@0.5 of 0.91 and 0.90 for M2 and M3 palettes, respectively. The late-fusion ensemble elevated mAP@0.5 for cracks to 0.96 and delamination to 0.95. It reduced end-to-end per-frame processing latency from 4.235 s to 2.858 s. Field validation demonstrated an error of 0.71 defects per inspected string compared to manual counts. Sensitivity analysis highlighted that a 10 m flight altitude provides an optimal balance, yielding 93% precision and 90% recall. Conclusions: Treating UAV inspection as an integrated cyber-physical service improves defect detection. This offers a scalable, real-time solution for preventive maintenance and automated fire-risk mitigation in renewable energy.

Method for predicting UAV trajectories and evasion for industrial autonomous missions in a dynamic environment

2026-03-31T23:49:48+03:00

This paper addresses the problem of predicting the trajectory and anticipatory evasion of an unmanned aerial vehicle (UAV) in industrial autonomous missions in a dynamic environment, subject to noise in navigation measurements and energy constraints, which pose risks of delayed or excessive maneuvering and increased deviation from the route. Objective. To develop and verify, using a simulation model, a method that ensures prediction-based obstacle avoidance with control of deviation from the reference trajectory and maneuvering energy consumption. Tasks. Formulate the architecture of the avoidance system; develop a predictor of future coordinates based on a recurrent neural network with long-term short-term memory; determine a method for assessing collision risk using a safety zone; implement a trajectory correction algorithm taking into account the “safety–deviation–energy consumption” trade-off; perform a comparative evaluation with baseline methods. Methods. Coordinate predictions are constructed based on time sequences of coordinates and motion parameters; collision risk is assessed by analyzing the intersection of the predicted trajectory with obstacle safety zones; trajectory correction is formalized as an optimization problem to select a maneuver that minimizes the total tracking error and the proximity penalty. The effectiveness was verified in a Python environment on standard trajectories (straight, circular, and polygonal) by comparison with pure tracking, line-of-sight, vector field, and nonlinear stabilization methods. Results. The proposed approach achieved the smallest mean deviation from the trajectory (14.95 m), the lowest maneuver energy consumption (72 conventional units), the highest tracking success rate (86.08%), and the highest overall productivity coefficient (0.494) among the algorithms considered; a trade-off was observed regarding the minimum distance to obstacles. Conclusions. The prediction-oriented evasion method improves the overall navigation efficiency in industrial mission models; further research involves field validation on real platforms and optimization of the predictor’s computational costs.

An approach to generating physically realistic data for crane safety monitoring systems based on finite element modeling

2026-03-31T23:57:15+03:00

Предметом дослідження є стійкість монорейкового пересувного крана, зокрема коефіцієнти на перекидання й на зсув. Проблема полягає в неадекватності статичних моделей, що базуються на припущенні про жорсткість конструкції, для визначення безпеки. Такі моделі ігнорують вплив деформацій, що призводить до неконсервативних оцінок стійкості та робить їх ненадійними для генерації даних для систем моніторингу на основі моделей штучного інтелекту. Метою дослідження є розроблення та обґрунтування підходу до створення даних для оцінювання стійкості монорейкових пересувних кранів, який ґрунтується на скінченно-елементному моделюванні та бере до уваги деформаційні ефекти конструкції. Для досягнення мети в статті визначено такі завдання: розроблення підходу до створення наборів даних для оцінювання стійкості монорейкового крана на основі скінченно-елементного моделювання; перевірка запропонованого підходу способом кількісного порівняльного аналізу згенерованих даних із даними, отриманими за статичним підходом, і доведення статистичної значущості розбіжностей. Для реалізації окреслених завдань використано такі методи: теорія експериментів, теорія балок Ейлера – Бернуллі, статистична теорія розподілів, теорія чисельного моделювання. Досягнуті результати. Отримано кількісні докази розбіжності між статичною та скінченно-елементною моделями. Для стійкості на перекидання середнє значення коефіцієнта β за скінченно-елементною моделлю становить 1.365 (на 2.4% нижче, ніж у статичній – 1.398). Для стійкості на зсув середній коефіцієнт γ зменшився майже удвічі – з 1.87 до 0.85. Статистичний аналіз підтвердив високу значущість цих розбіжностей (p < 0.001). Висновки. Скінченно-елементна модель бере до уваги геометричну нелінійність. Пружна деформація стріли змінює плече дії сили ваги вантажу, збільшуючи перекидальний момент і горизонтальну реакцію, що системно знижує запас стійкості. Відмінною рисою запропонованої моделі є її здатність виявляти "небезпечні зони" – режими роботи, які статична модель хибно ідентифікує як безпечні. Сферою практичного використання є розроблення систем моніторингу безпеки монорейкового крана на базі штучного інтелекту. Запропонована модель може бути застосована як "генератор" синтетичних даних для навчання та перевірки моделей машинного навчання, що дасть змогу створювати предиктивні системи для оцінювання стійкості монорейкового крана в реальному часі.

A method of semantic search for educational content based on multi-agent technologies

2026-04-01T00:04:02+03:00

The digital transformation of industry is accompanied by the active adoption of new technologies and the rapid evolution of production processes. A significant portion of educational materials is distributed across various information sources, including internal corporate systems, open educational platforms, and specialized web resources. Such resources often contain duplicates, redundant information, and heterogeneous metadata, which complicates the timely retrieval of relevant learning materials. The subject of the study is a method of semantic search for educational content in a distributed information environment using ontology-based knowledge models. The goal of the work is to investigate a method of semantic search for educational content in a distributed information environment based on a multi-agent organization of information resource processing and the use of ontology-based knowledge models. The objectives of the study are: to investigate the architectural model of a multi-agent search system; to develop a semantic selection algorithm based on the comparator identification method and an ontology-based model; to formalize a relevance evaluation predicate considering weighted metadata coefficients; to develop a multi-agent software platform; and to experimentally evaluate performance and resource consumption under different agent operating modes. Research methods include: the method of multi-agent organization of information resource processing with non-blocking message exchange; three-level URL deduplication; ontology-based term matching and a formalized relevance evaluation predicate; and experimental measurement of processing time, the number of processed links, and system resource consumption. Results: a model of a multi-agent system with four types of agents and a semantic search algorithm eliminating loops and duplicate links has been proposed; a software platform based on Kotlin using coroutines and asynchronous interaction between agents has been implemented; experimental results demonstrate that the proposed organization of processing provides higher performance compared to the sequential mode. Conclusions: the integration of semantic search and a multi-agent architecture enables efficient organization of the process of discovering and processing educational content in a distributed environment. The proposed method ensures coordinated operation of agents, eliminates link duplication, and provides a rational balance between search completeness and the use of computational resources.

Digital HR management system for security structures based on the employee's e-cabinet

2026-04-01T00:11:15+03:00

The subject is methods and models, operational processes of digitalization of HR management systems in security-oriented systems. The goal of the work is to develop methods and models for managing projects of digitalization of operational processes of HR management systems in security structures to increase their efficiency and adaptability. The article addresses the following tasks: a conceptual model of digital transformation of HR processes in security-oriented systems has been developed, a model of digitization of personnel processes based on the creation of the Rescuer+ e-cabinet has been developed, and algorithms for the automated generation of electronic contracts and their signing using electronic digital signatures have been developed and described. The following methods are used: modeling methods, in particular, graphical and structural-logical modeling, which made it possible to build a comprehensive model of digital transformation and a centralized management model. The following results were obtained: a conceptual model of digital transformation of HR processes in security-oriented systems was developed, which formalizes the interaction of personnel, organizational, and information security components. Based on the proposed model, a method for digitizing HR processes was substantiated and developed by creating the Rescuer+ employee e-cabinet, which ensures the centralization and automation of key HR operations. As part of the implementation of the method, an algorithm for the automated generation, storage, and signing of electronic contracts using a qualified electronic signature has been developed and described, which guarantees the legal significance and integrity of electronic document management. The process of managerial decision-making has been improved thanks to consolidated access to real-time data and HR analytics capabilities, the ability to use the developed models and methods as a basis for planning and managing digital transformation projects in other law enforcement agencies in Ukraine, adapting them to the specifics of particular departments. Conclusions: The implementation of a coordinated strategic policy for the development and use of information data will contribute to the modernization and acceleration of decision-making processes in various aspects of security-oriented systems. Thus, the above emphasizes the importance of digital technologies in the implementation of public service priorities, in particular their accessibility to personnel, as well as increasing efficiency and implementing innovative solutions.