This study proposes a supervised machine learning framework for vibration-based fault diagnosis of rotating machinery using integrated data-driven and physics-informed feature sets. A dataset acquired under variable load and multiple operating conditions was used for model training. Parallel signal processing techniques were applied to capture fault-related information across multiple frequency bands including time-domain analysis, frequency-domain analysis, baseband analysis, and envelope analysis. From the corresponding signal representations, statistical, spectral, and physics-based features associated with characteristic fault frequencies were extracted and combined into integrated feature sets. The diagnostic performance of models trained using purely data-driven features was systematically compared with models incorporating integrated data-driven and physics-informed features. Support Vector Machine, Random Forests, Gradient Boosting, and an ensemble classifier were evaluated using accuracy, precision, recall, and F1-score metrics. The proposed framework employs a two-layer classification strategy, where the first layer performs multiclass fault identification, while the second layer evaluates the presence of imbalance as a coexisting fault. In addition, the influence of different feature groups as well as individual measurement axes and their combinations on diagnostic performance were analyzed. Validation using a new dataset measured in laboratory conditions confirmed the robustness and generalization capability of the proposed diagnostic framework.

Urban noise is a major environmental concern that affects public health and quality of life, demanding new approaches beyond conventional noise level monitoring. This study investigates the development of an AI-driven Acoustic Event Detection and Classification (AED/C) system designed for urban sound recognition and its integration into smart city application. Using the UrbanSound8K dataset, five acoustic parameters—Mel Frequency Cepstral Coefficients (MFCC), Mel Spectrogram (MS), Spectral Contrast (SC), Tonal Centroid (TC), and Chromagram (Ch)—were mathematically modeled and applied to feature extraction. Their combinations were tested with three classical machine learning algorithms: Support Vector Machines (SVM), Random Forest (RF), Naive Bayes (NB) and a deep learning approach, i.e., Convolutional Neural Networks (CNN). A total of 52 models with the three ML algorithms were analyzed along with 4 models with CNN. The MFCC-based CNN models showed the highest accuracy, achieving up to 92.68% on test data. This achieved accuracy represents approximately +2% improvement compared to prior CNN-based approaches reported in similar studies. Additionally, the number of trained models, 56 in total, exceeds those presented in comparable research, ensuring more robust performance validation and statistical reliability. Real-time validation confirmed the applicability for IoT devices, and a low-cost wireless sensor unit (WSU) was developed with fog and cloud computing for scalable data processing. The constructed WSU demonstrates a cost reduction of at least four times compared to previously developed units, while maintaining good performance, enabling broader deployment potential in smart city applications. The findings demonstrate the potential of AI-based AED/C systems for continuous, source-specific noise classification, supporting sustainable urban planning and improved environmental management in smart cities.

The increasing demand for autonomous wireless sensors in Internet of Things (IoT) applications has intensified research on vibration energy harvesting, particularly in the low-frequency range where ambient vibrations are most prevalent. However, most vibration energy harvesters operate efficiently only at a single resonance mode, resulting in a narrow operational bandwidth and pronounced performance degradation under frequency detuning. To address this limitation, this paper proposes a multimodal hybrid piezoelectric–electromagnetic vibration energy harvester that exploits both the first and second resonance modes of a cantilever-based structure to achieve broadband low-frequency operation. The design is guided by the complementary utilization of strain-dominated and velocity-dominated regions associated with different vibration modes. Numerical modeling and finite element simulations are employed to investigate the influence of mass distribution, deformation characteristics, and relative velocity on energy conversion performance. A secondary cantilever carrying the electromagnetic coil is introduced to enhance the relative motion between the coil and the magnetic field, thereby extending the effective operational bandwidth. The experimental results demonstrate increased harvested power, improved energy conversion efficiency, and a significantly broadened effective frequency range compared to conventional single-mode piezoelectric and electromagnetic energy harvesters.

The lockdown that the city of Skopje has suffered during the months of April to June 2020 in order to control the spread of COVID-19 has significantly changed the acoustic environment in the urban parts of the city. The absence of vehicles, people on the streets and closed restaurants has led to a noise reduction captured by the low-cost wireless sensor network in the City of Skopje. The analysis carried out in this paper show reduction in noise pollution strongly correlated with the population's activity and behavior to the new circumstances. Overall, the sound pressure levels vary around 65 dB; however, some extreme decrease can be noticed, especially during the quarantine weekends. Also, the noise levels were compared between the same time periods during the year of 2019 and 2020, where it can be perceived reduction in the sound level for 36,5% for the day-evening-night noise level indicator (L). Significant variations occur for the indicators L, L, and L, especially during the lockdown weekends.

In an era of increasing digital connectivity and device autonomy, the Internet of Things (IoT) represents a fundamental pillar of modern digital transformation. As the number of sensors and devices operating in environments without a stable power supply continues to grow, the demand for sustainable and autonomous energy solutions becomes critical. Efficient conversion of vibrational energy into electrical energy requires a carefully optimized mechanical design to ensure effective energy transfer to the transducers, particularly under low-frequency and small-amplitude conditions typical of real-world IoT applications. This study presents a comprehensive multiphysics investigation and energy-dynamic characterization of an elastic cantilever system integrating piezoelectric and electromagnetic transducers for vibration energy harvesting. Analytical and numerical models are developed to describe the coupled mechanical-electrical behavior of both harvesting mechanisms, supported by detailed finite element simulations performed in COMSOL Multiphysics. The dynamic response of the cantilever is analyzed for multiple mass configurations. The results demonstrate that modifying the mass distribution significantly shifts the natural frequencies and redistributes kinetic and potential energy across vibration modes, enabling optimized placement of energy transducers. The findings confirm the potential of hybrid cantilever-based harvesters to provide efficient, multifunctional, and sustainable energy solutions for self-powered IoT and smart monitoring systems.

Urban air pollution in Skopje, a city with complex topography, is strongly influenced by traffic emissions, household heating, industrial activities, and meteorological conditions, leading to pronounced spatial and seasonal variability. The objective of this study is to assess the contribution of major urban emission sources to air quality in Skopje, with a focus on traffic pollution, and to quantify their seasonal influence on NO₂, PM₁₀, and PM_2.5 concentrations using a high-resolution urban dispersion modelling approach. The methodology is based on the ADMS-Urban dispersion modelling system, integrating traffic activity data as line sources, together with area sources representing household heating, point sources representing industrial facilities, and seasonally representative meteorological data. Model performance was evaluated through comparison with measurements from official urban monitoring stations. The results show that the model successfully reproduces the observed spatial gradients and seasonal trends of NO₂, PM₁₀, and PM_2.5 concentrations across the urban area. Source contribution analysis indicates that household heating dominates particulate matter pollution throughout the year, while traffic and industrial combustion are the main contributors to NO₂. The isolated traffic contribution exhibits clear seasonal variability, with the highest concentrations occurring during winter due to reduced atmospheric dispersion and increased traffic-related emissions. The model is primarily suitable for assessing spatial patterns and relative source contributions rather than accurate prediction of absolute concentration levels, due to the use of aggregated Tier 1 emission factors. The study confirms that physically based urban dispersion modelling provides a robust framework for identifying pollution hotspots, quantifying traffic contributions, and supporting targeted air quality management strategies in Skopje.