With industrial and residential electricity demands on the increase, electrification plays a fundamental role in providing the necessary infrastructure for this constant evolution, which fuels several sectors that are important to society. To distribute electricity on a large scale and over long distances, high-voltage transmission systems, which operate at very high voltages with multiple cables, are commonly used worldwide. These cables must be spaced apart to ensure that they do not come into contact. Thus, in this paper, we develop a robotic module coupled with a drone that is used to install spacers. Spacers are elements that are used between various cables within a power distribution system to keep them apart. The drone transports the robotic module to the cable, controls the robotic module using radio signals, and installs/uninstalls the spacer. After completing the procedure, the drone searches for the robotic module, returning it to the ground station to repeat the operation if necessary. This paper describes the development of the mathematical model followed by the CAD/CAE design. Computer simulations verified the feasibility of using this robotic module for installing spacers. It is noteworthy that this module can be placed on cables using a drone or a hot stick, eliminating the need for technicians to come into direct contact with the cables or to move along them. This drone–robot aims to reduce risks for technicians who carry out these operations by climbing the towers and moving on cables or in some cases may be suspended from a platform fixed to a helicopter.

Upper extremity injuries caused by trauma and repetitive movements are common among elderly individuals, manual labor workers, and athletes. A traditional rehabilitation program to restore range of motion and strength consists of massage, physiotherapy, and mechanotherapy. Although a great variety of commercial devices is available, the majority lack sensorization for performance monitoring. This paper presents a sensor-integrated rotating crank mechanism for upper limb rehabilitation by comparing the performance of three motor configurations and by validating the system's capability through quantitative motion parameters during exercise. Three prototype configurations were developed and tested. The first prototype (V0) was tested in passive mode without motor activation, then the stepper motor in (V1) was activated, and testing was carried out in semi-active mode. The third prototype (V3) was designed and tested in motorized mode. A 6-axis IMU sensor was integrated onto the crank to capture acceleration and orientation data. Testing was conducted with six volunteers (3 male, 3 female, aged 21-26 years) performing exercises in both horizontal and sagittal planes. The V0 configuration demonstrated predictable motion patterns with an acceleration magnitude ranging from approximately 9 to 13 m/s². Sagittal plane rotation exhibited higher variability with the range 5-20 m/s² due to gravitational effects. The V2 configuration introduced vibration and irregularities in motion smoothness with a magnitude acceleration range of 12 to 13.5 m/s². The most consistent performance was demonstrated by the brushless DC motor system (V3) with stable acceleration profiles (7-18 m/s² horizontal, 4-20 m/s² vertical). Acquired data revealed gender-related differences in peak acceleration, where male volunteers exhibited higher acceleration peaks, especially in the sagittal plane, where the Az component reached up to 25 m/s², whereas female volunteers showed lower and smoother acceleration profiles. The sensor-integrated system provided a reliable method for acquiring quantitative performance metrics, establishing a viable foundation for monitoring rehabilitation progress.

The continuous monitoring of patients affected by Alzheimer’s disease requires autonomous and reliable machine-based systems capable of operating under strict energy, size, and safety constraints. This work proposes an intelligent mechatronic architecture for an implantable monitoring device integrating embedded biomedical sensors, low-power processing units, secure wireless communication, and artificial intelligence for real-time data analysis. To address the limited availability of clinical datasets, a digital twin-based synthetic data generation framework is developed. The proposed system is evaluated on a multidimensional dataset composed of 10, 235 records, including physiological, behavioral, and cognitive parameters, with an 80/20 train-test split. Random Forest, Support Vector Machine, and Deep Neural Network models are implemented and compared using standardized classification and regression metrics, including accuracy, precision, recall, F1-score, confusion matrices, and error-based indicators. The experimental results show that the Deep Neural Network consistently outperforms classical machine learning models, achieving higher classification accuracy, reduced misclassification rates, and more stable convergence behavior, as confirmed by learning and loss curves. From a mechatronics perspective, the proposed solution emphasizes modular system integration, computational efficiency, and compatibility with implantable hardware constraints. The results demonstrate the feasibility of embedding intelligent decision-making capabilities into compact mechatronic systems, highlighting their relevance for intelligent machines and continuous monitoring applications.

Introduction: Severe weather events such as floods are increasing in frequency and pose significant risks to homes, infrastructure, and human safety. Traditional domestic flood protection systems often require manual setup or user intervention, limiting their effectiveness during sudden flood events. Addressing this gap, we propose an integrated automated barrier system that is designed to reduce internal flooding and associated damage for residences located on sloped streets.

Methods: The developed system combines mechanical components with sensor-driven automation to detect rising floodwater and react in real time. The key elements include an array of hinged diverting plates (lintels) and a vertical sealing barrier installed at the entrance threshold. When sensors register imminent water ingress, the system autonomously deploys the lintels to redirect surface flow away from the doorway, creating a localized dry area. Simultaneously, the sealing barrier engages to prevent water from penetrating the the building's interior. The design process incorporated standard door dimensions and complied with relevant safety and automation regulations.

Results: Prototype testing under simulated flood conditions demonstrated consistent activation without human intervention, effective water diversion away from the entryway, and reduced internal water penetration compared to conventional static barriers. The automated control logic reliably interpreted sensor inputs, triggering timely deployment and retraction of barrier elements.

Conclusions: The intelligent automated barrier shows promise as a proactive residential flood mitigation solution that enhances response times and protects property with minimal human input. Future work will refine sensor calibration and assess long-term field performance across diverse flood scenarios.

While bipedal locomotion for humanoid robots has improved significantly over the past 50 years of dedicated research, energy efficiency and load capacity remain challenging considerations in locomotion tasks. The LARMbot V.3 humanoid addresses these issues with custom parallel architectures in a low-cost and compact platform (850 mm in height, 462 mm in leg length, and 3.6 kg in weight) mainly for research and education purposes.

This work presents a performance evaluation of the LARMbot V.3 modular bipedal locomotion system, which is based on a parallel-serial (3-UPR)R mechanism for each leg. This structure offers proper precision, rigidity, and load capacity while maintaining the desired walking performance and movement stability. After we introduce the architecture, a prototype is presented and we describe the design, manufacturing, and assembly. Walking performance tests were conducted in two modes—free motion in the air and with ground contact—to measure power consumption, speed, and movement repeatability.

The tests demonstrated a stable and repeatable gait cycle with a step length of 80 mm, a step height of 20 mm, and a speed of 5 seconds per step. The acquired IMU data confirmed synchronized leg motion and hip orientation deviations within ±15°. Power measurements showed consistent and low values during the tests, with an average power consumption of 5.11 W. These results confirm the efficiency and reliability of the proposed LARMbot V.3 parallel biped locomotion system. This design enables stable and precise movement while maintaining low power consumption, and it is easy to manufacture. The module can serve as a foundation for further improvements and development of parallel locomotion for humanoid robots.

Collaborative manipulators are often deployed as robotic systems that are intended to operate safely and intuitively within shared workspaces alongside human users. Their effectiveness depends not only on mechanical design, but also on the reliability of the underlying control architecture, the quality of sensor feedback, and the system’s ability to adapt to human interaction in real time. This study provides a structured review of modern control strategies used in collaborative manipulators, examining the theoretical principles and practical implementation of impedance, admittance, hybrid, and sensor-based approaches. An analysis of the existing literature shows that impedance and admittance control techniques are particularly well suited for human–robot collaboration, as they help to maintain stability during contact while enabling compliant and responsive motion that aligns with human intent.

This paper introduces a conceptual multi-layer control architecture that integrates safety supervision, trajectory planning, and sensor fusion. Within this architecture, a hybrid control scheme that combines force and motion sensing is highlighted as a promising direction for achieving adaptive behavior. Such an approach supports real-time adjustment of dynamic parameters and aligns with the safety limits defined in ISO/TS 15066, ensuring controlled contact forces and safe motion near human operators.

Overall, the presented framework offers a comprehensive theoretical foundation for further development of adaptive and sensor-rich control systems. These insights are expected to contribute to subsequent simulation studies, prototype testing, and the broader implementation of collaborative robots in industrial, medical, and human-assistive applications.

The large-scale integration of renewable energy sources (RESs), electric vehicles (EVs), and battery energy storage systems (BESSs) has significantly reduced system inertia and intensified frequency–voltage coupling in modern interconnected power systems, thereby challenging the effectiveness of conventional secondary control strategies. To address these emerging issues, this paper proposes an intelligent, control-aware evolutionary multi-objective Neural Architecture Search (EMO–NAS) framework for coordinated Load Frequency Control (LFC) and Automatic Voltage Regulation (AVR) in renewable-dominated multi-area power systems. Unlike existing approaches that rely on fixed or heuristically selected controller structures, the proposed framework treats the controller architecture itself as an explicit decision variable and autonomously synthesizes task-specific control policies through multi-objective evolutionary optimization. The coordinated LFC–AVR problem is formulated by simultaneously minimizing frequency deviation, tie-line power oscillations, voltage deviation, rate of change of frequency (RoCoF), control effort, and robustness degradation, while satisfying practical operational constraints including generation rate limits, actuator bounds, and BESS state-of-charge restrictions. A structured NAS search space incorporating feedforward, recurrent, and temporal architectures is evaluated using closed-loop time-domain simulations under realistic disturbances, renewable intermittency, EV variability, and parameter uncertainty. Feasibility and stability are enforced through constraint-aware penalties and robust domain randomization. Comprehensive simulation studies on three- and four-area interconnected systems demonstrate that the proposed EMO–NAS controller achieves substantial performance improvements compared with optimally tuned fractional-order PID, robust sliding mode, and fixed-architecture neural controllers. Quantitatively, reductions of approximately 30–35% in frequency deviation, 25–35% in tie-line power oscillations, and up to 30% in RoCoF are achieved, while completely eliminating constraint violations. Robustness analysis under ±50% parameter uncertainty and Monte Carlo simulations further confirm superior stability, generalization, and scalability. These results establish architecture-level optimization as a powerful and systematic pathway for designing robust, coordinated secondary controllers in future low-inertia, renewable-dominated power systems.

Silicon carbide (SiC) is a wide-bandgap ceramic material recognized for its thermal stability, mechanical robustness, and optical reliability, making it highly suitable for machine-integrated functional components. In this study, SiC nanotubes synthesized via carbothermal reduction at 1800 °C were incorporated into a polyvinylpyrrolidone (PVP) matrix to fabricate SiC/PVP nanocomposite films with controlled filler loadings of 1–5 wt%. The structural, morphological, and optical properties were systematically investigated using X-ray diffraction (XRD), scanning electron microscopy with EDS mapping (SEM/EDS), UV–Vis spectroscopy, and FTIR analysis to establish structure–property relationships relevant to engineered machine materials.

At low filler concentrations (1–3 wt%), uniform dispersion of SiC nanotubes induces crystallite refinement, increased microstrain, and enhanced interfacial defect density within the polymer matrix. This results in a gradual reduction in the optical bandgap from 5.78 eV for pure PVP to 5.51 eV at 3 wt% SiC loading. At a higher concentration (5 wt%), nanotube aggregation reduces the effective interfacial area, leading to lower microstrain and a partial recovery of the bandgap to 5.70 eV. FTIR spectra confirm strong interfacial interactions, including hydrogen bonding and dipolar interactions between PVP functional groups and surface –OH/Si–O groups on SiC, without the formation of new chemical bonds. SEM/EDS mapping clearly illustrates the transition from homogeneous dispersion to clustered structures at elevated filler contents.

The results identify an optimal SiC loading of 2–3 wt% for achieving balanced microstructural uniformity and defect-mediated optical tunability. These findings demonstrate that SiC/PVP nanocomposites can be effectively engineered for lightweight machine components, optically functional layers, and integrated dielectric or transparent elements where controlled optical behavior, thermal stability, and polymer–ceramic synergy are essential.

The rapid development of intelligent transportation systems and connected vehicles has led to the generation of massive volumes of diverse and heterogeneous traffic data. Efficient analysis, interpretation, and classification of this data are crucial for enhancing mobility, traffic prediction, and safety in modern transportation networks. Recent studies have demonstrated that deep learning models are capable of effectively capturing both spatial and temporal dependencies in traffic datasets, enabling more accurate and reliable analysis and classification compared to traditional methods. However, most existing approaches focus solely on software-based implementations, often overlooking the practical challenges of deploying these models in real-time, resource-constrained embedded environments.

This review provides a comprehensive analysis of deep learning approaches applied to vehicular traffic data classification. It emphasizes the evaluation of model effectiveness, computational efficiency, and suitability for implementation in embedded systems, highlighting various optimization and adaptation strategies that make deployment feasible in hardware-constrained contexts.

The study also highlights current research trends, identifies critical open challenges in achieving real-time inference on limited-resource hardware, and discusses potential future directions for integrating deep learning methods with embedded systems. By bridging the gap between deep learning model design and practical hardware implementation, this review contributes to the development of intelligent, efficient, and deployable AI solutions for next-generation connected and autonomous vehicles, ultimately supporting safer and more effective transportation networks.

This paper presents the development and evaluation of an automated visual
inspection system applied to the analysis of battery solder joints in an electronic welding
process. In automated manufacturing environments, the quality of soldered connections is
critical in ensuring the electrical reliability, mechanical stability, and long-term
performance of electronic assemblies. Defects in the soldering process may lead to
intermittent connections, premature failures, or complete malfunctions of the final product,
making effective inspection mechanisms essential.
The proposed system focuses on the visual inspection of battery solder joints using
image-based classification techniques. Images acquired during the welding process were
analyzed using a convolutional neural network (CNN) trained to classify solder joints into
two distinct categories: acceptable (OK) and non-acceptable (NG). The inspection is
performed using a vision system based on a standard camera, allowing for the non-contact
and non-destructive evaluation of the solder quality.
The artificial intelligence model was trained and executed using a backend
application developed in Python that handled image acquisition, preprocessing, and
inference. A graphical user interface was implemented using Windows Forms in C# to enable
operator interaction, visualization of inspection results, and integration with the production
environment. This software architecture allows for a clear separation between the machine
learning processing layer and human–machine interface.
The experimental results indicate that the proposed system can reliably distinguish
between conforming and nonconforming solder joints under real production conditions. The
solution demonstrates the feasibility of applying convolutional neural networks for
automated inspection in electronics manufacturing, contributing to improved process
control, enhanced product quality, and reduced dependency on manual inspection.