Technologies

35 pages, 963 KiB

Open AccessArticle

Understanding Technology Perception in Autism with Separate Analyses for Anxiety and Depression Using Quantum Circuit Simulation Approach

by Gema Benedicto-Rodríguez, Vanessa Zorrilla-Muñoz, Nicolas Garcia-Aracil, Eduardo Fernandez and José Manuel Ferrández

Technologies 2025, 13(4), 165; https://doi.org/10.3390/technologies13040165 - 20 Apr 2025

Viewed by 111

Abstract

Background: This work explores the current use of technologies and the perception of their impact on people diagnosed with Autism Spectrum Disorder (ASD) and other comorbidities—Chronic Anxiety (CA) and Chronic Depression (CD). Autistic people often experience anxiety and/or depression. These mental health issues [...] Read more.

Background: This work explores the current use of technologies and the perception of their impact on people diagnosed with Autism Spectrum Disorder (ASD) and other comorbidities—Chronic Anxiety (CA) and Chronic Depression (CD). Autistic people often experience anxiety and/or depression. These mental health issues are exacerbated by social stigma, affecting their quality of life (QoL) and well-being. Aims: The study aims to analyze how emerging technologies can reduce communication difficulties, as well as stress, anxiety, and depression, and thus improve QoL for individuals with ASD and comorbidities like CA and CD. Methods: This study analyzes data from the secondary questionnaire ‘Encuesta de Discapacidad, Autonomía Personal y situaciones de Dependencia (EDAD)’ developed in 2020–2021 by the ‘Instituto Nacional de Estadística (INE)’ for people with ASD (n = 241), ASD and CA (n = 61), and ASD and CD (n = 29). The analysis includes Pearson correlation tests to examine the relationship between various factors affecting QoL. Results: The results highlight differences in difficulties affecting the QoL of ASD persons. Pearson correlation analysis showed significant negative correlations in communication and learning for ASD patients, with similar patterns in the separate analyses of CA and CD. More significant variables were found in the ‘Learning’ and ‘Communication’ indices for ASD, while CA and CD represented more significant variables in ‘Mobility’ index. Conclusions: This work suggests that technological interventions, such as the integration of advanced technologies, could enhance emotional regulation and social skills in individuals with ASD. In this sense, the quantum computing approach could help in the emerging technologies impact evaluation, analyzing devices adapted to the user to optimize their QoL and well-being. Full article

(This article belongs to the Special Issue Assistive Technologies in Care and Rehabilitation: Research, Developments, and International Initiatives)

► Show Figures

Figure 1

22 pages, 8938 KiB

Open AccessArticle

Enhancing Hand Gesture Image Recognition by Integrating Various Feature Groups

by Ismail Taha Ahmed, Wisam Hazim Gwad, Baraa Tareq Hammad and Entisar Alkayal

Technologies 2025, 13(4), 164; https://doi.org/10.3390/technologies13040164 - 19 Apr 2025

Viewed by 193

Abstract

Human gesture image recognition is the process of identifying, deciphering, and classifying human gestures in images or video frames using computer vision algorithms. These gestures can vary from the simplest hand motions, body positions, and facial emotions to complicated gestures. Two significant problems [...] Read more.

Human gesture image recognition is the process of identifying, deciphering, and classifying human gestures in images or video frames using computer vision algorithms. These gestures can vary from the simplest hand motions, body positions, and facial emotions to complicated gestures. Two significant problems affecting the performance of human gesture picture recognition methods are ambiguity and invariance. Ambiguity occurs when gestures have the same shape but different orientations, while invariance guarantees that gestures are correctly classified even when scale, lighting, or orientation varies. To overcome this issue, hand-crafted features can be combined with deep learning to greatly improve the performance of hand gesture image recognition models. This combination improves the model’s overall accuracy and dependability in identifying a variety of hand movements by enhancing its capacity to record both shape and texture properties. Thus, in this study, we propose a hand gesture recognition method that combines Reset50 model feature extraction with the Tamura texture descriptor and uses the adaptability of GAM to represent intricate interactions between the features. Experiments were carried out on publicly available datasets containing images of American Sign Language (ASL) gestures. As Tamura-ResNet50-OptimizedGAM achieved the highest accuracy rate in the ASL datasets, it is believed to be the best option for human gesture image recognition. According to the experimental results, the accuracy rate was 96%, which is higher than the total accuracy of the state-of-the-art techniques currently in use. Full article

(This article belongs to the Section Information and Communication Technologies)

► Show Figures

Figure 1

18 pages, 574 KiB

Open AccessArticle

Leveraging IPv6 and ICMPv6 for Delay-Tolerant Networking in Deep Space

by Umberto Pirovano, Oriol Fusté and Anna Calveras

Technologies 2025, 13(4), 163; https://doi.org/10.3390/technologies13040163 - 18 Apr 2025

Viewed by 99

Abstract

Communications in delay-tolerant networking (DTN) environments like deep space face significant challenges due to immense distances and the intermittent nature of links. Overcoming these issues requires moving beyond the assumptions of immediacy and reliability that underpin traditional terrestrial Internet Protocol (IP) networks. Historically, [...] Read more.

Communications in delay-tolerant networking (DTN) environments like deep space face significant challenges due to immense distances and the intermittent nature of links. Overcoming these issues requires moving beyond the assumptions of immediacy and reliability that underpin traditional terrestrial Internet Protocol (IP) networks. Historically, deep-space networks have relied on custom architectures or protocols like the Bundle Protocol (BP) to address these challenges; however, such solutions impose the constraint that nodes must implement the chosen protocol for proper operation, thereby not providing interoperability with standard IP-based nodes. This paper proposes an alternative approach, leveraging innovations in IP version 6 (IPv6) and Internet Control Message Protocol version 6 (ICMPv6) to integrate delay-tolerant features directly at Layer 3. By embedding these functionalities within the existing IPv6 framework, the proposed IP-compliant solution enhances interoperability, with terrestrial networks enabling DTN nodes to seamlessly communicate with compliant IPv6 nodes. This study provides a detailed comparison of the capabilities of IPv6 and BP version 7, highlighting gaps and opportunities. Based on this analysis, a node architecture is designed to implement the necessary functionalities for DTN, paving the way for more seamless integration of deep-space and terrestrial networks while reducing complexity and improving scalability. Full article

(This article belongs to the Section Information and Communication Technologies)

► Show Figures

Graphical abstract

22 pages, 7303 KiB

Open AccessArticle

Ground Segmentation for LiDAR Point Clouds in Structured and Unstructured Environments Using a Hybrid Neural–Geometric Approach

by Antonio Santo, Enrique Heredia, Carlos Viegas, David Valiente and Arturo Gil

Technologies 2025, 13(4), 162; https://doi.org/10.3390/technologies13040162 - 16 Apr 2025

Viewed by 335

Abstract

Ground segmentation in LiDAR point clouds is a foundational capability for autonomous systems, enabling safe navigation in applications ranging from urban self-driving vehicles to planetary exploration rovers. Reliably distinguishing traversable surfaces in geometrically irregular or sensor-sparse environments remains a critical challenge. This paper [...] Read more.

Ground segmentation in LiDAR point clouds is a foundational capability for autonomous systems, enabling safe navigation in applications ranging from urban self-driving vehicles to planetary exploration rovers. Reliably distinguishing traversable surfaces in geometrically irregular or sensor-sparse environments remains a critical challenge. This paper introduces a hybrid framework that synergizes multi-resolution polar discretization with sparse convolutional neural networks (SCNNs) to address these challenges. The method hierarchically partitions point clouds into adaptive sectors, leveraging PCA-derived geometric features and dynamic variance thresholds for robust terrain modeling, while a SCNN resolves ambiguities in data-sparse regions. Evaluated in structured (SemanticKITTI) and unstructured (Rellis-3D) environments, two different versions of the proposed method are studied, including a purely geometric method and a hybrid approach that exploits deep learning techniques. A comparison of the proposed method with its purely geometric version is made for the purpose of highlighting the strengths of each approach. The hybrid approach achieves state-of-the-art performance, attaining an F1-score of 95.4% in urban environments, surpassing the purely geometric (91.4%) and learning-based baselines. Conversely, in unstructured terrains, the geometric variant demonstrates superior metric balance (80.8% F1) compared to the hybrid method (75.8% F1), highlighting context-dependent trade-offs between precision and recall. The framework’s generalization is further validated on custom datasets (UMH-Gardens, Coimbra-Liv), showcasing robustness to sensor variations and environmental complexity. The code and datasets are openly available to facilitate reproducibility. Full article

(This article belongs to the Special Issue Advanced Autonomous Systems and Artificial Intelligence Stage)

► Show Figures

Graphical abstract

33 pages, 14735 KiB

Open AccessArticle

Artificial Vision System for Autonomous Mobile Platform Used in Intelligent and Flexible Indoor Environment Inspection

by Marius Cristian Luculescu, Luciana Cristea and Attila Laszlo Boer

Technologies 2025, 13(4), 161; https://doi.org/10.3390/technologies13040161 - 16 Apr 2025

Viewed by 221

Abstract

The widespread availability of artificial intelligence (AI) tools has facilitated the development of complex, high-performance applications across a broad range of domains. Among these, processes related to the surveillance and assisted verification of indoor environments have garnered significant interest. This paper presents the [...] Read more.

The widespread availability of artificial intelligence (AI) tools has facilitated the development of complex, high-performance applications across a broad range of domains. Among these, processes related to the surveillance and assisted verification of indoor environments have garnered significant interest. This paper presents the implementation, testing, and validation of an autonomous mobile platform designed for the intelligent and flexible inspection of such spaces. The artificial vision system, the main component on which the study focuses, was built using a Raspberry Pi 5 development module supplemented by a Raspberry Pi AI Kit to enable hardware acceleration for image recognition tasks using AI techniques. Some of the most recognized neural network models were evaluated in line with the application’s specific requirements. Utilizing transfer learning techniques, these models were further developed and trained with additional image datasets tailored to the inspection tasks. The performance of these networks was then tested and validated on new images, facilitating the selection of the model with the best results. The platform’s flexibility was ensured by mounting the artificial vision system on a mobile structure capable of autonomously navigating indoor environments and identifying inspection points, markers, and required objects. The platform could generate reports, make decisions based on the detected conditions, and be easily reconfigured for alternative inspection tasks. Finally, the intelligent and flexible inspection system was tested and validated for its deep learning-based vision system performance. Full article

(This article belongs to the Special Issue Advanced Autonomous Systems and Artificial Intelligence Stage)

► Show Figures

Graphical abstract

31 pages, 9296 KiB

Open AccessArticle

An Experimental and Numerical Analysis of the Influence of Surface Roughness on Supersonic Flow in a Nozzle Under Atmospheric and Low-Pressure Conditions

by Pavla Šabacká, Jiří Maxa, Robert Bayer, Tomáš Binar, Petr Bača, Jana Švecová, Jaroslav Talár and Martin Vlkovský

Technologies 2025, 13(4), 160; https://doi.org/10.3390/technologies13040160 - 16 Apr 2025

Viewed by 237

Abstract

The ongoing research in Environmental Scanning Electron Microscopy (ESEM) is contributed to in this paper. Specifically, this study investigates supersonic flow in a nozzle aperture under low-pressure conditions at the continuum mechanics boundary. This phenomenon is prevalent in the differentially pumped chamber of [...] Read more.

The ongoing research in Environmental Scanning Electron Microscopy (ESEM) is contributed to in this paper. Specifically, this study investigates supersonic flow in a nozzle aperture under low-pressure conditions at the continuum mechanics boundary. This phenomenon is prevalent in the differentially pumped chamber of an ESEM, which separates two regions with a significant pressure gradient using an aperture with a pressure ratio of approximately 10:1 in the range of 10,000 to 100 Pa. The influence of nozzle wall roughness on the boundary layer characteristics and its subsequent impact on the oblique shock wave behavior, and consequently, on the static pressure distribution along the flow axis, is solved in this paper. It demonstrates the significant effect of varying inertial-to-viscous force ratios at low pressures on the resulting impact of roughness on the oblique shock wave characteristics. The resulting oblique shock wave distribution significantly affects the static pressure profile along the axis, which can substantially influence the scattering and loss of the primary electron beam traversing the differential pumping stage. This, in turn, affects the sharpness of the resulting image. The boundary layer within the nozzle plays a crucial role in determining the overall flow characteristics and indirectly affects beam scattering. This study examines the influence of surface roughness and quality of the manufactured nozzle on the resulting flow behavior. The initial results obtained from experimental measurements using pressure sensors, when compared to CFD simulation results, demonstrate the necessity of accurately setting roughness values in CFD calculations to ensure accurate results. The CFD simulation has been validated against experimental data, enabling further simulations. The research combines physical theory, CFD simulations, advanced experimental sensing techniques, and precision manufacturing technologies for the critical components of the experimental setup. Full article

(This article belongs to the Special Issue New Technologies for Sensors)

► Show Figures

Figure 1

28 pages, 5163 KiB

Open AccessArticle

Design of High-Pass and Low-Pass Active Inverse Filters to Compensate for Distortions in RC-Filtered Electrocardiograms

by Dobromir Dobrev, Tatyana Neycheva, Vessela Krasteva and Irena Jekova

Technologies 2025, 13(4), 159; https://doi.org/10.3390/technologies13040159 - 15 Apr 2025

Viewed by 308

Abstract

Distortions of electrocardiograms (ECGs) caused by mandatory high-pass and low-pass analog RC filters in ECG devices are always present. The fidelity of the ECG waveform requires limiting the RC cutoff frequencies of the diagnostic (0.05–150 Hz) and monitoring systems (0.5–40 Hz). However, the [...] Read more.

Distortions of electrocardiograms (ECGs) caused by mandatory high-pass and low-pass analog RC filters in ECG devices are always present. The fidelity of the ECG waveform requires limiting the RC cutoff frequencies of the diagnostic (0.05–150 Hz) and monitoring systems (0.5–40 Hz). However, the use of fixed frequency bands is a compromise between enhanced noise immunity and ECG distortions. This study aims to propose active inverse high-pass and low-pass filters which are able to compensate for distortions in digital recordings of RC-filtered ECGs, thereby overcoming the limitations imposed by analog filtering. A new straightforward design of an inverse high-pass filter (IHPF) uses an integrator as the forward-path gain block, with a feedback loop containing an active digital filter equivalent to the analog RC high-pass filter. In contrast, the inverse low-pass filter (ILPF) employs a constant-gain block in the forward path to ensure stability and prevent phase delay, while its feedback path features an active digital counterpart of the RC low-pass filter. Second-order inverse filters are created by cascading two first-order stages. The proposed filters were validated according to essential performance requirements for electrocardiographs. The low-frequency (impulse) responses of IHPFs with cutoff frequencies of 0.05–5 Hz exhibit no overshoot and undershoot by magnitudes of 0.1–25 µV, well within the ±100 µV compliance limit defined for a test rectangular pulse (3 mV, 100 ms). The high-frequency responses of ILPFs with cutoff frequencies of 10–150 Hz present a relative amplitude drop of only 0.2–2.5%, far below the 10% limit for peak amplitude reduction of a triangular pulse (1.5 mV) with 20 ms vs. 200 ms widths. For any of the eight ECG leads (I, II, and V1–V6) available in the standard signal (ANE20000), the IHPF (0.05–5 Hz) presents ST-segment deviations <5 μV (within the ±25 μV limit) and R- and S-peak deviations <±3.5% (within the ±5% limit). The ILPF (10–150 Hz) preserves R- and S-peak amplitudes with deviations less than −1%. Diagnostic-level recovery of ECG waveforms distorted by first- and second-order analog RC filters in ECG devices is possible with the innovative and comprehensive inverse filter design presented in this study. This approach offers a significant advancement in ECG signal processing, effectively restoring essential waveform components even after aggressive, noise-robust analog filtering in ECG acquisition circuits. Although validated for ECG signals, the proposed inverse filters are also applicable to other biosignal front-end circuits employing RC coupling. Full article

(This article belongs to the Special Issue Digital Data Processing Technologies: Trends and Innovations)

► Show Figures

Figure 1

22 pages, 26135 KiB

Open AccessArticle

New Approach for Mapping Land Cover from Archive Grayscale Satellite Imagery

by Mohamed Rabii Simou, Mohamed Maanan, Safia Loulad, Mehdi Maanan and Hassan Rhinane

Technologies 2025, 13(4), 158; https://doi.org/10.3390/technologies13040158 - 14 Apr 2025

Viewed by 314

Abstract

This paper examines the use of image-to-image translation models to colorize grayscale satellite images for improved built-up segmentation of Agadir, Morocco, in 1967 and Les Sables-d’Olonne, France, in 1975. The proposed method applies advanced colorization techniques to historical remote sensing data, enhancing the [...] Read more.

This paper examines the use of image-to-image translation models to colorize grayscale satellite images for improved built-up segmentation of Agadir, Morocco, in 1967 and Les Sables-d’Olonne, France, in 1975. The proposed method applies advanced colorization techniques to historical remote sensing data, enhancing the segmentation process compared to using the original grayscale images. In this study, spatial data such as Landsat 5TM satellite images and declassified satellite images were collected and prepared for analysis. The models were trained and validated using Landsat 5TM RGB images and their corresponding grayscale versions. Once trained, these models were applied to colorize the declassified grayscale satellite images. To train the segmentation models, colorized Landsat images were paired with built-up-area masks, allowing the models to learn the relationship between colorized features and built-up regions. The best-performing segmentation model was then used to segment the colorized declassified images into built-up areas. The results demonstrate that the Attention Pix2Pix model successfully learned to colorize grayscale satellite images accurately, improving the PSNR by up to 27.72 and SSIM by 0.96. Furthermore, the results of segmentation were highly satisfactory, with UNet++ identified as the best-performing model with an mIoU of 96.95% in Greater Agadir and 95.42% in Vendée. These findings indicate that the application of the developed method can achieve accurate and reliable results that can be utilized for future LULC change studies. The innovative approach of the study has significant implications for land planning and management, providing accurate LULC information to inform decisions related to zoning, environmental protection, and disaster management. Full article

(This article belongs to the Section Environmental Technology)

► Show Figures

Graphical abstract

21 pages, 10169 KiB

Open AccessArticle

Reinforcement Learning for Fail-Operational Systems with Disentangled Dual-Skill Variables

by Taewoo Kim and Shiho Kim

Technologies 2025, 13(4), 156; https://doi.org/10.3390/technologies13040156 - 13 Apr 2025

Viewed by 247

Abstract

We present a novel approach to reinforcement learning (RL) specifically designed for fail-operational systems in critical safety applications. Our technique incorporates disentangled skill variables, significantly enhancing the resilience of conventional RL frameworks against mechanical failures and unforeseen environmental changes. This innovation arises from [...] Read more.

We present a novel approach to reinforcement learning (RL) specifically designed for fail-operational systems in critical safety applications. Our technique incorporates disentangled skill variables, significantly enhancing the resilience of conventional RL frameworks against mechanical failures and unforeseen environmental changes. This innovation arises from the imperative need for RL mechanisms to sustain uninterrupted and dependable operations, even in the face of abrupt malfunctions. Our research highlights the system’s ability to swiftly adjust and reformulate its strategy in response to sudden disruptions, maintaining operational integrity and ensuring the completion of tasks without compromising safety. The system’s capacity for immediate, secure reactions is vital, especially in scenarios where halting operations could escalate risks. We examine the system’s adaptability in various mechanical failure scenarios, highlighting its effectiveness in maintaining safety and functionality in unpredictable situations. Our research represents a significant advancement in the safety and performance of RL systems, paving the way for their deployment in safety-critical environments. Full article

(This article belongs to the Section Assistive Technologies)

► Show Figures

Figure 1

20 pages, 6224 KiB

Open AccessArticle

Exploring Heart Rate Variability and Mental Effects of Gameplay in Virtual Reality and 3D Morphing Animation

by Penio Lebamovski

Technologies 2025, 13(4), 157; https://doi.org/10.3390/technologies13040157 - 12 Apr 2025

Viewed by 254

Abstract

This study presents a research approach to creating 3D animations in a virtual reality game using the morphing technique as well as heart rate variability (HRV) analysis. The aim is to investigate the mental effects of the game on players by analysing electrocardiographic [...] Read more.

This study presents a research approach to creating 3D animations in a virtual reality game using the morphing technique as well as heart rate variability (HRV) analysis. The aim is to investigate the mental effects of the game on players by analysing electrocardiographic signals recorded before and during the game. The animations were created using Java(ver.1.8)/Java3D(ver.1.6), Blender(ver.3.1.2), and Unity(ver. 2021.3.6f1). The techniques used are Morph Interpolator in Java3D, as well as Blend Shapes and Keyframes in Blender. Animation in Unity does not have direct support for morphing, which necessitates the use of Blend Shapes and Blender. Formats such as OBJ and FBX were used to transfer data between the platforms. In addition to the software implementation of the game animation, the study offers a comparative analysis between two of the platforms (Java/Java3D and Blender) in terms of their effectiveness, advantages, and disadvantages in implementing morphing animations. The software solutions used create high-quality animations, which are necessary for generating an interactive virtual environment leading to mental stress during the game. The effectiveness of the proposed approach is proven through HRV analysis, with the results showing the psychological effect of the game on users, expressed in a decrease in HRV. Full article

(This article belongs to the Section Information and Communication Technologies)

► Show Figures

Figure 1

21 pages, 2693 KiB

Open AccessArticle

Implementation of Kolmogorov–Arnold Networks for Efficient Image Processing in Resource-Constrained Internet of Things Devices

by Anargul Shaushenova, Oleksandr Kuznetsov, Ardak Nurpeisova and Maral Ongarbayeva

Technologies 2025, 13(4), 155; https://doi.org/10.3390/technologies13040155 - 12 Apr 2025

Viewed by 269

Abstract

This research investigates the implementation of Kolmogorov–Arnold networks (KANs) for image processing in resource-constrained IoTs devices. KANs represent a novel neural network architecture that offers significant advantages over traditional deep learning approaches, particularly in applications where computational resources are limited. Our study demonstrates [...] Read more.

This research investigates the implementation of Kolmogorov–Arnold networks (KANs) for image processing in resource-constrained IoTs devices. KANs represent a novel neural network architecture that offers significant advantages over traditional deep learning approaches, particularly in applications where computational resources are limited. Our study demonstrates the efficiency of KAN-based solutions for image analysis tasks in IoTs environments, providing comparative performance metrics against conventional convolutional neural networks. The experimental results indicate substantial improvements in processing speed and memory utilization while maintaining competitive accuracy. This work contributes to the advancement of AI-driven IoTs applications by proposing optimized KAN-based implementations suitable for edge computing scenarios. The findings have important implications for IoTs deployment in smart infrastructure, environmental monitoring, and industrial automation where efficient image processing is critical. Full article

(This article belongs to the Topic Internet of Things Architectures, Applications, and Strategies: Emerging Paradigms, Technologies, and Advancing AI Integration)

► Show Figures

Figure 1

23 pages, 57584 KiB

Open AccessArticle

Pix2Next: Leveraging Vision Foundation Models for RGB to NIR Image Translation

by Youngwan Jin, Incheol Park, Hanbin Song, Hyeongjin Ju, Yagiz Nalcakan and Shiho Kim

Technologies 2025, 13(4), 154; https://doi.org/10.3390/technologies13040154 - 11 Apr 2025

Viewed by 229

Abstract

This paper proposes Pix2Next, a novel image-to-image translation framework designed to address the challenge of generating high-quality Near-Infrared (NIR) images from RGB inputs. Our method leverages a state-of-the-art Vision Foundation Model (VFM) within an encoder–decoder architecture, incorporating cross-attention mechanisms to enhance feature integration. [...] Read more.

This paper proposes Pix2Next, a novel image-to-image translation framework designed to address the challenge of generating high-quality Near-Infrared (NIR) images from RGB inputs. Our method leverages a state-of-the-art Vision Foundation Model (VFM) within an encoder–decoder architecture, incorporating cross-attention mechanisms to enhance feature integration. This design captures detailed global representations and preserves essential spectral characteristics, treating RGB-to-NIR translation as more than a simple domain transfer problem. A multi-scale PatchGAN discriminator ensures realistic image generation at various detail levels, while carefully designed loss functions couple global context understanding with local feature preservation. We performed experiments on the RANUS and IDD-AW datasets to demonstrate Pix2Next’s advantages in quantitative metrics and visual quality, highly improving the FID score compared to existing methods. Furthermore, we demonstrate the practical utility of Pix2Next by showing improved performance on a downstream object detection task using generated NIR data to augment limited real NIR datasets. The proposed method enables the scaling up of NIR datasets without additional data acquisition or annotation efforts, potentially accelerating advancements in NIR-based computer vision applications. Full article

► Show Figures

Graphical abstract

25 pages, 20259 KiB

Open AccessArticle

From Antenna Optimization to MIMO Structures: A Unified Design Framework

by Claudia Constantinescu, Claudia Pacurar, Adina Giurgiuman, Calin Munteanu, Sergiu Andreica, Marian Gliga and Laura Grindei

Technologies 2025, 13(4), 153; https://doi.org/10.3390/technologies13040153 - 10 Apr 2025

Viewed by 188

Abstract

Considering the improvements to modern communications and its requirements, antennas need to operate optimally at certain frequencies and have smaller dimensions. This study considered the optimization of two single antennas functioning at 2.4 GHz through geometry modification while preserving or even improving their [...] Read more.

Considering the improvements to modern communications and its requirements, antennas need to operate optimally at certain frequencies and have smaller dimensions. This study considered the optimization of two single antennas functioning at 2.4 GHz through geometry modification while preserving or even improving their bandwidth, while also considering their gain. At first, the research was conducted using numerical modeling and, based on the conclusions drawn following this analysis, the next step was the experimental analysis of the structures. Due to their different geometrical appearances, the optimized antennas were compared, and then an optimum two-antenna MIMO structure was determined for both structures using different methods to decrease the mutual coupling. The optimum structure was obtained for both antennas. The new antennas functioned at 2.4 GHz but had different dimensions, thus a study into the decoupling methods was needed to see if the same methods were best for both cases. It was determined that shifting the two antennas in the MIMO was the better method when leaving a distance of λ/2 cannot be considered due to an increase in the dimensions of the structures, followed by a 90° shifting of the antennas. Also, the modification of the gain representation was observed through implementing the different decoupling methods to determine their influence on the beamforming. Full article

► Show Figures

Figure 1

31 pages, 2939 KiB

Open AccessArticle

Analysis of Multimodal Sensor Systems for Identifying Basic Walking Activities

by John C. Mitchell, Abbas A. Dehghani-Sanij, Sheng Q. Xie and Rory J. O’Connor

Technologies 2025, 13(4), 152; https://doi.org/10.3390/technologies13040152 - 10 Apr 2025

Viewed by 292

Abstract

Falls are a major health issue in societies globally and the second leading cause of unintentional death worldwide. To address this issue, many studies aim to remotely monitor gait to prevent falls. However, these activity data collected in studies must be labelled with [...] Read more.

Falls are a major health issue in societies globally and the second leading cause of unintentional death worldwide. To address this issue, many studies aim to remotely monitor gait to prevent falls. However, these activity data collected in studies must be labelled with the appropriate environmental context through Human Activity Recognition (HAR). Multimodal HAR datasets often achieve high accuracies at the cost of cumbersome sensor systems, creating a need for these datasets to be analysed to identify the sensor types and locations that enable high-accuracy HAR. This paper analyses four datasets, USC-HAD, HuGaDB, Camargo et al.’s dataset, and CSL-SHARE, to find optimal models, methods, and sensors across multiple datasets. Regarding window size, optimal windows are found to be dependent on the sensor modality of a dataset but mostly occur in the 2–5 s range. Support Vector Machines (SVMs) and Artificial Neural Networks (ANNs) are found to be the highest-performing models overall. ANNs are further used to create models trained on the features from individual sensors of each dataset. From this analysis, Inertial Measurement Units (IMUs) and three-axis goniometers are shown to be individually capable of high classification accuracy, with Electromyography (EMG) sensors exhibiting inconsistent and reduced accuracies. Finally, it is shown that the thigh is the optimal location for IMU sensors, with accuracy decreasing as IMUs are placed further down away from the thigh. Full article

(This article belongs to the Special Issue Assistive Technologies in Care and Rehabilitation: Research, Developments, and International Initiatives)

► Show Figures

Figure 1

30 pages, 1564 KiB

Open AccessArticle

RACER: A Lightweight Distributed Consensus Algorithm for the IoT with Peer-Assisted Latency-Aware Traffic Optimisation

by Zachary Auhl, Harsha Moraliyage, Naveen Chilamkurti and Damminda Alahakoon

Technologies 2025, 13(4), 151; https://doi.org/10.3390/technologies13040151 - 9 Apr 2025

Viewed by 286

Abstract

Internet-of-Things (IoT) devices are interconnected objects embedded with sensors and software, enabling data collection and exchange. These devices encompass a wide range of applications, from household appliances to industrial systems, designed to enhance connectivity and automation. In distributed IoT networks, achieving reliable decision-making [...] Read more.

Internet-of-Things (IoT) devices are interconnected objects embedded with sensors and software, enabling data collection and exchange. These devices encompass a wide range of applications, from household appliances to industrial systems, designed to enhance connectivity and automation. In distributed IoT networks, achieving reliable decision-making necessitates robust consensus mechanisms that allow devices to agree on a shared state of truth without reliance on central authorities. Such mechanisms are critical for ensuring system resilience under diverse operational conditions. Recent research has identified three common limitations in existing consensus mechanisms for IoT environments: dependence on synchronised networks and clocks, reliance on centralised coordinators, and suboptimal performance. To address these challenges, this paper introduces a novel consensus mechanism called Randomised Asynchronous Consensus with Efficient Real-time Sampling (RACER). The RACER framework eliminates the need for synchronised networks and clocks by implementing the Sequenced Probabilistic Double Echo (SPDE) algorithm, which operates asynchronously without timing assumptions. Furthermore, to mitigate the reliance on centralised coordinators, RACER leverages the SPDE gossip protocol, which inherently requires no leaders, combined with a lightweight transaction ordering mechanism optimised for IoT sensor networks. Rather than using a blockchain for transaction ordering, we opted for an eventually consistent transaction ordering mechanism to specifically deal with high churn, asynchronous networks and to allow devices to independently and deterministically order transactions. To enhance the throughput of IoT networks, this paper also proposes a complementary algorithm, Peer-assisted Latency-Aware Traffic Optimisation (PLATO), designed to maximise efficiency within RACER-based systems. The combination of RACER and PLATO is able to maintain a throughput of above 600 mb/s on a 100-node network, significantly outperforming the compared consensus mechanisms in terms of network node size and performance. Full article

(This article belongs to the Special Issue IoT-Enabling Technologies and Applications)

► Show Figures

Figure 1

30 pages, 7670 KiB

Open AccessArticle

Comparative Analysis of Energy Consumption and Performance Metrics in Fuel Cell, Battery, and Hybrid Electric Vehicles Under Varying Wind and Road Conditions

by Ahmed Hebala, Mona I. Abdelkader and Rania A. Ibrahim

Technologies 2025, 13(4), 150; https://doi.org/10.3390/technologies13040150 - 9 Apr 2025

Viewed by 632

Abstract

As global initiatives to reduce greenhouse gas emissions and combat climate change expand, electric vehicles (EVs) powered by fuel cells and lithium-ion batteries are gaining global recognition as solutions for sustainable transportation due to their high energy conversion efficiency. Considering the driving range [...] Read more.

As global initiatives to reduce greenhouse gas emissions and combat climate change expand, electric vehicles (EVs) powered by fuel cells and lithium-ion batteries are gaining global recognition as solutions for sustainable transportation due to their high energy conversion efficiency. Considering the driving range limitations of battery electric vehicles (BEVs) and the low efficiency of internal combustion engines (ICEs), fuel cell hybrid vehicles offer a compelling alternative for long-distance, low-emission driving with less refuelling time. To facilitate their wider scale adoption, it is essential to understand their energy performance through models that consider external weather effects, driving styles, road gradients, and their simultaneous interaction. This paper presents a microlevel, multicriteria assessment framework to investigate the performance of BEVs, fuel cell electric vehicles (FCEVs), and hybrid electric vehicles (HEVs), with a focus on energy consumption, drive systems, and emissions. Simulation models were developed using MATLAB 2021a Simulink environment, thus enabling the integration of standardized driving cycles with real-world wind and terrain variations. The results are presented for various trip scenarios, employing quantitative and qualitative analysis methods to identify the most efficient vehicle configuration, also validated through the simulation of three commercial EVs. Predictive modelling approaches are utilized to estimate a vehicle’s performance under unexplored conditions. Results indicate that trip conditions have a significant impact on the performance of all three vehicles, with HEVs emerging as the most efficient and balanced option, followed by FCEVs, making them strong candidates compared with BEVs for broader adoption in the transition toward sustainable transportation. Full article

(This article belongs to the Special Issue Next-Generation Distribution System Planning, Operation, and Control)

► Show Figures

Figure 1

57 pages, 13801 KiB

Open AccessArticle

Integrating Social Sustainability into Supply Chain Design: Optimization of a Capacitated Two-Echelon Location-Routing Problem

by Mohamed Nafea, Lamia A. Shihata and Maggie Mashaly

Technologies 2025, 13(4), 149; https://doi.org/10.3390/technologies13040149 - 9 Apr 2025

Viewed by 287

Abstract

Over recent years, location-routing problems have become popular, since they tackle multiple major decisions in supply chains. With the focus now on sustainable supply chains, the problem has become sought-after, with the emphasis on complying with goals set by world leaders such as [...] Read more.

Over recent years, location-routing problems have become popular, since they tackle multiple major decisions in supply chains. With the focus now on sustainable supply chains, the problem has become sought-after, with the emphasis on complying with goals set by world leaders such as complying with environmental rules and social equity. As a result, this has opened multiple research directions within the location-routing problem. In the literature, the focus has merely been on two of the sustainability pillars, the environmental and economic pillars, with no integration of the social pillar. In this article, the aim is to integrate the social pillar alongside the environmental and economic ones by modeling the system as a capacitated two-echelon location-routing problem tackling multiple scenarios. Under the umbrella of optimization technology, the algorithm used to solve the problem is a genetic algorithm. This article also demonstrates the process of designing the experimentation phase and selecting variables aiming to fine-tune the models. Multiple parent selection methods and crossover methods were tested, among other variables. The algorithm has proven its success in finding a near-optimum value when compared to the benchmark solution, with an error less than 0.05%. Tournament has performed better as a parent selection method in contrast to stochastic universal sampling, and has proved to be more stable in the face of the stochastic noise induced in the models. This study shows that the social pillar, like the other two pillars, can be integrated in the location-routing problem at an extensive level, beyond what is normally implemented. Full article

► Show Figures

Figure 1

34 pages, 24917 KiB

Open AccessArticle

Autonomous Real-Time Mass Center Location and Inertia Identification for Grappling Space Robotics

by Timothy Sands

Technologies 2025, 13(4), 148; https://doi.org/10.3390/technologies13040148 - 8 Apr 2025

Viewed by 325

Abstract

Grappling actions by space robots for the purposes of stabilizing, refueling, repair, and equipment replacement necessitate the autonomous abilities of a single grappling space robot to rapidly contend with large variations in total system inertia rapidly shifting a system’s center of mass, as [...] Read more.

Grappling actions by space robots for the purposes of stabilizing, refueling, repair, and equipment replacement necessitate the autonomous abilities of a single grappling space robot to rapidly contend with large variations in total system inertia rapidly shifting a system’s center of mass, as targets can be massive with possibly unknown or poorly known mass inertia properties. Grappling actions yield opportunities for a novel online calculation of the time-varying location of the combined system’s center of mass. Two-norm optimal nonlinear, projection regression-based learning is implemented and juxtaposed to a comparative benchmark both qualitatively and quantitatively supported by a comparison of enhancements of Luenberger observers. Analysis precedes modeling and simulation to verify the design, and then, spaceflight experiments are proposed for the sequel to validate the simulation results. Time-varying mass locations are discerned, and the time-varying location of the mass center is revealed to be 36–95 percent different than initially assumed, and 58–317 percent corrections to inertia identification are demonstrated. Combined three-dimensional maneuvers obscures identification compared to single-axis maneuvering. Full article

(This article belongs to the Special Issue Advanced Autonomous Systems and Artificial Intelligence Stage)

► Show Figures

Figure 1

22 pages, 7716 KiB

Open AccessArticle

A Deep-Learning Approach to Heart Sound Classification Based on Combined Time-Frequency Representations

by Leonel Orozco-Reyes, Miguel A. Alonso-Arévalo, Eloísa García-Canseco, Roilhi F. Ibarra-Hernández and Roberto Conte-Galván

Technologies 2025, 13(4), 147; https://doi.org/10.3390/technologies13040147 - 7 Apr 2025

Viewed by 352

Abstract

Worldwide, heart disease is the leading cause of mortality. Cardiac auscultation, when conducted by a trained professional, is a non-invasive, cost-effective, and readily available method for the initial assessment of cardiac health. Automated heart sound analysis offers a promising and accessible approach to [...] Read more.

Worldwide, heart disease is the leading cause of mortality. Cardiac auscultation, when conducted by a trained professional, is a non-invasive, cost-effective, and readily available method for the initial assessment of cardiac health. Automated heart sound analysis offers a promising and accessible approach to supporting cardiac diagnosis. This work introduces a novel method for classifying heart sounds as normal or abnormal by leveraging time-frequency representations. Our approach combines three distinct time-frequency representations—short-time Fourier transform (STFT), mel-scale spectrogram, and wavelet synchrosqueezed transform (WSST)—to create images that enhance classification performance. These images are used to train five convolutional neural networks (CNNs): AlexNet, VGG-16, ResNet50, a CNN specialized in STFT images, and our proposed CNN model. The method was trained and tested using three public heart sound datasets: PhysioNet/CinC Challenge 2016, CirCor DigiScope Phonocardiogram Dataset 2022, and another open database. While individual representations achieve maximum accuracy of ≈85.9%, combining STFT, mel, and WSST boosts accuracy to ≈99%. By integrating complementary time-frequency features, our approach demonstrates robust heart sound analysis, achieving consistent classification performance across diverse CNN architectures, thus ensuring reliability and generalizability. Full article

(This article belongs to the Special Issue Advancements in Medical and Assistive Technologies Using Artificial Intelligence and Deep Learning Techniques)

► Show Figures

Figure 1

23 pages, 1931 KiB

Open AccessArticle

A Study on Chatbot Development Using No-Code Platforms by People with Disabilities for Their Peers at a Sheltered Workshop

by Sara Hamideh Kerdar, Britta Marleen Kirchhoff, Lars Adolph and Liane Bächler

Technologies 2025, 13(4), 146; https://doi.org/10.3390/technologies13040146 - 4 Apr 2025

Viewed by 418

Abstract

No-code (NC) platforms empower individuals without IT experience to create tailored applications and websites. While these platforms are accessible to a broader audience, their usability for people with disabilities remains underexplored. This study investigated whether, with targeted training, people with disabilities could effectively [...] Read more.

No-code (NC) platforms empower individuals without IT experience to create tailored applications and websites. While these platforms are accessible to a broader audience, their usability for people with disabilities remains underexplored. This study investigated whether, with targeted training, people with disabilities could effectively use NC platforms to develop customized tools for their workplace, and whether these tools would be adopted by their peers. Conducted in collaboration with a sheltered workshop in Germany, the study had three phases. Phase I involved a brainstorming session with employees, which shaped the study design and product development. In Phase II, six participants with disabilities received a one-week training to develop chatbots. Phase III implemented the chatbots in the workshop. In Phase II, each participant successfully developed four chatbots, which increased the participants’ skills and motivation. Based on the phase III results, users rated the developed chatbots highly (the System Usability Scale (SUS) questionnaire was delivered in the form of a chatbot), indicating their user-friendliness (M = 88.9, SD = 11.2). This study suggests that with appropriate training, individuals with disabilities can use NC platforms to create impactful, customized tools that are user-friendly and accessible to their peers. Full article

(This article belongs to the Special Issue Assistive Technologies in Care and Rehabilitation: Research, Developments, and International Initiatives)

► Show Figures

Figure 1

34 pages, 2629 KiB

Open AccessArticle

AOAFS: A Malware Detection System Using an Improved Arithmetic Optimization Algorithm

by Rafat Alshorman, Bilal H. Abed-alguni and Yaqeen E. Alqudah

Technologies 2025, 13(4), 145; https://doi.org/10.3390/technologies13040145 - 4 Apr 2025

Viewed by 235

Abstract

Malware detection datasets often contain a huge number of features, many of which are irrelevant, noisy, and duplicated. This issue diminishes the efficacy of Machine Learning models used for malware detection. Feature Selection (FS) is an approach commonly used to reduce the number [...] Read more.

Malware detection datasets often contain a huge number of features, many of which are irrelevant, noisy, and duplicated. This issue diminishes the efficacy of Machine Learning models used for malware detection. Feature Selection (FS) is an approach commonly used to reduce the number of features in a malware detection dataset to a smaller set of features to facilitate the ease of the Machine Learning process. The Arithmetic Optimization Algorithm (AOA) is a relatively new efficient optimization algorithm that can be used for FS. This work introduces a new malware detection system called the improved AOA method for FS (AOAFS) that enhances the performance of Machine Learning techniques for malware detection. The AOAFS contains three key enhancements. First, the K-means clustering method is used to improve the initial population of the AOAFS. Second, sixteen Binary Transfer Functions are applied to model the binary solution space for FS in the AOAFS. Finally, Dynamic Opposition-based Learning is utilized to improve the mutation capability of the AOAFS. Several malware datasets were used to compare the AOAFS to four popular Machine Learning algorithms and eight famous wrapper-based optimization algorithms. Statistically, the AOAFS was evaluated using the Friedman Test for ranking the tested algorithms, while the Wilcoxon Signed-Rank Test was employed for pairwise comparisons. The results indicated that the AOAFS achieves the highest classification accuracy with the smallest feature set across all datasets. Full article

► Show Figures

Figure 1

53 pages, 2538 KiB

Open AccessSystematic Review

Assistive and Emerging Technologies to Detect and Reduce Neurophysiological Stress and Anxiety in Children and Adolescents with Autism and Sensory Processing Disorders: A Systematic Review

by Pantelis Pergantis, Victoria Bamicha, Aikaterini Doulou, Antonios I. Christou, Nikolaos Bardis, Charalabos Skianis and Athanasios Drigas

Technologies 2025, 13(4), 144; https://doi.org/10.3390/technologies13040144 - 4 Apr 2025

Viewed by 1123

Abstract

This systematic review aims to investigate the ways in which assistive and developing technologies can help children and adolescents with autism spectrum disorder (ASD) experience less stress and neurophysiological distress. According to recent CDC data, the prevalence of ASD in the United States [...] Read more.

This systematic review aims to investigate the ways in which assistive and developing technologies can help children and adolescents with autism spectrum disorder (ASD) experience less stress and neurophysiological distress. According to recent CDC data, the prevalence of ASD in the United States has climbed to 1 in 36 children. The symptoms of ASD can manifest in a wide range of ways, and the illness itself exhibits significant variations. Furthermore, it has been closely linked to experiencing stress and worry in one’s life, which many people refer to as sensory processing disorder (SPD). SPD is a disorder that describes how people behave when they are exposed to environmental stimuli that they may not normally process by feeling more intense than what is causing them to worry and distress. One of the most significant limiting factors that can prevent someone from engaging in what they need to do in their everyday lives is stress. Individuals with ASD deal with stress on a regular basis, which has a big impact on how they function. In order to address a significant research vacuum concerning the use of assistive and emerging technologies to reduce stress in individuals with ASD, this systematic review aims to investigate performance, measuring techniques, and interventions by gathering data from the past 10 years. In order to determine the research hypothesis, particular research questions, and the inclusion and exclusion criteria for the studies, the research process entails gathering studies through systematic review analysis in accordance with the PRISMA principles. Experimental and observational studies on the use of assistive and emerging technologies for stress and anxiety management in children and adolescents with ASD that were published only in English met the inclusion criteria. Research not directly related to stress and anxiety outcomes, articles published in languages other than English, and research conducted outside of the designated time frame were also excluded. The study’s findings demonstrated that the technologies under examination had beneficial impacts on reducing stress; nonetheless, notable limitations were found that could compromise the replication and generalizability of legitimate and dependable applications in their utilization. Full article

(This article belongs to the Special Issue Assistive Technologies in Care and Rehabilitation: Research, Developments, and International Initiatives)

► Show Figures

Figure 1

25 pages, 13176 KiB

Open AccessArticle

Deep Object Occlusion Relationship Detection Based on Associative Embedding Clustering

by Peiyong Gong, Kai Zheng, Ting Liu, Yi Jiang and Huixuan Zhao

Technologies 2025, 13(4), 143; https://doi.org/10.3390/technologies13040143 - 4 Apr 2025

Viewed by 290

Abstract

Visual relationship detection is crucial for understanding scenes depicted in images when aiming to detect objects within the image and recognize the visual relationships between each pair of objects. Nevertheless, profound occlusion, as a typical visual relationship existing between objects and constituting a [...] Read more.

Visual relationship detection is crucial for understanding scenes depicted in images when aiming to detect objects within the image and recognize the visual relationships between each pair of objects. Nevertheless, profound occlusion, as a typical visual relationship existing between objects and constituting a pivotal semantic feature, has regrettably been subjected to insufficient scrutiny. To address this issue, we propose a pioneering approach termed DOORD-AEC, which is specifically designed for detecting occlusion spatial relationships among targets. DOORD-AEC introduces associative embedding clustering to supervise a convolutional neural network with two branches, enabling it to take in an input image and produce a triplet set representing occlusion spatial relationships. The network learns to simultaneously identify all of the targets and occlusions that make up the triplet set and group them together using associative embedding clustering. Additionally, we contribute the KORD dataset, which is a novel and challenging dataset for occlusion spatial relationships among targets. We demonstrate the effectiveness of our DOORD-AEC method using this dataset. Full article

► Show Figures

Figure 1

10 pages, 208 KiB

Open AccessArticle

Leveraging Artificial Intelligence for Personalized Rehabilitation Programs for Head and Neck Surgery Patients

by Gianluca Marcaccini, Ishith Seth, Jennifer Novo, Vicki McClure, Brett Sacks, Kaiyang Lim, Sally Kiu-Huen Ng, Roberto Cuomo and Warren M. Rozen

Technologies 2025, 13(4), 142; https://doi.org/10.3390/technologies13040142 - 4 Apr 2025

Viewed by 325

Abstract

Background: Artificial intelligence (AI) and large language models (LLMs) are increasingly used in healthcare, with applications in clinical decision-making and workflow optimization. In head and neck surgery, postoperative rehabilitation is a complex, multidisciplinary process requiring personalized care. This study evaluates the feasibility of [...] Read more.

Background: Artificial intelligence (AI) and large language models (LLMs) are increasingly used in healthcare, with applications in clinical decision-making and workflow optimization. In head and neck surgery, postoperative rehabilitation is a complex, multidisciplinary process requiring personalized care. This study evaluates the feasibility of using LLMs to generate tailored rehabilitation programs for patients undergoing major head and neck surgical procedures. Methods: Ten hypothetical head and neck surgical clinical scenarios were developed, representing oncologic resections with complex reconstructions. Four LLMs, ChatGPT-4o, DeepSeek V3, Gemini 2, and Copilot, were prompted with identical queries to generate rehabilitation plans. Three senior clinicians independently assessed their quality, accuracy, and clinical relevance using a five-point Likert scale. Readability and quality metrics, including the DISCERN score, Flesch Reading Ease, Flesch–Kincaid Grade Level, and Coleman–Liau Index, were applied. Results: ChatGPT-4o achieved the highest clinical relevance (Likert mean of 4.90 ± 0.32), followed by DeepSeek V3 (4.00 ± 0.82) and Gemini 2 (3.90 ± 0.74), while Copilot underperformed (2.70 ± 0.82). Gemini 2 produced the most readable content. A statistical analysis confirmed significant differences across the models (p < 0.001). Conclusions: LLMs can generate rehabilitation programs with varying quality and readability. ChatGPT-4o produced the most clinically relevant plans, while Gemini 2 generated more readable content. AI-generated rehabilitation plans may complement existing protocols, but further clinical validation is necessary to assess their impact on patient outcomes. Full article

(This article belongs to the Special Issue Assistive Technologies in Care and Rehabilitation: Research, Developments, and International Initiatives)

► Show Figures

Graphical abstract

18 pages, 3425 KiB

Open AccessArticle

Improvement of Bank Fraud Detection Through Synthetic Data Generation with Gaussian Noise

by Fray L. Becerra-Suarez, Halyn Alvarez-Vasquez and Manuel G. Forero

Technologies 2025, 13(4), 141; https://doi.org/10.3390/technologies13040141 - 4 Apr 2025

Viewed by 440

Abstract

Bank fraud detection faces critical challenges in imbalanced datasets, where fraudulent transactions are rare, severely impairing model generalization. This study proposes a Gaussian noise-based augmentation method to address class imbalance, contrasting it with SMOTE and ADASYN. By injecting controlled perturbations into the minority [...] Read more.

Bank fraud detection faces critical challenges in imbalanced datasets, where fraudulent transactions are rare, severely impairing model generalization. This study proposes a Gaussian noise-based augmentation method to address class imbalance, contrasting it with SMOTE and ADASYN. By injecting controlled perturbations into the minority class, our approach mitigates overfitting risks inherent in interpolation-based techniques. Five classifiers, including XGBoost and a convolutional neural network (CNN), were evaluated on augmented datasets. XGBoost achieved superior performance with Gaussian noise-augmented data (accuracy: 0.999507, AUC: 0.999506), outperforming SMOTE and ADASYN. These results underscore Gaussian noise’s efficacy in enhancing fraud detection accuracy, offering a robust alternative to conventional oversampling methods. Our findings emphasize the pivotal role of augmentation strategies in optimizing classifier performance for imbalanced financial data. Full article

► Show Figures

Figure 1

28 pages, 3613 KiB

Open AccessArticle

Chatbot Based on Large Language Model to Improve Adherence to Exercise-Based Treatment in People with Knee Osteoarthritis: System Development

by Humberto Farías, Joaquín González Aroca and Daniel Ortiz

Technologies 2025, 13(4), 140; https://doi.org/10.3390/technologies13040140 - 4 Apr 2025

Viewed by 432

Abstract

Knee osteoarthritis (KOA) is a prevalent condition globally, leading to significant pain and disability, particularly in individuals over the age of 40. While exercise has been shown to reduce symptoms and improve physical function and quality of life in patients with KOA, long-term [...] Read more.

Knee osteoarthritis (KOA) is a prevalent condition globally, leading to significant pain and disability, particularly in individuals over the age of 40. While exercise has been shown to reduce symptoms and improve physical function and quality of life in patients with KOA, long-term adherence to exercise programs remains a challenge due to the lack of ongoing support. To address this, a chatbot was developed using large language models (LLMs) to provide evidence-based guidance and promote adherence to treatment. A systematic review conducted under the PRISMA framework identified relevant clinical guidelines that served as the foundational knowledge base for the chatbot. The Mistral 7B model, optimized with Parameter-Efficient Fine-Tuning (PEFT) and Mixture-of-Experts (MoE) techniques, was integrated to ensure computational efficiency and mitigate hallucinations, a critical concern in medical applications. Additionally, the chatbot employs Self-Reflective Retrieval-Augmented Generation (SELF-RAG) combined with Chain of Thought (CoT) reasoning, enabling dynamic query reformulation and the generation of accurate, evidence-based responses tailored to patient needs. The chatbot was evaluated by comparing pre- and post-improvement versions and against a reference model (ChatGPT), using metrics of accuracy, relevance, and consistency. The results demonstrated significant improvements in response quality and conversational coherence, emphasizing the potential of integrating advanced LLMs with retrieval and reasoning methods to address critical challenges in healthcare. This approach not only enhances treatment adherence but also strengthens patient–provider interactions in managing chronic conditions like KOA. Full article

(This article belongs to the Special Issue Advancements in Medical and Assistive Technologies Using Artificial Intelligence and Deep Learning Techniques)

► Show Figures

Figure 1

22 pages, 2418 KiB

Open AccessArticle

An Implementation of Creep Test Assisting System with Dial Gauge Needle Reading and Smart Lighting Function for Laboratory Automation

by Dezheng Kong, Nobuo Funabiki, Shihao Fang, Noprianto, Mitsuhiro Okayasu and Pradini Puspitaningayu

Technologies 2025, 13(4), 139; https://doi.org/10.3390/technologies13040139 - 2 Apr 2025

Viewed by 184

Abstract

For decades, analog dial gauges have been essential for measuring and monitoring data at various industrial instruments including production machines and laboratory equipment. Among them, we focus on the instrument for creep test in a mechanical engineering laboratory, which evaluates material strength under [...] Read more.

For decades, analog dial gauges have been essential for measuring and monitoring data at various industrial instruments including production machines and laboratory equipment. Among them, we focus on the instrument for creep test in a mechanical engineering laboratory, which evaluates material strength under sustained stress. Manual reading of gauges imposes significant labor demands, especially in long-duration tests. This burden further increases under low-lighting environments, where poor visibility can lead to misreading data points, potentially compromising the accuracy of test results. In this paper, to address the challenges, we implement a creep test assisting system that possesses the following features: (1) to save the installation cost, a web camera and Raspberry Pi are employed to capture images of the dial gauge and automate the needle reading by image processing in real time, (2) to ensure reliability under low-lighting environments, a smart lighting mechanism is integrated to turn on a supplementary light when the dial gauge is not clearly visible, and (3) to allow a user to stay in a distant place from the instrument during a creep test, material break is detected and the corresponding message is notified to a laboratory staff using LINE automatically. For evaluations, we install the implemented system into a material strength measuring instrument at Okayama University, Japan, and confirm the effectiveness and accuracy through conducting experiments under various lighting conditions. Full article

► Show Figures

Figure 1

18 pages, 1707 KiB

Open AccessArticle

Resonance-Induced Capacitively Coupled Contactless Conductivity Detection (ReC⁴D) Unit for Nucleic Acid Amplification Testing

by Roberto G. Ramírez-Chavarría, Jorge A. Uc-Martín, Bryan E. Alvarez-Serna and Ramón F. Padilla-Morán

Technologies 2025, 13(4), 138; https://doi.org/10.3390/technologies13040138 - 2 Apr 2025

Viewed by 278

Abstract

Nucleic acid amplification (NAA) is a technique that increases the number of copies of a gene, making it possible to detect microorganisms. This technique is often used in clinical tests, biochemical analysis, and environmental assays, to mention only a few. However, developing portable, [...] Read more.

Nucleic acid amplification (NAA) is a technique that increases the number of copies of a gene, making it possible to detect microorganisms. This technique is often used in clinical tests, biochemical analysis, and environmental assays, to mention only a few. However, developing portable, robust, and low-cost measurement platforms to evaluate NAA products remains a technological challenge. Therefore, in this work, we introduce an attractive unit for detecting and quantifying nucleic acids based on the capacitively coupled contactless conductivity detection (

C^{4}

D) principle. The proposed unit, Re

C^{4}

D, combines electrical resonance with

C^{4}

D to enhance sensitivity when evaluating an NAA reaction. The Re

C^{4}

D units advantages are twofold: (i) the transducer is electrically isolated to allow its reuse, and (ii) the induced electrical resonance in the Re

C^{4}

D unit minimizes the stray capacitances of the conventional

C^{4}

D assays, which enhances sensitivity, increases the linear operating range, and improves the limit of detection (LoD). Furthermore, we evaluated the proposed device for quantifying different concentrations of SARS-CoV-2 genetic material and compared it with measurements from a conventional

C^{4}

D unit. Thus, we demonstrate that the Re

C^{4}

D unit can measure concentrations of NAA products with an LoD of 0.24

\frac{copy}{μ L}

and a sensitivity of 5.618

\frac{kHz}{\log (\frac{copy}{μ L})}

. These results position the Re

C^{4}

D unit close to the state-of-the-art NAA testing platforms, with the added value of a low cost, robustness, reusability, and affordability. Full article

(This article belongs to the Special Issue New Technologies for Sensors)

► Show Figures

Graphical abstract

17 pages, 14217 KiB

Open AccessArticle

DeepSTAS: DL-assisted Semantic Transmission Accuracy Enhancement Through an Attention-driven HAPS Relay System

by Pascal Nkurunziza and Daisuke Umehara

Technologies 2025, 13(4), 137; https://doi.org/10.3390/technologies13040137 - 2 Apr 2025

Viewed by 225

Abstract

Semantic communication technology, as it allows for source data meaning extraction and the transmission of appropriate semantic information only, has the potential to extend Shannon’s paradigm, which is concerned with the reproduction of a message from one location to another, regardless of its [...] Read more.

Semantic communication technology, as it allows for source data meaning extraction and the transmission of appropriate semantic information only, has the potential to extend Shannon’s paradigm, which is concerned with the reproduction of a message from one location to another, regardless of its meaning. Nevertheless, some user terminals (UTs) may experience inadequate service due to their geolocation in reference to the base stations, which may entirely affect the accuracy of transmission and complicate deployment and implementation. A High-Altitude Platform Station (HAPS) serves as a key enabler for the deployment of wireless broadband in inaccessible areas, such as in coastal, desert, and mountainous areas. This paper proposes a novel HAPS relay-based semantic communication scheme, named DeepSTAS, which leverages deep learning techniques to enhance transmission accuracy. The proposed scheme focuses on attention-based semantic signal decoding, denoising, and forwarding modes; thus, called a CSA-DCGAN SDF HAPS relay network. The simulation results reveal that the proposed system with attention mechanisms significantly outperforms the system without attention mechanisms, both in peak signal-to-noise ratio (PSNR) and multi-scale structural similarity index (MS-SSIM); the proposed system can achieve a 2 dB gain when leveraging the attention mechanisms, and a PSNR of 38.5 dB can be obtained, with an MS-SSIM exceeding 0.999 at an approximate SNR of only 20 dB. The system provides considerable performance, more than 37 dB, and a corresponding MS-SSIM close to 0.999 at an estimated SNR of 20 dB when the CIFAR-100 dataset is considered and an MS-SSIM of 0.965 at an approximate SNR of only 10 dB on the Kodak dataset. The proposed system holds promise to maintain consistent performance even at low SNRs across various channel conditions. Full article

(This article belongs to the Section Information and Communication Technologies)

► Show Figures

Figure 1

30 pages, 1096 KiB

Open AccessReview

Next-Generation Smart Inverters: Bridging AI, Cybersecurity, and Policy Gaps for Sustainable Energy Transition

by Hilmy Awad and Ehab H. E. Bayoumi

Technologies 2025, 13(4), 136; https://doi.org/10.3390/technologies13040136 - 1 Apr 2025

Viewed by 367

Abstract

Smart inverters are pivotal in modern renewable energy systems, enabling efficient grid integration, stability, and advanced control of distributed energy resources. While existing literature addresses their technical functionalities, significant research gaps persist in areas such as interoperability, cybersecurity, standardization, and the integration of [...] Read more.

Smart inverters are pivotal in modern renewable energy systems, enabling efficient grid integration, stability, and advanced control of distributed energy resources. While existing literature addresses their technical functionalities, significant research gaps persist in areas such as interoperability, cybersecurity, standardization, and the integration of artificial intelligence for adaptive control. This article provides a comprehensive review of smart inverter technologies, emphasizing their role in renewable energy applications, advanced control strategies, and unresolved challenges. By systematically analyzing recent advancements and case studies, the paper identifies critical limitations in current practices, including economic barriers, regulatory misalignments, and fault tolerance under dynamic grid conditions. The review contributes to the field by synthesizing dispersed knowledge, highlighting under-researched areas, and proposing actionable pathways for future innovation. The main findings reveal the transformative potential of AI-driven grid-forming inverters for enhancing grid stability and resilience. However, their widespread adoption is hindered by the absence of harmonized standards and misaligned policy frameworks. Consequently, this review underscores the urgent need for policymakers to develop and implement supportive regulatory structures that facilitate the deployment of AI-enabled smart inverters and establish unified standards to ensure interoperability and cybersecurity. This work serves as a foundational reference for researchers and policymakers aiming to address technical and systemic bottlenecks in smart inverter deployment. Full article

► Show Figures

Figure 1

Journal Menu

Journal Browser

Technologies, Volume 13, Issue 4 (April 2025) – 42 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI