Journal of Sensor and Actuator Networks

16 pages, 15073 KB

Open AccessArticle

A Bidirectional, Full-Duplex, Implantable Wireless CMOS System for Prosthetic Control

by Riccardo Collu, Cinzia Salis, Elena Ferrazzano and Massimo Barbaro

J. Sens. Actuator Netw. 2025, 14(5), 92; https://doi.org/10.3390/jsan14050092 - 10 Sep 2025

Implantable medical devices present several technological challenges, one of the most critical being how to provide power supply and communication capabilities to a device hermetically sealed within the body. Using a battery as a power source represents a potential harm for the individual’s [...] Read more.

Implantable medical devices present several technological challenges, one of the most critical being how to provide power supply and communication capabilities to a device hermetically sealed within the body. Using a battery as a power source represents a potential harm for the individual’s health because of possible toxic chemical release or overheating, and it requires periodic surgery for replacement. This paper proposes a batteryless implantable device powered by an inductive link and equipped with bidirectional wireless communication channels. The device, designed in a 180 nm CMOS process, is based on two different pairs of mutually coupled inductors that provide, respectively, power and a low-bitrate bidirectional communication link and a separate, high-bitrate, one-directional upstream connection. The main link is based on a 13.56 MHz carrier and allows power transmission and a half-duplex two-way communication at 106 kbps (downlink) and 30 kbps (uplink). The secondary link is based on a 27 MHz carrier, which provides one-way communication at 2.25 Mbps only in uplink. The low-bitrate links are needed to send commands and monitor the implanted system, while the high-bitrate link is required to receive a continuous stream of information from the implanted sensing devices. The microchip acts as a hub for power and data wireless transmission capable of managing up to four different neural recording and stimulation front ends, making the device employable in a complex, distributed, bidirectional neural prosthetic system. Full article

(This article belongs to the Topic Wireless Energy Harvesting and Power Transfer for Communications and Networks)

► Show Figures

Figure 1

24 pages, 3485 KB

Open AccessArticle

Impact Evaluation of Sound Dataset Augmentation and Synthetic Generation upon Classification Accuracy

by Eleni Tsalera, Andreas Papadakis, Gerasimos Pagiatakis and Maria Samarakou

J. Sens. Actuator Netw. 2025, 14(5), 91; https://doi.org/10.3390/jsan14050091 - 9 Sep 2025

Abstract

We investigate the impact of dataset augmentation and synthetic generation techniques on the accuracy of supervised audio classification based on state-of-the-art neural networks used as classifiers. Dataset augmentation techniques are applied upon the raw sound and its transformed image format. Specifically, sound augmentation [...] Read more.

We investigate the impact of dataset augmentation and synthetic generation techniques on the accuracy of supervised audio classification based on state-of-the-art neural networks used as classifiers. Dataset augmentation techniques are applied upon the raw sound and its transformed image format. Specifically, sound augmentation techniques are applied prior to spectral-based transformation and include time stretching, pitch shifting, noise addition, volume controlling, and time shifting. Image augmentation techniques are applied after the transformation of the sound into a scalogram, involving scaling, shearing, rotation, and translation. Synthetic sound generation is based on the AudioGen generative model, triggered through a series of customized prompts. Augmentation and synthetic generation are applied to three sound categories: (a) human sounds, (b) animal sounds, and (c) sounds of things, with each category containing ten sound classes with 20 samples retrieved from the ESC-50 dataset. Sound- and image-orientated neural network classifiers have been used to classify the augmented datasets and their synthetic additions. VGGish and YAMNet (sound classifiers) employ spectrograms, while ResNet50 and DarkNet53 (image classifiers) employ scalograms. The streamlined AI-based process of augmentation and synthetic generation, enhanced classifier fine-tuning and inference allowed for a consistent, multicriteria-comparison of the impact. Classification accuracy has increased for all augmentation and synthetic generation scenarios; however, the increase has not been uniform among the techniques, the sound types, and the percentage of the training set population increase. The average increase in classification accuracy ranged from 2.05% for ResNet50 to 9.05% for VGGish. Our findings reinforce the benefit of audio augmentation and synthetic generation, providing guidelines to avoid accuracy degradation due to overuse and distortion of key audio features. Full article

(This article belongs to the Special Issue AI-Assisted Machine-Environment Interaction)

► Show Figures

Figure 1

28 pages, 9195 KB

Open AccessArticle

DAR-MDE: Depth-Attention Refinement for Multi-Scale Monocular Depth Estimation

by Saddam Abdulwahab, Hatem A. Rashwan, Moumen T. El-Melegy and Domenec Puig

J. Sens. Actuator Netw. 2025, 14(5), 90; https://doi.org/10.3390/jsan14050090 - 1 Sep 2025

Viewed by 451

Abstract

Monocular Depth Estimation (MDE) remains a challenging problem due to texture ambiguity, occlusion, and scale variation in real-world scenes. While recent deep learning methods have made significant progress, maintaining structural consistency and robustness across diverse environments remains difficult. In this paper, we propose [...] Read more.

Monocular Depth Estimation (MDE) remains a challenging problem due to texture ambiguity, occlusion, and scale variation in real-world scenes. While recent deep learning methods have made significant progress, maintaining structural consistency and robustness across diverse environments remains difficult. In this paper, we propose DAR-MDE, a novel framework that combines an autoencoder backbone with a Multi-Scale Feature Aggregation (MSFA) module and a Refining Attention Network (RAN). The MSFA module enables the model to capture geometric details across multiple resolutions, while the RAN enhances depth predictions by attending to structurally important regions guided by depth-feature similarity. We also introduce a multi-scale loss based on curvilinear saliency to improve edge-aware supervision and depth continuity. The proposed model achieves robust and accurate depth estimation across varying object scales, cluttered scenes, and weak-texture regions. We evaluated DAR-MDE on the NYU Depth v2, SUN RGB-D, and Make3D datasets, demonstrating competitive accuracy and real-time inference speeds (19 ms per image) without relying on auxiliary sensors. Our method achieves a

δ

< 1.25 accuracy of 87.25% and a relative error of 0.113 on NYU Depth v2, outperforming several recent state-of-the-art models. Our approach highlights the potential of lightweight RGB-only depth estimation models for real-world deployment in robotics and scene understanding. Full article

► Show Figures

Figure 1

26 pages, 17668 KB

Open AccessArticle

Real-Time Damage Detection and Localization on Aerospace Structures Using Graph Neural Networks

by Emiliano Del Priore and Luca Lampani

J. Sens. Actuator Netw. 2025, 14(5), 89; https://doi.org/10.3390/jsan14050089 - 29 Aug 2025

Viewed by 420

Abstract

This work presents a novel Graph Neural Network (GNN) based framework for structural damage detection and localization in composite aerospace structures. The sensor network is modeled as a graph whose nodes correspond to the strain measurement points placed on the system, while the [...] Read more.

This work presents a novel Graph Neural Network (GNN) based framework for structural damage detection and localization in composite aerospace structures. The sensor network is modeled as a graph whose nodes correspond to the strain measurement points placed on the system, while the edges capture spatial and structural relationships among sensors. Strain mode shapes, extracted via Automated Operational Modal Analysis (AOMA), are used as input features for the GNN. Two architectures are developed: one for binary damage detection and another for damage localization, the latter outputting a spatial probability distribution of damage over the structure. Both networks are trained and validated on synthetic datasets generated from high-fidelity finite element transient simulations performed on a composite wing equipped with 40 strain sensors. The obtained results show strong effectiveness in both detection and localization tasks, thus highlighting the potential of leveraging GNNs for topology-aware Structural Health Monitoring applications. In particular, the proposed framework achieves an AUC of 0.97 for damage detection and a mean localization error of approximately 3% of the wingspan on the synthetic dataset. The performance of the GNN is also compared with a fully connected and a convolutional neural network, demonstrating significant improvements in the localization accuracy. Full article

(This article belongs to the Special Issue Fault Diagnosis in the Internet of Things Applications)

► Show Figures

Figure 1

15 pages, 4071 KB

Open AccessArticle

Electrostatic MEMS Phase Shifter for SiN Photonic Integrated Circuits

by Seyedfakhreddin Nabavi, Michaël Ménard and Frederic Nabki

J. Sens. Actuator Netw. 2025, 14(5), 88; https://doi.org/10.3390/jsan14050088 - 29 Aug 2025

Viewed by 528

Abstract

Optical phase modulation is essential for a wide range of silicon photonic integrated circuits used in communication applications. In this study, an optical phase shifter utilizing photo-elastic effects is proposed, where mechanical stress is induced by electrostatic micro-electro-mechanical systems (MEMS) with actuators arranged [...] Read more.

Optical phase modulation is essential for a wide range of silicon photonic integrated circuits used in communication applications. In this study, an optical phase shifter utilizing photo-elastic effects is proposed, where mechanical stress is induced by electrostatic micro-electro-mechanical systems (MEMS) with actuators arranged in a comb drive configuration. The design incorporates suspended serpentine silicon nitride (SiN) optical waveguides. Through extensive numerical simulations, it is shown that the change in the effective refractive index (n_eff) of the optical waveguide is a function of the voltage applied to the electrostatic actuators and that such n_eff tuning can be achieved for a broad range of wavelengths. Implemented within one arm of an unbalanced Mach–Zehnder interferometer (MZI), the phase shifter achieves a phase change of π when the stressed optical path measures 4.7 mm, and the actuators are supplied with 80 V DC and consume almost no power. This results in a half-wave voltage-length product (V_πL) of 37.6 V·cm. Comparative analysis with contemporary optical phase shifters highlights the proposed design’s superior power efficiency, compact footprint, and simplified fabrication process, making it a highly efficient component for reconfigurable MEMS-based silicon nitride photonic integrated circuits. Full article

► Show Figures

Figure 1

36 pages, 1564 KB

Open AccessReview

Hybrid Path Planning Algorithm for Autonomous Mobile Robots: A Comprehensive Review

by Mithun Shanmugaraja, Mohanraj Thangamuthu and Sivasankar Ganesan

J. Sens. Actuator Netw. 2025, 14(5), 87; https://doi.org/10.3390/jsan14050087 - 28 Aug 2025

Viewed by 540

Abstract

Path planning is a complex task in robotics, requiring an efficient and adaptive algorithm to find the shortest path in a dynamic environment. The traditional path planning methods, such as graph-based algorithms, sampling-based algorithms, reaction-based algorithms, and optimization-based algorithms, have limitations in computational [...] Read more.

Path planning is a complex task in robotics, requiring an efficient and adaptive algorithm to find the shortest path in a dynamic environment. The traditional path planning methods, such as graph-based algorithms, sampling-based algorithms, reaction-based algorithms, and optimization-based algorithms, have limitations in computational efficiency, real-time adaptability, and obstacle avoidance. To address these challenges, hybrid path planning algorithms combine the strengths of multiple techniques to enhance performance. This paper includes a comprehensive review of hybrid approaches based on graph-based algorithms, sampling-based algorithms, reaction-based algorithms, and optimization-based algorithms. Also, this article discusses the advantages and limitations, supported by a comparative evaluation of computational complexity, path optimization, and finding the shortest path in a dynamic environment. Finally, we propose an AI-driven adaptive path planning approach to solve the difficulties. Full article

(This article belongs to the Special Issue AI-Assisted Machine-Environment Interaction)

► Show Figures

Figure 1

11 pages, 650 KB

Open AccessArticle

Efficient and Low-Cost Modular Polynomial Multiplier for WSN Security

by Fariha Haroon and Hua Li

J. Sens. Actuator Netw. 2025, 14(5), 86; https://doi.org/10.3390/jsan14050086 - 25 Aug 2025

Viewed by 362

Abstract

Wireless Sensor Network (WSN) technology has constrained computing resources that require efficient and low-cost cryptographic hardware to provide security services, particularly when dealing with large modular polynomial multiplication in cryptography. In this paper, a cost-efficient reconfigurable Karatsuba modular polynomial multiplier is proposed for [...] Read more.

Wireless Sensor Network (WSN) technology has constrained computing resources that require efficient and low-cost cryptographic hardware to provide security services, particularly when dealing with large modular polynomial multiplication in cryptography. In this paper, a cost-efficient reconfigurable Karatsuba modular polynomial multiplier is proposed for general modulus polynomials. The modulus polynomial can be changed easily depending on the application. The proposed modular polynomial multiplier is synthesized and simulated by the AMD Vivado Design Tool. The design’s performance on ADP (Area Delay Product) has been improved compared to previous designs. It can be applied in ECC encryption to speed up the security services in WSN. Full article

(This article belongs to the Special Issue Applications of Wireless Sensor Networks: Innovations and Future Trends)

► Show Figures

Figure 1

Journal Menu

Journal Browser

J. Sens. Actuator Netw., Volume 14, Issue 5 (October 2025) – 7 articles

Further Information

Guidelines

MDPI Initiatives

Follow MDPI