Search Results (373)

Urban noise levels that exceed the World Health Organization (WHO) recommendations have become a growing concern due to their adverse effects on public health. In Bogotá, Colombia, studies by the District Department of Environment (SDA) indicate that 11.8% of the population is exposed to noise levels above the WHO limits. This research aims to identify and categorize environmental noise sources in real time using an embedded intelligent system. A total of 657 labeled audio clips were collected across eight classes and processed using a 60/20/20 train–validation–test split, ensuring that audio segments from the same continuous recording were not mixed across subsets. The system was implemented on a Raspberry Pi 2W equipped with a UMIK-1 microphone and powered by a 90 W solar panel with a 12 V battery, enabling autonomous operation. The TinyML-based model achieved precision and recall values between 0.92 and 1.00, demonstrating high performance under real urban conditions. Heavy vehicles and motorcycles accounted for the largest proportion of classified samples. Although airplane-related events were less frequent, they reached maximum sound levels of up to 88.4 dB(A), exceeding the applicable local limit of 70 dB(A) by approximately 18 dB(A) rather than by percentage. In conclusion, the results demonstrate that on-device TinyML classification is a feasible and effective strategy for urban noise monitoring. Local inference reduces latency, bandwidth usage, and privacy risks by eliminating the need to transmit raw audio to external servers. This approach provides a scalable and sustainable foundation for noise management in smart cities and supports evidence-based public policies aimed at improving urban well-being. This work presents an introductory and exploratory study on the application of TinyML for acoustic environmental monitoring, aiming to evaluate its feasibility and potential for large-scale implementation. Full article

In recent years, the demand for efficient neural networks in embedded contexts has grown, driven by the need for real-time inference with limited resources. While GPUs offer high performance, their size, power consumption, and cost often make them unsuitable for constrained or large-scale applications. FPGAs have therefore emerged as a promising alternative, combining reconfigurability, parallelism, and increasingly favorable cost–performance ratios. They are especially relevant in domains such as robotics, IoT, and autonomous drones, where rapid sensor fusion and low power consumption are critical. This work presents the full implementation of a neural network on a low-cost FPGA, targeting real-time image and video recognition for drone applications. The workflow included training and quantizing a YOLOv3-Tiny model with Brevitas and PyTorch, converting it into hardware logic using the FINN framework, and optimizing the hardware design to maximize use of the reprogrammable silicon area and inference time. A custom driver was also developed to allow the device to operate as a TPU. The resulting accelerator, deployed on a Xilinx Zynq-7020, could recognize 208 frames per second (FPS) when running at a 200 MHz clock frequency, while consuming only 2.55 W. Compared to Google’s Coral Edge TPU, the system offers similar inference speed with greater flexibility, and outperforms other FPGA-based approaches in the literature by a factor of three to seven in terms of FPS/W. Full article

This paper introduces a hybrid learning framework that synergistically combines Reinforcement Learning (RL) and Supervised Learning (SL) to train autonomous cyber-defense agents capable of operating effectively in dynamic and adversarial environments. The proposed approach leverages RL for strategic exploration and policy development, while incorporating SL to distill high-reward trajectories into refined policy updates, enhancing sample efficiency, learning stability, and robustness. The framework first targets specialized agent training, where each agent is optimized against a specific adversarial behavior. Subsequently, it is extended to enable the training of a generalized agent that learns to counter multiple, diverse attack strategies through multi-task and curriculum learning techniques. Comprehensive experiments conducted in the CybORG simulation environment demonstrate that the hybrid RL–SL framework consistently outperforms pure RL baselines across both specialized and generalized settings, achieving higher cumulative rewards. Specifically, hybrid-trained agents achieve up to 23% higher cumulative rewards in specialized defense tasks and approximately 18% improvements in generalized defense scenarios compared to RL-only agents. Moreover, incorporating temporal context into the observation space yields a further 4–6% performance gain in policy robustness. Furthermore, we investigate the impact of augmenting the observation space with historical actions and rewards, revealing consistent, albeit incremental, gains in SL-based learning performance. Key contributions of this work include: $\left(i\right)$ a novel hybrid learning paradigm that integrates RL and SL for effective cyber-defense policy learning, $\left(ii\right)$ a scalable extension for training generalized agents across heterogeneous threat models, and $\left(iii\right)$ empirical analysis on the role of temporal context in agent observability and decision-making. Collectively, the results highlight the promise of hybrid learning strategies for building intelligent, resilient, and adaptable cyber-defense systems in evolving threat landscapes. Full article

Accurate vehicle state estimation in a lane coordinate system is essential for safe and reliable operation of Advanced Driver Assistance Systems (ADASs) and autonomous driving. However, achieving robust lane-based state estimation using only low-cost sensors, such as a camera, an IMU, and a steering angle sensor, remains challenging due to the complexity of vehicle dynamics and the inherent signal delays in vision systems. This paper presents a lane-coordinate-based vehicle state estimator that addresses these challenges by combining a vehicle dynamics-based bicycle model with an Extended Kalman Filter (EKF) and a signal delay compensation algorithm. The estimator performs real-time estimation of lateral position, lateral velocity, and heading angle, including the unmeasurable lateral velocity about the lane, by predicting the vehicle’s state evolution during camera processing delays. A computationally efficient camera processing pipeline, incorporating lane segmentation via a pre-trained network and lane-based state extraction, is implemented to support practical applications. Validation using real vehicle driving data on straight and curved roads demonstrates that the proposed estimator provides continuous, high-accuracy, and delay-compensated lane-coordinate-based vehicle states. Compared to conventional camera-only methods and estimators without delay compensation, the proposed approach significantly reduces estimation errors and phase lag, enabling the reliable and real-time acquisition of vehicle-state information critical for ADAS and autonomous driving applications. Full article

Background/Objectives: Autonomously functioning thyroid nodules (AFTNs) are most frequently diagnosed as benign. However, they show high ratings in ultrasound (US) risk stratification systems (RSSs) that utilize the current clinical standard methodology of conventional static image capture (SIC) documentation. The objective of this study was to evaluate the RSS ratings and respective fine needle cytology (FNC) recommendations of cine loop (CL) video sequences in comparison to SIC. Methods: 407 patients with 424 AFTNs were enrolled in this unicentric, retrospective study between 11/2015 and 11/2023. Recorded US CL and SIC were analyzed lesion-wise and compared regarding US features, Kwak and ACR TIRADS, ACR FNC recommendations, as well as assessment difficulties and artifacts. Statistical analyses were conducted using the Chi² test and Spearman’s correlation coefficient in SPSS software. p-values < 0.05 were considered significant. Results: Strong to very strong correlations were observed for all US features, RSS ratings, and ACR FNC recommendations (Spearman’s correlation: each p < 0.001), comparing CL and SIC. For >60% of the AFTNs, ACR FNC recommendation was given. Kwak TIRADS were more consistent with the benign nature of AFTNs than the ACR ratings. CL captured significantly more “echogenic foci” than SIC (Chi²: p < 0.001). Artifacts (poor image quality, acoustic shadowing, sagittal incompletely displayed AFTN) were significantly more common on CL, affecting ~40% of AFTNs, compared to ~15% on SIC (Chi²: each p < 0.05). Weak correlation was observed for assessment confidence between CL and SIC, with SIC outperforming CL (Spearman’s correlation: each p < 0.001). Conclusions: A strong correlation was identified between CL and SIC in terms of RSS ratings and ACR FNC recommendations. Kwak is a superior representative of the benign character of AFTNs than ACR. However, CL provided more detailed information while being associated with decreased observer confidence and more artifacts. Specific operator training and technical improvements, including AI implementation, could improve image quality in future. Full article

Unlike traditional control systems that rely on physical input devices, facial gesture-based interaction offers a contactless and intuitive method for operating autonomous systems. Recent advances in computer vision and deep learning have enabled the use of facial expressions and movements for command recognition in human–robot interaction. In this work, we propose a lightweight, real-time facial gesture recognition method, GhostNet-BiLSTM-Attention (GBA), which integrates GhostNet and BiLSTM with an attention mechanism, is trained on the FaceGest dataset, and is integrated with a 3D robot simulation in Unity. The system is designed to recognize predefined facial gestures such as head tilts, eye blinks, and mouth movements with high accuracy and low inference latency. Recognized gestures are mapped to specific robot commands and transmitted to a Unity-based simulation environment via socket communication across machines. This framework enables smooth and immersive robot control without the need for conventional controllers or sensors. Real-time evaluation demonstrates the system’s robustness and responsiveness under varied user and lighting conditions, achieving a classification accuracy of 99.13% on the FaceGest dataset. The GBA holds strong potential for applications in assistive robotics, contactless teleoperation, and immersive human–robot interfaces. Full article

This study presents the design and evaluation of AquaSignalNet, a deep learning-based system for recognizing underwater visual commands to enable the autonomous navigation of a Remotely Operated Vehicle (ROV). The system is built on a ResNet50 V2 architecture and trained with a custom dataset, UVSRD, comprising 33,800 labeled images across 12 gesture classes, including directional commands, speed values, and vertical motion instructions. The model was deployed on a Raspberry Pi 4 integrated with a TIVA C microcontroller for real-time motor control, a PID-based depth control loop, and an MPU9250 sensor for orientation tracking. Experiments were conducted in a controlled pool environment using printed signal cards to define two autonomous trajectories. In the first trajectory, the system achieved 90% success, correctly interpreting a mixed sequence of turns, ascents, and speed changes. In the second, more complex trajectory, involving a rectangular inspection loop and multi-layer navigation, the system achieved 85% success, with failures mainly due to misclassification resulting from lighting variability near the water surface. Unlike conventional approaches that rely on QR codes or artificial markers, AquaSignalNet employs markerless visual cues, offering a flexible alternative for underwater inspection, exploration, and logistical operations. The results demonstrate the system’s viability for real-time gesture-based control. Full article

This paper presents a neural network–driven framework for solving the inverse problem of initial parameter estimation in air-dropped package missions. Unlike traditional analytical methods, which are computationally intensive and often impractical in real time, the proposed system leverages the flexibility of multilayer perceptrons to model both forward and inverse relationships between drop conditions and flight outcomes. In the forward stage, a trained network predicts range, flight time, and impact velocity from predefined release parameters. In the inverse stage, a deeper neural model reconstructs the required release velocity, angle, and altitude directly from the desired operational outcomes. By employing a hybrid workflow—combining physics-based simulation with neural approximation—our approach generates large, high-quality datasets at low computational cost. Results demonstrate that the inverse network achieves high accuracy across deterministic and stochastic tests, with minimal error when operating within the training domain. The study confirms the suitability of neural architectures for tackling complex, nonlinear identification tasks in precision airdrop operations. Beyond their technical efficiency, such models enable agile, GPS-independent mission planning, offering a reliable and low-cost decision support tool for humanitarian aid, scientific research, and defense logistics. This work highlights how artificial intelligence can transform conventional trajectory design into a fast, adaptive, and autonomous capability. Full article

With the growing application of unmanned aerial vehicles (UAVs) in complex, stochastic environments, autonomous navigation and obstacle avoidance represent critical technical challenges requiring urgent solutions. This study proposes an innovative deep reinforcement learning (DRL) framework that leverages multimodal perception through the fusion of LiDAR and depth camera data. A sophisticated multi-sensor data preprocessing mechanism is designed to extract multimodal features, significantly enhancing the UAV’s situational awareness and adaptability in intricate, stochastic environments. In the high-level decision-maker of the framework, to overcome the intrinsic limitation of low sample efficiency in DRL algorithms, this study introduces an advanced decision-making algorithm, Soft Actor-Critic with Prioritization (SAC-P), which markedly accelerates model convergence and enhances training stability through optimized sample selection and utilization strategies. Validated within a high-fidelity Robot Operating System (ROS) and Gazebo simulation environment, the proposed framework achieved a task success rate of 81.23% in comparative evaluations, surpassing all baseline methods. Notably, in generalization tests conducted in previously unseen and highly complex environments, it maintained a success rate of 72.08%, confirming its robust and efficient navigation and obstacle avoidance capabilities in complex, densely cluttered environments with stochastic obstacle distributions. Full article

An increased level of autonomy is attractive above all in the framework of proximity operations, and researchers are focusing more and more on artificial intelligence techniques to improve spacecraft’s capabilities in these scenarios. This work presents an autonomous AI-based guidance algorithm to plan the path of a chaser spacecraft for the map reconstruction of an artificial uncooperative target, coupled with Model Predictive Control for the tracking of the generated trajectory. Deep reinforcement learning is particularly interesting for enabling spacecraft’s autonomous guidance, since this problem can be formulated as a Partially Observable Markov Decision Process and because it leverages domain randomization well to cope with model uncertainty, thanks to the neural networks’ generalizing capabilities. The main drawback of this method is that it is difficult to verify its optimality mathematically and the constraints can be added only as part of the reward function, so it is not guaranteed that the solution satisfies them. To this end a convex Model Predictive Control formulation is employed to track the DRL-based trajectory, while simultaneously enforcing compliance with the constraints. Two neural network architectures are proposed and compared: a recurrent one and the more recent transformer. The trained reinforcement learning agent is then tested in an end-to-end AI-based pipeline with image generation in the loop, and the results are presented. The computational effort of the entire guidance and control strategy is also verified on a Raspberry Pi board. This work represents a viable solution to apply artificial intelligence methods for spacecraft’s autonomous motion, still retaining a higher level of explainability and safety than that given by more classical guidance and control approaches. Full article

Cybersecurity has become a critical concern in the automotive sector, where the increasing connectivity and complexity of modern vehicles—particularly in the context of autonomous driving—have significantly expanded the attack surface. In response to these challenges, this paper presents the Welcome To The Machine (WTTM) framework, developed to support proactive and structured cyber risk management throughout the entire vehicle lifecycle. Specifically tailored to the automotive domain, the framework encompasses four core actions: detection, analysis, response, and remediation. A central element of WTTM is the WTTM Questionnaire, designed to assess the organizational cybersecurity maturity of automotive manufacturers and suppliers. The questionnaire addresses six key areas: Governance, Risk Management, Concept and Design, Security Requirements, Validation and Testing, and Supply Chain. This paper focuses on the development and validation of WTTM-Q. Statistical validation was performed using responses from 43 participants, demonstrating high internal consistency (Cronbach’s alpha > 0.70) and strong construct validity (CFI = 0.94, RMSEA = 0.061). A supervised classifier (XGBoost), trained on 115 hypothetical response configurations, was employed to predict a priori risk classes, achieving 78% accuracy and a ROC AUC of 0.84. The WTTM framework, supported by a Vehicle Security Operations Center, provides a scalable, standards-aligned solution for enhancing cybersecurity in the automotive industry. Full article

Intelligent autonomous systems in open-pit mining operations face critical challenges in perception and decision-making due to sensor-based visual degradations, particularly lens soiling and sun glare, which significantly compromise the performance and safety of integrated mining automation systems. We propose a comprehensive intelligent framework leveraging single-domain generalization with traditional data augmentation techniques, specifically Photometric Distortion (PD) and Contrast Limited Adaptive Histogram Equalization (CLAHE), integrated within the BiSeNetV1 architecture. Our systematic approach evaluated four state-of-the-art backbones: ResNet-50, MobileNetV2 (Convolutional Neural Networks (CNN)-based), SegFormer-B0, and Twins-PCPVT-S (ViT-based) within an end-to-end autonomous system architecture. The model was trained on clean images from the AutoMine dataset and tested on degraded visual conditions without requiring architectural modifications or additional training data from target domains. ResNet-50 demonstrated superior system robustness with mean Intersection over Union (IoU) of 84.58% for lens soiling and 80.11% for sun glare scenarios, while MobileNetV2 achieved optimal computational efficiency for real-time autonomous systems with 55.0 Frames Per Second (FPS) inference speed while maintaining competitive accuracy (81.54% and 71.65% mIoU respectively). Vision Transformers showed superior stability in system performance but lower overall performance under severe degradations. The proposed intelligent augmentation-based approach maintains high accuracy while preserving real-time computational efficiency, making it suitable for deployment in autonomous mining vehicle systems. Traditional augmentation approaches achieved approximately 30% superior performance compared to advanced GAN-based domain generalization methods, providing a practical solution for robust perception systems without requiring expensive multi-domain training datasets. Full article

The possibility of developing autonomous vessels has recently become increasingly interesting. As most vessels are powered by diesel engines, the idea of developing a method to detect engine malfunctions by analyzing signals from microphones placed near the engine and accelerometers mounted on the engine housing is intriguing. This paper presents a method for detecting engine malfunctions by analyzing signals obtained from the output of a microphone and accelerometer. The algorithm is based on signal analysis in the frequency domain using discrete Fourier transform (DFT), and the same procedure is applied to both acoustic and vibration data. The proposed method was tested on a six-cylinder marine diesel engine where a fault was emulated by deactivating one cylinder. In controlled experiments across five rotational speeds, the method achieved an accuracy of approximately 98.3% when trained on 75 operating cycles and evaluated over 15 cycles. The average precision and recall across all sensors exceeded 97% and 96%, respectively. The ability of the algorithm to treat microphone and accelerometer signals identically simplifies implementation, and the detection accuracy can be increased further by adding additional sensors. Full article

This paper presents a deep reinforcement learning-based demand response (DR) optimization framework for active distribution networks under uncertainty and user heterogeneity. The proposed model utilizes a Double Deep Q-Network (Double DQN) to learn adaptive, multi-period DR strategies across residential, commercial, and electric vehicle (EV) participants in a 24 h rolling horizon. By incorporating a structured state representation—including forecasted load, photovoltaic (PV) output, dynamic pricing, historical DR actions, and voltage states—the agent autonomously learns control policies that minimize total operational costs while maintaining grid feasibility and voltage stability. The physical system is modeled via detailed constraints, including power flow balance, voltage magnitude bounds, PV curtailment caps, deferrable load recovery windows, and user-specific availability envelopes. A case study based on a modified IEEE 33-bus distribution network with embedded PV and DR nodes demonstrates the framework’s effectiveness. Simulation results show that the proposed method achieves significant cost savings (up to 35% over baseline), enhances PV absorption, reduces load variance by 42%, and maintains voltage profiles within safe operational thresholds. Training curves confirm smooth Q-value convergence and stable policy performance, while spatiotemporal visualizations reveal interpretable DR behavior aligned with both economic and physical system constraints. This work contributes a scalable, model-free approach for intelligent DR coordination in smart grids, integrating learning-based control with physical grid realism. The modular design allows for future extension to multi-agent systems, storage coordination, and market-integrated DR scheduling. The results position Double DQN as a promising architecture for operational decision-making in AI-enabled distribution networks. Full article

Autonomous docking is a critical capability for enabling fully automated operations in industrial and logistics environments using Autonomous Mobile Robots (AMRs). Traditional rule-based docking approaches often struggle with generalization and robustness in complex, dynamic scenarios. This paper presents a deep reinforcement learning-based autonomous docking framework that integrates Proximal Policy Optimization (PPO) with multi-sensor fusion. It includes YOLO-based vision detection, depth estimation, and LiDAR-based orientation correction. A concise 4D state vector, comprising relative position and angle indicators, is used to guide a continuous control policy. The outputs are linear and angular velocity commands for smooth and accurate docking. The training is conducted in a Gym-compatible Gazebo simulation, acting as a digital twin of the real-world system, and incorporates realistic variations in lighting, obstacle placement, and marker visibility. A designed reward function encourages alignment accuracy, progress, and safety. The final policy is deployed on a real robot via a sim-to-real transfer pipeline, supported by a ROS-based transfer node. Experimental results demonstrate that the proposed method achieves robust and precise docking behavior under diverse real-world conditions, validating the effectiveness of PPO-based learning and sensor fusion for practical autonomous docking applications. Full article