Search Results (149)

Groundwater plays a crucial role in sustaining agriculture and livelihoods in the arid Middle Draa Valley (MDV) of southeastern Morocco. However, increasing groundwater extraction, declining rainfall, and the absence of effective floodwater harvesting systems have led to severe aquifer depletion. This study applies and compares six machine learning (ML) algorithms—decision trees (CART), ensemble methods (random forest, LightGBM, XGBoost), distance-based learning (k-nearest neighbors), and support vector machines—integrating GIS, satellite data, and field observations to delineate zones suitable for groundwater recharge. The results indicate that ensemble tree-based methods yielded the highest predictive accuracy, with LightGBM outperforming the others by achieving an overall accuracy of 0.90. Random forest and XGBoost also demonstrated strong performance, effectively identifying priority areas for artificial recharge, particularly near ephemeral streams. A feature importance analysis revealed that soil permeability, elevation, and stream proximity were the most influential variables in recharge zone delineation. The generated maps provide valuable support for irrigation planning, aquifer conservation, and floodwater management. Overall, the proposed machine learning–geospatial framework offers a robust and transferable approach for mapping groundwater recharge zones (GWRZ) in arid and semi-arid regions, contributing to the achievement of Sustainable Development Goals (SDGs))—notably SDG 6 (Clean Water and Sanitation), by enhancing water-use efficiency and groundwater recharge (Target 6.4), and SDG 13 (Climate Action), by supporting climate-resilient aquifer management. Full article

Accurately hindcasting waves in semi-enclosed, fetch-limited basins remains challenging for reanalysis models, which tend to underestimate storm peaks near the coast. We developed interpretable ML models for Rijeka Bay (northern Adriatic) using only wind observations from two land-based wind stations to predict buoy H_m0 measurements spanning 2009–2011 (testing) and 2019–2021 (training and validation). The tested tree-based models included Random Forest, XGBoost, and Explainable Boosting Machine. This study introduces a novel approach in the literature by employing weighted schemes and feature engineering to enhance the predictive performance of interpretable, low-complexity machine learning models in hindcasting waves. Representing wind direction as sine–cosine components generally reduced RMSE and BIAS relative to traditional speed–direction inputs, while an exponential sample weight scheme that emphasized storm waves halved extreme H_m0 underprediction without inflating overall RMSE. The best-performing model, a Random Forest model, achieved an RMSE of 0.096 m and a correlation of 0.855 on the unseen test set—30% lower overall RMSE and 50% lower extreme wave RMSE than the MEDSEA and COEXMED hindcasts. Additionally, the underprediction was reduced by 90% compared to these reanalysis models. The method offers a computationally lightweight, transferable supplement to numerical wave guidance for coastal engineering and harbor operations. Full article

Process control and quality assurance of wire + arc additive manufacturing (WAAM) and automated welding rely heavily on in-process monitoring videos to quantify variables such as melt pool geometry, location and size of droplet transfer, arc characteristics, etc. To enable feedback control based upon this information, an automatic and robust segmentation method for monitoring of videos and images is required. However, video segmentation in WAAM and welding is challenging due to constantly fluctuating arc brightness, which varies with deposition and welding configurations. Additionally, conventional computer vision algorithms based on greyscale value and gradient lack flexibility and robustness in this scenario. Deep learning offers a promising approach to WAAM video segmentation; however, the prohibitive time and cost associated with creating a well-labelled, suitably sized dataset have hindered its widespread adoption. The emergence of large computer vision models, however, has provided new solutions. In this study a semi-automatic annotation tool for WAAM videos was developed based upon the computer vision foundation model SAM and the video object tracking model XMem. The tool can enable annotation of the video frames hundreds of times faster than traditional manual annotation methods, thus making it possible to achieve rapid quantitative analysis of WAAM and welding videos with minimal user intervention. To demonstrate the effectiveness of the tool, three cases are demonstrated: online wire position closed-loop control, droplet transfer behaviour analysis, and assembling a dataset for dedicated deep learning segmentation models. This work provides a broader perspective on how to exploit large models in WAAM and weld deposits. Full article

Machine learning (ML) has gained significant attention for unraveling the complex, nonlinear relationships between soil moisture (SM) and various predictive variables, including remote sensing (RS; reflectance, brightness temperature, backscatter coefficients) and biophysical (topographic, soil, vegetation, and weather) variables. We reviewed the literature to extract and synthesize ML algorithms, reliable input features, and challenges in SM estimation using RS data. We analyzed results from 144 articles published from 2010 to 2024. Random forest (40 out of 67 studies), support vector regressor (13 out of 39 studies), and artificial neural networks (12 out of 27 studies) often outperformed other algorithms to estimate SM using RS datasets. Multi-source RS data often outperformed single-source data in SM estimation. Satellite-derived features, such as vegetation indices and backscattering coefficients, provided critical information on surface SM (SSM) variability to estimate SSM. For root zone SM estimation, soil properties and SSM generally were more reliable predictors than surface information derived solely from RS. Two recent advances—the use of semi-empirical models and L-band SAR to mitigate vegetation effects, and transfer learning to improve model transferability—have shown promise in addressing key challenges in SM estimation. Full article

The adoption of AI vision models in specialized industries is often hindered by the substantial requirement for extensive, manually annotated image datasets. Even when employing transfer learning, robust model development typically necessitates tens of thousands of such images, a process that is time-consuming, costly, and demands consistent expert annotation. This technical note introduces a semi-automated method to significantly reduce this annotation burden. The proposed approach utilizes two general-purpose vision-transformer-to-caption (GP-ViTC) models to generate descriptive text from images. These captions are then processed by a custom-developed semantic classifier (SC), which requires only minimal training to predict the correct image class. This GP-ViTC + SC system demonstrated exemplary classification rates in test cases and can subsequently be used to automatically annotate large image datasets. While the inference speed of the GP-ViTC models is not suited for real-time applications (approximately 10 s per image), this method substantially lessens the labor and expertise required for dataset creation, thereby facilitating the development of new, high-speed, custom AI vision models for niche applications. This work details the approach and its successful application, offering a cost-effective pathway for generating tailored image training sets. Full article

The accurate delineation of shrub crown boundaries is critical for ecological monitoring, land management, and understanding vegetation dynamics in fragile ecosystems such as semi-arid shrublands. While traditional image processing techniques often struggle with overlapping canopies, deep learning methods, such as convolutional neural networks (CNNs), offer promising solutions for precise segmentation. This study employed high-resolution imagery captured by unmanned aircraft systems (UASs) throughout the shrub growing season and explored the effectiveness of transfer learning for both semantic segmentation (Attention U-Net) and instance segmentation (Mask R-CNN). It utilized pre-trained model weights from two previous studies that originally focused on tree crown delineation to improve shrub crown segmentation in non-forested areas. Results showed that transfer learning alone did not achieve satisfactory performance due to differences in object characteristics and environmental conditions. However, fine-tuning the pre-trained models by unfreezing additional layers improved segmentation accuracy by around 30%. Fine-tuned pre-trained models show limited sensitivity to shrubs in the early growing season (April to June) and improved performance when shrub crowns become more spectrally unique in late summer (July to September). These findings highlight the value of combining pre-trained models with targeted fine-tuning to enhance model adaptability in complex remote sensing environments. The proposed framework demonstrates a scalable solution for ecological monitoring in data-scarce regions, supporting informed land management decisions and advancing the use of deep learning for long-term environmental monitoring. Full article

Meningiomas are the most common primary intracranial tumors in adults. For most cases, surgical resection is effective in mitigating recurrence risk. Accurate visualization of meningiomas helps radiologists assess the distribution and volume of the tumor within the brain while assisting neurosurgeons in preoperative planning. This paper introduces an innovative realistic 3D medical visualization system, namely SegR3D. It incorporates a 3D medical image segmentation pipeline, which preprocesses the data via semi-supervised learning-based multi-target segmentation to generate masks of the lesion areas. Subsequently, both the original medical images and segmentation masks are utilized as non-scalar volume data inputs into the realistic rendering pipeline. We propose a novel importance transfer function, assigning varying degrees of importance to different mask values to emphasize the areas of interest. Our rendering pipeline integrates physically based rendering with advanced illumination techniques to enhance the depiction of the structural characteristics and shapes of lesion areas. We conducted a user study involving medical practitioners to evaluate the effectiveness of SegR3D. Our experimental results indicate that SegR3D demonstrates superior efficacy in the visual analysis of meningiomas compared to conventional visualization methods. Full article

Remaining useful life (RUL) estimation of a bearing is a methodology to monitor rolling bearings for a system’s performance and reliability. It predicts the exact residual time without operational interruptions until complete bearing failure by training a deep learning model to predict the remaining time of working using extracted signal features. Extracting features is one of the most important subjects since its quality directly influences the performance of predicting RUL. Features should gradually and consistently increase over time and capture sudden deterioration within normalized specific thresholds. However, recent studies have not addressed feature extraction methods that consider all of these aspects. Moreover, some bearings exhibit a “self-healing” phenomenon, in which bearing conditions appear to temporarily improve, and this complicates the accurate representation of consistent performance degradation. However, very few studies have properly addressed this issue. Meanwhile, transfer learning is frequently used when training the RUL deep learning model because there is a lack of data for run-to-failure experiments. Most RUL estimation methodologies pre-train and apply deep learning models with supervised learning. But supervised transfer learning supposes that researchers already have access to end-of-life (EOL) data—often unavailable in industrial settings—limiting their practicality. To address these challenges, this paper proposes a novel semi-supervised transfer learning methodology that integrates an anti-self-healing health indicator (ASH-HI) with a transformer-based architecture. ASH-HI is a health indicator that quantifies the power spectrum density (PSD) difference between normal and abnormal states using skewness-based parameter selection, eliminating the need for manual parameter tuning. Also, it overcomes the self-healing problem by measuring the difference not only between normal and abnormal states but also between “correction” and abnormal states. Also, this paper presents a new semi-supervised transfer learning method without EOL information. The proposed methodology is validated using the PHM 2012, NASA IMS, and an experimental setup. This study is the first to attempt transfer learning using more than three datasets simultaneously, resulting in significantly improved performance. Full article

Ground penetrating radar (GPR) is an effective and efficient nondestructive testing technology for subsurface investigations. Deep learning-based methods have been successfully used in automatic underground target recognition. However, these methods are mostly based on supervised learning, requiring large amounts of labeled training data to guarantee high accuracy and generalization ability, which is a challenge in GPR fields due to time-consuming and labor-intensive data annotation work. To alleviate the demand for abundant labeled data, a semi-supervised deep learning method named attention–temporal ensembling (Attention-TE) is proposed for underground target recognition using GPR B-scan images. This method integrates a semi-supervised temporal ensembling architecture with a triplet attention module to enhance the classification performance. Experimental results of laboratory and field data demonstrate that the proposed method can automatically recognize underground targets with an average accuracy of above 90% using less than 30% of labeled data in the training dataset. Ablation experimental results verify the efficiency of the triplet attention module. Moreover, comparative experimental results validate that the proposed Attention-TE algorithm outperforms the supervised method based on transfer learning and four semi-supervised state-of-the-art methods. Full article

Timely identification of crop conditions is relevant for informed decision-making in precision agriculture. The initial step in determining the conditions that crops require involves isolating the components that constitute them, including the leaves and fruits of the plants. An alternative method for conducting this separation is to utilize intelligent digital image processing, wherein plant elements are labeled for subsequent analysis. The application of Deep Learning algorithms offers an alternative approach for conducting segmentation tasks on images obtained from complex environments with intricate patterns that pose challenges for separation. One such application is semantic segmentation, which involves assigning a label to each pixel in the processed image. This task is accomplished through training various models of Convolutional Neural Networks. This paper presents a comparative analysis of semantic segmentation performance using a convolutional neural network model with different backbone architectures. The task focuses on pixel-wise classification into three categories: leaves, fruits, and background, based on images of semi-hydroponic tomato crops captured in greenhouse settings. The main contribution lies in identifying the most efficient backbone-UNet combination for segmenting tomato plant leaves and fruits under uncontrolled conditions of lighting and background during image acquisition. The Convolutional Neural Network model UNet is is implemented with different backbones to use transfer learning to take advantage of the knowledge acquired by other models such as MobileNet, VanillaNet, MVanillaNet, ResNet, VGGNet trained with the ImageNet dataset, in order to segment the leaves and fruits of tomato plants. Highest percentage performance across five metrics for tomato plant fruit and leaves segmentation is the MVanillaNet-UNet and VGGNet-UNet combination with $0.88089$ and $0.89078$ respectively. A comparison of the best results of semantic segmentation versus those obtained with a color-dominant segmentation method optimized with a greedy algorithm is presented. Full article

Cytochrome P450 enzymes (CYPs) are versatile heme-containing monooxygenases involved in the metabolism of endogenous and exogenous compounds, as well as natural product biosynthesis. Their ability to catalyze regio- and stereoselective oxidation reactions makes them valuable in pharmaceuticals, fine chemicals, and biocatalysis. However, wild-type CYPs suffer from low catalytic efficiency, limited substrate specificity, and instability under industrial conditions. Recent advances in protein engineering—rational design, semi-rational design, and directed evolution—have enhanced their activity, stability, and substrate scope. These strategies have enabled CYPs to be engineered for applications like C–H functionalization, carbene transfer, and complex molecule biosynthesis. Despite progress, challenges remain in optimizing efficiency, expanding substrate ranges, and scaling production for industrial use. Future directions include integrating CYPs with other biocatalysts, improving high-throughput screening, and applying machine learning to enzyme design. This review highlights recent developments and the promising future of engineered CYPs in sustainable chemistry, drug development, and high-value chemical production. Full article

At the beginning of human history, surface water, especially from rivers and springs, was the most frequent water supply source. Groundwater was used in arid and semi-arid regions, e.g., eastern Crete (Greece). As the population increased, periodic water shortages occurred, which led to the development of sophisticated hydraulic structures for water transfer and for the collection and storage of rainwater, as seen, for example, in Early Minoan times (ca 3200–2100 BC). Water supply and urban planning had always been essentially related: the urban water supply systems that existed in Greece since the Bronze Age (ca 3200–1100 BC) were notably advanced, well organized, and operable. Water supply systems evolved considerably during the Classical and Hellenistic periods (ca 480–31 BC) and during the Roman period (ca 31 BC–480 AD). Also, early Indian society was an amazing vanguard of technology, planning, and vision, which significantly impacted India’s architectural and cultural heritage, thus laying the foundation for sustainable urban living and water resource management. In ancient Egypt, the main source of freshwater was the Nile River; Nile water was conveyed by open and closed canals to supply water to cities, temples, and fields. Underground stone-built aqueducts supplied Nile water to so-called Nile chambers in temples. The evolution of water supply and urban planning approaches from ancient simple systems to complex modern networks demonstrates the ingenuity and resilience of human communities. Many lessons can be learned from studying traditional water supply systems, which could be re-considered for today’s urban sustainable development. By digging into history, measures for overcoming modern problems can be found. Rainwater harvesting, establishing settlements in proximity of water sources to facilitate access to water, planning, and adequate drainage facilities were the characteristics of ancient civilizations since the ancient Egyptian, Minoan, Mohenjo-Daro, Mesopotamian, and Roman eras, which can still be adopted for sustainability. This paper presents significant lessons on water supply around the world from ancient times to the present. This diachronic survey attempts to provide hydro-technology governance for the present and future. Full article

Complex dynamics and nonlinear systems play a critical role in industrial processes, where complex interactions, high uncertainty, and external disturbances can significantly impact efficiency, stability, and safety. In sectors such as mining, manufacturing, and energy networks, even small perturbations can lead to unexpected system behaviors, operational inefficiencies, or cascading failures. Understanding and controlling these dynamics is essential for developing robust, adaptive, and resilient industrial systems. This study conducts a systematic literature review covering 2015–2025 in Scopus and Web of Science, initially retrieving 2628 (Scopus) and 343 (WoS) articles. After automated filtering (Python) and applying inclusion/exclusion criteria, a refined dataset of 2900 references was obtained, from which 89 highly relevant studies were selected. The literature was categorized into six key areas: (i) heat transfer with magnetized fluids, (ii) nonlinear control, (iii) big-data-driven optimization, (iv) energy transition via SOEC, (v) fault detection in control valves, and (vi) stochastic modeling with semi-Markov switching. Findings highlight the convergence of robust control, machine learning, IoT, and Industry 4.0 methodologies in tackling industrial challenges. Cybersecurity and sustainability also emerge as critical factors in developing resilient models, alongside barriers such as limited data availability, platform heterogeneity, and interoperability gaps. Future research should integrate multiscale analysis, deterministic chaos, and deep learning to enhance the adaptability, security, and efficiency of industrial operations in high-complexity environments. Full article

A substantial amount of the literature has examined the impact of participation in sport-based youth development programming and its global contribution to the lives of young people. In a similar vein, the outcomes of sport-based leadership programs are heavily influenced by the relationships and life skills acquired. One often overlooked demographic in this literature is incarcerated youth, a unique population who’s time spent in juvenile justice is fundamentally designed to prioritize rehabilitation and development. This paper sought to understand youths’ experiences in a sport-based leadership prison program with regards to content, relationship building, and transfer. This study included semi-structured interviews with three, currently incarcerated, adolescent black males, exploring their experiences as participants in their sport leadership program. Results included themes around the program meaning, relationship enhancers, and life skill learnings. Findings explore how sport-based prison programs may consider the importance of physical and psychological safety, relationship building, and life skill teachings as crucial components of a program that remain with participants well into their reintegration within society. Full article

Internet of Things (IoT) networks’ wide range and heterogeneity make them prone to cyberattacks. Most IoT devices have limited resource capabilities (e.g., memory capacity, processing power, and energy consumption) to function as conventional intrusion detection systems (IDSs). Researchers have applied many approaches to lightweight IDSs, including energy-based IDSs, machine learning/deep learning (ML/DL)-based IDSs, and federated learning (FL)-based IDSs. FL has become a promising solution for IDSs in IoT networks because it reduces the overhead in the learning process by engaging IoT devices during the training process. Three FL architectures are used to tackle the IDSs in IoT networks, including centralized (client–server), decentralized (device-to-device), and semi-decentralized. However, none of them has solved the heterogeneity of IoT devices while considering lightweight-ness and performance at the same time. Therefore, we propose a semi-decentralized FL-based model for a lightweight IDS to fit the IoT device capabilities. The proposed model is based on clustering the IoT devices—FL clients—and assigning a cluster head to each cluster that acts on behalf of FL clients. Consequently, the number of IoT devices that communicate with the server is reduced, helping to reduce the communication overhead. Moreover, clustering helps in improving the aggregation process as each cluster sends the average model’s weights to the server for aggregation in one FL round. The distributed denial-of-service (DDoS) attack is the main concern in our IDS model, since it easily occurs in IoT devices with limited resource capabilities. The proposed model is configured with three deep learning techniques—LSTM, BiLSTM, and WGAN—using the CICIoT2023 dataset. The experimental results show that the BiLSTM achieves better performance and is suitable for resource-constrained IoT devices based on model size. We test the pre-trained semi-decentralized FL-based model on three datasets—BoT-IoT, WUSTL-IIoT-2021, and Edge-IIoTset—and the results show that our model has the highest performance in most classes, particularly for DDoS attacks. Full article