ABSTRACT Fault detection in district heating (DH) substations is critical for energy efficiency and reliability. However, it is challenged by scarce fault labels, low‐frequency data, privacy concerns, and battery‐constrained gateways. We propose a novel hybrid semi‐supervised federated domain adaptation architecture for fault detection in DH. We use a one‐class variational autoencoder (VAE) to leverage heterogeneous sensor streams from 434 distributed substations. First, we perform cross‐network unsupervised pre‐training on multi‐sourced data from two independent real‐world DH networks, fusing their return temperature dynamics into a robust shared manifold. Second, we leverage maintenance metadata to selectively allow verified‐normal clients for per‐round fine‐tuning of the model. Third, we drastically reduce uplink costs by compressing each client's weight delta using 10% top‐ k sparsification and demonstrate that our pipeline enables robust few‐shot finetuning with 20% of the normal operational data while retaining high detection performance. By strategically training, our method achieves F 1 and G‐mean scores of up to 97% and an AUC ≥ 99% on real‐world DH data. To our knowledge, this is the first work to study cross‐domain data fusion in the DH field for fault detection, aiming to enhance and enable effective, scalable, and energy‐efficient monitoring of substations.

Automatic fault detection and diagnosis (FDD) are critical for maintaining reliable and efficient industrial systems. However, conventional methods rely heavily on manual inspections or threshold-based techniques, which often fail to capture the dynamic patterns in time series (TS) sensor data. As a result, faults persist for extended periods, leading to suboptimal system operations, increased energy waste, and significant economic losses. This work proposes a cross-modal framework that facilitates the efficient deployment of state-of-the-art pretrained vision models for enhanced FDD, with two novel TS-to-image transformations: first, an adapter deep encoder that learns optimal, task-specific representations from raw sensor data while generating outputs that are input-compliant with pretrained models. Second, an enhanced line plot that creates geometric shapes of two related signals. Comparative experiments against fixed methods, including spectrograms, Gramian angular fields, Markov transition fields, recurrence plots, and five deep learning baseline models, showed substantial performance gains across diverse domains. InceptionTime achieved the highest average baseline performance with an F$_{1}$of 88.6%, while the adapter and shapes achieved 94.4% and 92.4%, respectively. The findings highlight the potential of the cross-modal framework for FDD to facilitate early intervention and efficient system maintenance in industrial settings.

Abstract This paper proposes SHIODEG, a hybrid metaheuristic that integrates the success-history intelligent optimizer (SHIO) with differential evolution (DE) and a Gaussian transformation (GT) to tackle two persistent challenges in optimization for engineering design: (i) the absence of a universally best optimizer across problem classes (as implied by the No-Free-Lunch perspective) and (ii) the limited ability of purely gradient-based methods to produce substantial improvements in complex, constrained, and often non-smooth real-world problems, motivating hybrid strategies that balance exploration and exploitation. SHIODEG follows a staged search process in which DE generates diverse trial solutions, GT injects normally distributed perturbations to reduce premature convergence and diversity collapse, and SHIO refines promising regions using success-history guidance from the best three leaders. SHIODEG is evaluated on the IEEE CEC2022 benchmark suite (12 functions) using 30 independent runs, a population size of 100, and a budget of 1000D function evaluations. The results show that SHIODEG consistently delivers top-tier performance across the benchmark suite, showing strong competitiveness, low variability, and statistically significant improvements over a wide range of alternative optimizers. It also demonstrates robust effectiveness on multiple constrained engineering design problems, achieving high-quality solutions across diverse real-world constraints.

This study introduces FLAME (Fault Localization using Augmented Model Enhancement), a novel fault diagnosis framework for District Heating (DH) substations. Automated Fault Detection and Diagnosis (FDD) has become imperative as many DH substations perform sub-optimal due to faults. The main challenges complicating accurate fault diagnosis are increasing operational complexities and a scarcity of labelled data. FLAME integrates a hybrid Convolutional Neural Network and Long Short-Term Memory (CNN-LSTM) model with an attention mechanism and introduces the Fault Augmentation Signature Technique (FAST). FAST overcomes the limitations of traditional stochastic data augmentation methods by leveraging pattern mixing of the time series. The FLAME framework uses transfer learning, initially trained on augmented data using FAST and fine-tuned using original substation data. Experimental results reveal that FLAME outperforms conventional methods, obtaining F1 scores of approximately 0.95 and 0.92 on lab-simulated and real-world datasets, respectively. Additionally, the research found the importance of the temperature difference measurement ( Δ T ) and median-based sampling strategies for optimal fault pattern identification. These findings establish FLAME as a new benchmark in DH system diagnostics, offering a robust framework to enhance fault diagnosis accuracy and operational efficiency of DH substations.

Abstract Large amount of data are generated from in-situ monitoring of additive manufacturing (AM) processes which is later used in prediction modelling for defect classification to speed up quality inspection of products. A high volume of this process data is defect-free (majority class) and a lower volume of this data has defects (minority class) which result in the class-imbalance issue. Using imbalanced datasets, classifiers often provide sub-optimal classification results, i.e. better performance on the majority class than the minority class. However, it is important for process engineers that models classify defects more accurately than the class with no defects since this is crucial for quality inspection. Hence, we address the class-imbalance issue in manufacturing process data to support in-situ quality control of additive manufactured components. For this, we propose cluster-based adaptive data augmentation (CADA) for oversampling to address the class-imbalance problem. Quantitative experiments are conducted to evaluate the performance of the proposed method and to compare with other selected oversampling methods using AM datasets from an aerospace industry and a publicly available casting manufacturing dataset. The results show that CADA outperformed random oversampling and the SMOTE method and is similar to random data augmentation and cluster-based oversampling. Furthermore, the results of the statistical significance test show that there is a significant difference between the studied methods. As such, the CADA method can be considered as an alternative method for oversampling to improve the performance of models on the minority class.

Abstract In this paper, we investigate the current state and development of personalized smart immersive extended reality environments (PSI-XR). PSI-XR has gained increasing traction across various fields such as education, entertainment, and healthcare, offering customized immersive experiences that address users’ personalized needs. This study performs a systematic literature review by collecting and analyzing related journal and conference papers in the domain. Following a comprehensive search across three databases, which yielded 1276 papers, a refined selection of 94 publications was made to conduct an in-depth analysis of cutting-edge research in the field of PSI-XR. This review focused on examining application domains, relevant technologies, and smart techniques, including artificial intelligence, with particular emphasis on advancements in personalization. The study provides insights into prospective advancements while also identifying the opportunities and challenges in this evolving field. This review is beneficial for both researchers and developers interested in exploring the state-of-the-art personalized perspective in a smart immersive extended reality environment.

BACKGROUND: Chronic obstructive pulmonary disease (COPD) is a severe condition affecting millions worldwide, leading to numerous annual deaths. The absence of significant symptoms in its early stages promotes high underdiagnosis rates for the affected people. Besides pulmonary function failure, another harmful problem of COPD is the systemic effects, e.g., heart failure or voice distortion. However, the systemic effects of COPD might provide valuable information for early detection. In other words, symptoms caused by systemic effects could be helpful to detect the condition in its early stages. OBJECTIVE: The proposed study aims to explore whether the voice features extracted from the vowel "a" utterance carry any information that can be predictive of COPD by employing Machine Learning (ML) on a newly collected voice dataset. METHODS: Forty-eight participants were recruited from the pool of research clinic visitors at Blekinge Institute of Technology (BTH) in Sweden between January 2022 and May 2023. A dataset consisting of 1246 recordings from 48 participants was gathered. The collection of voice recordings containing the vowel "a" utterance commenced following an information and consent meeting with each participant using the VoiceDiagnostic application. The collected voice data was subjected to silence segment removal, feature extraction of baseline acoustic features, and Mel Frequency Cepstrum Coefficients (MFCC). Sociodemographic data was also collected from the participants. Three ML models were investigated for the binary classification of COPD and healthy controls: Random Forest (RF), Support Vector Machine (SVM), and CatBoost (CB). A nested k-fold cross-validation approach was employed. Additionally, the hyperparameters were optimized using grid-search on each ML model. For best performance assessment, accuracy, F1-score, precision, and recall metrics were computed. Afterward, we further examined the best classifier by utilizing the Area Under the Curve (AUC), Average Precision (AP), and SHapley Additive exPlanations (SHAP) feature-importance measures. RESULTS: The classifiers RF, SVM, and CB achieved a maximum accuracy of 77 %, 69 %, and 78 % on the test set and 93 %, 78 % and 97 % on the validation set, respectively. The CB classifier outperformed RF and SVM. After further investigation of the best-performing classifier, CB demonstrated the highest performance, producing an AUC of 82 % and AP of 76 %. In addition to age and gender, the mean values of baseline acoustic and MFCC features demonstrate high importance and deterministic characteristics for classification performance in both test and validation sets, though in varied order. CONCLUSION: This study concludes that the utterance of vowel "a" recordings contain information that can be captured by the CatBoost classifier with high accuracy for the classification of COPD. Additionally, baseline acoustic and MFCC features, in conjunction with age and gender information, can be employed for classification purposes and benefit healthcare for decision support in COPD diagnosis. CLINICAL TRIAL REGISTRATION NUMBER: NCT05897944.

This paper aims to address data labelling issues in process data to support in-situ process monitoring of additive manufactured components. For this, we adopted an active learning (AL) approach to minimise the manual effort for data labelling for classification models. In this study, we present an approach that utilises pre-trained models to extract deep features from images, and clustering and query by committee sampling to select the representative samples to build defect classification models. We conduct quantitative experiments to evaluate the proposed method’s performance and compare it with other selected state-of-the-art AL approaches using a dataset of additive manufacturing (AM) and a publicly available dataset. The experimental results show that the proposed approach outperforms AL with committee based sampling, and AL with clustering and random sampling. The results of the statistical significance test show that there is a significant difference between the studied AL approaches. Hence, the proposed AL approach can be considered an alternative method to reduce labelling costs when building defects classification models, whose generalizability is most likely plausible.

Abstract This paper presents a digital image dataset of historical handwritten birth records stored in the archives of several parishes across Sweden, together with the corresponding metadata that supports the evaluation of document analysis algorithms’ performance. The dataset is called SHIBR (the Swedish Historical Birth Records). The contribution of this paper is twofold. First, we believe it is the first and the largest Swedish dataset of its kind provided as open access (15,000 high-resolution colour images of the era between 1800 and 1840). We also perform some data mining of the dataset to uncover some statistics and facts that might be of interest and use to genealogists. Second, we provide a comprehensive survey of contemporary datasets in the field that are open to the public along with a compact review of word spotting techniques. The word transcription file contains 17 columns of information pertaining to each image (e.g., child’s first name, birth date, date of baptism, father's first/last name, mother’s first/last name, death records, town, job title of the father/mother, etc.). Moreover, we evaluate some deep learning models, pre-trained on two other renowned datasets, for word spotting in SHIBR. However, our dataset proved challenging due to the unique handwriting style. Therefore, the dataset could also be used for competitions dedicated to a large set of document analysis problems, including word spotting.

Abstract This paper introduces a new image-based handwritten historical digit dataset named Arkiv Digital Sweden (ARDIS). The images in ARDIS dataset are extracted from 15,000 Swedish church records which were written by different priests with various handwriting styles in the nineteenth and twentieth centuries. The constructed dataset consists of three single-digit datasets and one-digit string dataset. The digit string dataset includes 10,000 samples in red–green–blue color space, whereas the other datasets contain 7600 single-digit images in different color spaces. An extensive analysis of machine learning methods on several digit datasets is carried out. Additionally, correlation between ARDIS and existing digit datasets Modified National Institute of Standards and Technology (MNIST) and US Postal Service (USPS) is investigated. Experimental results show that machine learning algorithms, including deep learning methods, provide low recognition accuracy as they face difficulties when trained on existing datasets and tested on ARDIS dataset. Accordingly, convolutional neural network trained on MNIST and USPS and tested on ARDIS provide the highest accuracies $$58.80\%$$ 58.80 % and $$35.44\%$$ 35.44 % , respectively. Consequently, the results reveal that machine learning methods trained on existing datasets can have difficulties to recognize digits effectively on our dataset which proves that ARDIS dataset has unique characteristics. This dataset is publicly available for the research community to further advance handwritten digit recognition algorithms.

OBJECTIVE: We provide an e-Science perspective on the workflow from risk factor discovery and classification of disease to evaluation of personalized intervention programs. As case studies, we use personalized prostate and breast cancer screenings. MATERIALS AND METHODS: We describe an e-Science initiative in Sweden, e-Science for Cancer Prevention and Control (eCPC), which supports biomarker discovery and offers decision support for personalized intervention strategies. The generic eCPC contribution is a workflow with 4 nodes applied iteratively, and the concept of e-Science signifies systematic use of tools from the mathematical, statistical, data, and computer sciences. RESULTS: The eCPC workflow is illustrated through 2 case studies. For prostate cancer, an in-house personalized screening tool, the Stockholm-3 model (S3M), is presented as an alternative to prostate-specific antigen testing alone. S3M is evaluated in a trial setting and plans for rollout in the population are discussed. For breast cancer, new biomarkers based on breast density and molecular profiles are developed and the US multicenter Women Informed to Screen Depending on Measures (WISDOM) trial is referred to for evaluation. While current eCPC data management uses a traditional data warehouse model, we discuss eCPC-developed features of a coherent data integration platform. DISCUSSION AND CONCLUSION: E-Science tools are a key part of an evidence-based process for personalized medicine. This paper provides a structured workflow from data and models to evaluation of new personalized intervention strategies. The importance of multidisciplinary collaboration is emphasized. Importantly, the generic concepts of the suggested eCPC workflow are transferrable to other disease domains, although each disease will require tailored solutions.

Abstract An important application in wireless networks is data collection. It aims to gather and deliver specific data for concerned authorities. Many researchers invest in vehicular ad hoc networks for that purpose to acquire data from different sources on the roads as from its vicinity. A vehicle is considered as a mobile data collector, it gathers real‐time or delay‐tolerant data such as road traffic, environmental information, and event advertisements. In a previous work, we have proposed a novel clustered data gathering protocol (CDGP) for vehicular ad hoc network, which improves the collection performance by implementing a new space division multiple access technique called dynamic space division multiple access and a retransmission mechanism in case of errors. However, CDGP supports only delay‐tolerant data as it does not use any aggregation technique. In this paper, we propose an enhancement of this protocol by extending it to support: (i) both real‐time and delay‐tolerant applications; (ii) multiple types of data; and (iii) aggregation of collected data prior to sending them to the initiator. We present the plausible analytical complexity of the extended CDGP, as we illustrate the superiority of its performance throughout the results obtained from simulation experiments, using a Freeway mobility model. Copyright © 2015 John Wiley & Sons, Ltd.

Since it was first presented in 2002, optical projection tomography (OPT) has emerged as a powerful tool for the study of biomedical specimen on the mm to cm scale. In this paper, we present computational tools to further improve OPT image acquisition and tomographic reconstruction. More specifically, these methods provide: semi-automatic and precise positioning of a sample at the axis of rotation and a fast and robust algorithm for determination of postalignment values throughout the specimen as compared to existing methods. These tools are easily integrated for use with current commercial OPT scanners and should also be possible to implement in "home made" or experimental setups for OPT imaging. They generally contribute to increase acquisition speed and quality of OPT data and thereby significantly simplify and improve a number of three-dimensional and quantitative OPT based assessments.

In this paper, the concept of object-oriented embedding (OOE) is introduced into information hiding in general and particularly to steganography which is the science that involves undetectable communication of secret data in an appropriate multimedia carrier. The proposal takes advantage of computer vision to orient the embedding process. Although, any existing algorithm can benefit from this technique to enhance its performance against steganalysis attacks, however this work also considers a new embedding algorithm in the wavelet domain using the Binary Reflected Gray Code (BRGC). In the realm of information hiding, one wing focuses on robustness, i.e., watermarking, and another wing focuses on imperceptibility, i.e., steganography. This work advocates for a new steganographic model that meets both robustness as well as imperceptibility. Resilience against common steganalysis attacks including the 274-D merged Markov and DCT features while surviving various image processing manipulations are reported. A neural network classifier was trained with features derived from 400 images. Comparisons with existing systems will also be highlighted.

This paper presents a novel dynamic threshold approach to discriminate skin pixels and non-skin pixels in color images. Fixed decision boundaries (or fixed threshold) classification approaches are successfully applied to segment human skin. These fixed thresholds mostly failed in two situations as they only search for a certain skin color range: 1) any non-skin object may be classified as skin if non-skin objects's color values belong to fixed threshold range. 2) any true skin may be mistakenly classified as non-skin if that skin color values do not belong to fixed threshold range. Therefore in this paper, instead of predefined fixed thresholds, novel online learned dynamic thresholds are used to overcome the above drawbacks. The experimental results show that our method is robust in overcoming these drawbacks.

The majority of existing methods have one thing in common which is the de-correlation of luminance from the considered colour channels. It is believed that the luminance is underestimated here since it is seen as the least contributing colour component to skin colour detection. This work questions this claim by showing that luminance can be useful in separating skin and non-skin clusters. To this end, this work uses a new colour space which contains error signals derived from differentiating the grayscale map and the non-encoded-red grayscale version. The advantages of this approach are the reduction of space dimensionality from 3D to 1D space and the construction of a rapid classifier necessary for real time applications. This method is meant to assist digital image steganography to orient the embedding process since skin information is deemed to be psycho-visually redundant.

Previous studies have shown that Steganography tools such as the S-Tools application outperforms counterpart tools in hiding data in the spatial domain. Another tool called F5 is identified as a robust tool in implementing Steganography in the frequency domain. This paper presents a Steganographic system which exploits the YCbCr colour space. YCbCr is intended to take advantage of human colour-response characteristics. Results show that our algorithm outperforms both F5 and S-Tools. Moreover, our system demonstrates improved performance in retrieving the hidden data after applying image processing attacks in the form of additive artificial noise. As a performance measurement for image distortion Peak-Signal-to-Noise Ratio (PSNR), which is classified under the difference distortion metrics, can be applied to the Stego images. This study also shows that by adopting an object oriented Steganography mechanism, in the sense that we track skin tone objects in video frames, we get a higher PSNR value.