Hotspot Detection Research Articles

DMFR-YOLO: an infrared small hotspot detection algorithm based on double multi-scale feature fusion

Abstract Hotspot automatic detection is an effective strategy to realize intelligent maintenance of photovoltaic (PV) modules. However, it is challenging to detect small hotspots accurately using deep learning-based object detection methods due to the weak information and small area of the target. This study presents a double multi-scale feature reconstruction (DMFR)-YOLO to detect small hotspots in PV modules. In the proposed DMFR-YOLO, the backbone of YOLOv8n is improved by eliminating redundant deep layers and introducing a shallow detection head to enhance its perception and recognition ability for small objects. Moreover, a weighted multi-layer feature reconstruction (MLFR) module is introduced to fuse the features of different depths and a multi-receptive field feature reconstruction (MRFR) module is designed to fuse the information of different receptive fields. Finally, ablation and comparison experiments are conducted to evaluate the performance of the proposed method. The model with both MLFR and MRFR can achieve a recall rate of 88.6%, with a mean average precision (mAP@0.5) of 93.3%. Experimental results demonstrate that the proposed DMFR-YOLO is able to achieve small hotspot detection in IR images with high accuracy.

Quantitative Estimation and Analysis of Spatiotemporal Delay Effects in Expressway Traffic Accidents

Expressway traffic accidents often result in severe congestion, with their unpredictable nature complicating timely and effective response measures. This paper presents a comprehensive method for accurately estimating and analyzing the spatiotemporal delay effects of expressway accidents through the integration of multi-source geographic data. The innovation lies in utilizing real-world vehicle trajectory data, combined with a Traffic Performance Index (TPI), to quantitatively assess delay impacts. By applying spatial clustering and hotspot detection techniques, we investigate the distribution patterns of delays and further employ a Spatial Error Model (SEM) to examine the relationships between accident characteristics and associated delay effects. Using expressway accident data and vehicle trajectory records from Hunan Province, the results demonstrate that the TPI-based approach effectively captures the duration, extent, and severity of traffic delays. Moreover, significant correlations are identified between delay impacts and specific accident characteristics, such as accident type, road type, road environment, pre-accident vehicle speed, and secondary accidents. This approach provides traffic management authorities with actionable insights into the overall roadway impact, facilitating targeted emergency response strategies and informing road usage policies tailored to the characteristics of accident impacts, thus helping to mitigate future risks.

Riemannian Manifold-Aided Data-driven Diagnosis of Photovoltaic Hot Spots

Abstract This paper proposes a novel data-driven approach for hot-spot fault detection in photovoltaic (PV) modules, utilizing a curved Riemannian manifold (RM) to characterize the sample space. The proposed method detects hot spots in PV systems by mining geometric features in high-dimensional data. The RM-based method combines the geometric properties of Riemannian manifolds for fault detection. In addition, it also provides information about the severity of the hot spots. The proposed method has three main advantages:
1) it is capable of actively learning fault characteristics and can be applied to diagnose various degrees of hot spots;
2) unlike other data-driven methods, the proposed method considers non-Euclidean spatial data features, which further improves the accuracy of the method and reduces the false alarm rate (FAR);
3) it does not rely on mathematical models or expert knowledge, making it able to meet the real-time monitoring needs of actual PV systems.
The effectiveness and superiority of the proposed approach are verified through 12 sets of hot spots tests conducted on a PV experimental platform.

Advanced ERT techniques for methane potential evaluation in controlled dump sites: A forward modeling approach

Methane (CH4) emissions from large controlled dump sites play a significant role in global warming. However, when captured efficiently, CH4 is a valuable energy resource. This study optimizes electrical resistivity tomography (ERT) for improved CH4 detection in controlled dump sites, focusing on refining electrode array configurations. We assessed three configurations: dipole-dipole, Wenner-Schlumberger (WS), and Wenner arrays. The results were compared with synthetic data simulated using the forward modeling technique, and the model's accuracy was evaluated using the Nash-Sutcliffe model efficiency coefficient (NSE), model sensitivity, and root mean square error (RMSE). Additionally, we analyzed the model's correlation with CH4 flux measurements. The WS array demonstrated the best performance, achieving a lower RMSE (23.4 %), a higher NES (0.90), and model sensitivity (0.90), along with a moderate negative correlation (-0.50) with CH4 flux measurements. This configuration effectively identified subsurface features, such as organic waste layers and leachate zones, which are crucial for CH₄ capture and landfill gas collection. Forward modeling confirmed the WS array's ability to provide high-resolution subsurface imaging, enhancing CH₄ hotspot detection. By improving the spatial and temporal accuracy of CH₄ detection, this approach addresses limitations in traditional methods and supports the design of more efficient gas capture systems. Applied in Thailand, these findings highlight the potential of optimized ERT techniques for enhancing CH₄ recovery in waste-to-energy systems, contributing to more sustainable waste management practices and reducing greenhouse gas emissions.

Rancang Bangun Sistem Pendeteksi Kebakaran Hutan Menggunakan Drone Berbasis Computer Vision

The danger of forest and land fires in Indonesia is a serious threat that continues to recurevery year. According to U.S. data Fire Service, there are more than 700 forest fires everyyear burning more than 7 million hectares of land. These fires not only damage ecosystems,but also cause the loss of important biodiversity. Many of Indonesia's endemic flora andfauna have fallen victim, increasing the risk of extinction. The use of UAVs can also changedisaster management. This technology can be integrated with early warning systems basedon analytical data and artificial intelligence, increasing the accuracy of fire predictions andmore proactive responses. Thus, the application of this technology is not just a temporarysolution, but a long-term strategic step to maintain environmental sustainability andcommunity welfare. Technological solutions such as UAVs and cooperation betweenvarious parties are needed to overcome this challenge effectively. The aim of this researchis to create a computer vision-based system that uses deep learning algorithms, especiallyYou Only Look Once (YOLO), to detect forest fires quickly and accurately. In mitigating forest fire disasters, early detection of hotspots is an important step to prevent fires fromspreading to other areas and reduce the damage caused by fires to humans and theenvironment. To achieve the objectives of this study, several actions were taken. First,forest fire image data is collected and processed for use in training the YOLO model. Next,the model is trained using a dataset that covers various forest fire conditions, such as fireintensity, time of day, and weather conditions, to ensure that the model can find firehotspots in various conditions.

Bayesian multivariate spatiotemporal statistical modeling of bus and taxi ridership

Statistical modeling of ridership over both space and time provides valuable insights on transportation planning and policies. Existing spatiotemporal studies, however, predominantly focus on analyzing a single type rather than multiple types of ridership, thus cannot leverage the correlation between different types of ridership. This study proposes a Bayesian multivariate spatiotemporal statistical model to jointly analyze multiple ridership over time. Specifically, the model accounts for correlation between multiple ridership based on different assumptions of space-time interactions (i.e., departures from the main spatial and temporal patterns) between different types of ridership as well as if covariates are included in the model. Using hourly bus and taxi ridership in the city of Wuhu, China as an example, the case study indicates that accounting for the correlation between the space-time interactions of each ridership, beyond the correlation between the main spatial patterns of the two ridership, further improves the statistical inferences of ridership modeling. In addition, the proposed approach enables the detection of spatial and spatiotemporal hotspots of each ridership as well as bus-taxi ratio hotspots using posterior probabilities. It also supports visual inspections regarding how the inclusion of covariates explains these hotspots. The proposed approach not only advances multivariate spatiotemporal statistical modeling of ridership, but can also provide useful insights on space- and time-specific transport policies at a granular resolution.

Assessment of VOCs emission inventory in Seoul through spatiotemporal observations using passive and online PAMS measurements

We have conducted a quantitative analysis of volatile organic compounds (VOCs) in Seoul from September 2022 to June 2023, employing passive air samplers (PAS) at 25 sites and continuous monitoring at five photochemical assessment monitoring stations (PAMS). Comparing VOCs concentrations and compositions from PAS and PAMS during the same time periods, we found that most VOCs showed no significant difference (p-value >0.05) between the methods, except for olefins, which were overestimated by 24 % in passive sampling, confirming PAS as an effective tool for assessing the spatial distribution of VOCs species over large areas. Throughout Seoul, median values of total volatile organic compounds (TVOCs) concentrations collected with passive samplers remained stable with a standard deviation of 1.22 ppbv, typically ranging from 11 to 13 ppbv, except for an increase during winter in the southwestern regions of the city, where intense industrial and vehicular emissions are reported in the current Clean Air Policy Support System (CAPSS) emissions inventory. Positive matrix factorization (PMF) analysis using the passive data revealed general consistency in the spatial distribution of area and road emissions when compared with CAPSS emission inventory, although localized discrepancies were observed. In quantitative assessments of TVOCs comparing PAS observations with emission-based models, modeled values were within 1.5 times the interquartile range of observed PAS concentrations over Seoul. However, the lack of detection of emission hotspots in southeastern Seoul in the PAS data, coupled with the omission of photochemical loss in our dispersion-only models, indicates that the current CAPSS emission inventory may significantly underestimate actual ambient VOCs levels, especially in summer.

Clinical and Technical Validation of OncoIndx® Assay-A Comprehensive Genome Profiling Assay for Pan-Cancer Investigations.

Comprehensive next-generation sequencing (NGS) assays enable the identification of clinically relevant mutations, enhancing the capability for targeted therapeutic interventions. In addition, genomic alterations driving the oncogenic roadmap and leading to resistance mechanisms are reshaping precision oncology. We report the workflow and clinical and technical validation of the OncoIndx® NGS platform-a comprehensive genomic profiling (CGP)-based assay for pan-cancer investigation. We evaluated the concordance between the OncoIndx® test findings and clinically established hotspot detection using SeraSeq reference standards. OncoIndx is a hybridization capture-based NGS assay for the targeted deep sequencing of all exons and selected introns of 1080 cancer-related genes. We show the outcome in the form of tier I and tier II single nucleotide variants (SNVs), copy number alterations (CNAs), and specific gene fusions. OncoIndx® also informs genome-wide tumor mutational burden (TMB), microsatellite instability (MSI), homologous recombination deficiency (HRD), and genomic loss of heterozygosity (gLOH). A total of 63 samples were utilized for validation with reference standards, clinical samples, and orthogonal assessment for genomic alterations. In addition, 49 cross-laboratory samples were validated for microsatellite instability (MSI), and for the tumor mutation burden (TMB), 18 samples as reference standards, 6 cross-laboratory samples, and 29 TCGA samples were utilized. We show a maximum clinical sensitivity of 98% and a positive predictive value (PPV) of 100% for the clinically actionable genomic variants detected by the assay. In addition, we demonstrate analytical validation with the performance of the assay, limit of detection (LoD), precision, and orthogonal concordance for various types of SVs, CNAs, genomic rearrangements, and complex biomarkers like TMB, MSI, and HRD. The assay offers reliable genomic predictions with the high-precision detection of actionable variants, validated by established reference standards.

Smart hotspot detection using geospatial artificial intelligence: A machine learning approach to reduce flood risk

This study employs Geospatial Artificial Intelligence (GeoAI) and the Random Forest Machine Learning (ML) algorithm to enhance flood hazard assessments in Portugal. It utilizes NASA's LP DAAC (2023) Digital Elevation Model (DEM) and slope data from EPIC WEBGIS PORTUGAL DATA, offering detailed topographical insights for environmental planning. Additionally, it incorporates data on proximity to water bodies from the Portuguese Environment Agency and the European Environment Agency, and soil characteristics from EPIC WEBGIS PORTUGAL DATA, facilitating a thorough examination of flood risks. This approach prioritizes long-term land features over short-term weather patterns, providing a comprehensive understanding of flood vulnerability. The study processes data at a 1 km x 1 km resolution, adapting TIFF maps for compatibility with the Random Forest model. The produced flood hazard maps identify potential flood hotspots at both national and city levels, crucial for urban planning. These maps aid in assessing the vulnerability of key infrastructure and assets, such as transport networks and buildings. The research highlights the importance of integrating additional data on assets and socioeconomic factors to enhance urban resilience. It sets the stage for future research aimed at improving predictive accuracy and underscores the necessity of extensive geospatial analytics in managing infrastructure risks.

Avian Influenza: Lessons from Past Outbreaks and an Inventory of Data Sources, Mathematical and AI Models, and Early Warning Systems for Forecasting and Hotspot Detection to Tackle Ongoing Outbreaks.

The ongoing avian influenza (H5N1) outbreak, one of the most widespread and persistent in recent history, has significantly impacted public health and the poultry and dairy cattle industries. This review covers lessons from past outbreaks, risk factors for transmission, molecular epidemiology, clinical features, surveillance strategies, and socioeconomic impacts. Since 1997, H5N1 has infected over 900 individuals globally, with a fatality rate exceeding 50%. Key factors influencing infection rates include demographic, socioeconomic, environmental, and ecological variables. The virus's potential for sustained human-to-human transmission remains a concern. The current outbreak, marked by new viral clades, has complicated containment efforts. This review discusses how to integrate technological advances, such as mathematical modeling and artificial intelligence (AI), to improve forecasting, hotspot detection, and early warning systems. We provide inventories of data sources, covering both conventional and unconventional data streams, as well as those of mathematical and AI models, which can be vital for comprehensive surveillance and outbreak responses. In conclusion, integrating AI, mathematical models, and technological innovations into a One-Health approach is essential for improving surveillance, forecasting, and response strategies to mitigate the impacts of the ongoing avian influenza outbreak. Strengthening international collaboration and biosecurity measures will be pivotal in controlling future outbreaks and protecting both human and animal populations from this evolving global threat.

Deep Learning-Driven Thermal Imaging Hotspot Detection in Solar Photovoltaic Arrays Using YOLOv10

The effective management of solar photovoltaic (PV) arrays is vital to maximising energy generation and ensuring long-term performance reliability. The presence of faults, or partial shading, can give rise to hotspots in PV arrays, making their detection critical for timely maintenance and avoiding further impacts on their performance. This paper describes a new approach to the hotspot detection issue in thermal images of solar PV arrays by leveraging deep learning. This study employs the YOLOv10 (You Only Look Once, version 10) model to deliver very high accuracy and speed in identifying and localising hotspots under defined regions of interest (ROIs). The proposed method begins with taking thermal images from multiple solar PV installations while capturing a range of operational conditions and fault types. These images are annotated with hotspot regions to create a robust training dataset. Given YOLO's capability for real-time object detection with high precision and mean average precision (mAP), the YOLOv10 model is then implemented to train on this dataset. Multiple changes were made to the YOLOv10 hyperparameters to optimise them for detecting hotspots in thermal images. The experimental scenario of this paper demonstrates that this approach indeed performs significantly better than standard image processing methods as well as prior deep learning models for detecting both hotspot accuracy as well as speed of processing the images. The YOLOv10 method demonstrated the highest available classification performance with a mAP at intersection over union (IoU) of 0.5 accuracy of 0.91 with inference time suitable for real-time applications. The sample results demonstrated the model's continuous ability to detect hotspots and provide location data under various conditions, such as crossing through different times of day and weather. The results of this study demonstrate that YOLOv10 can be used for improved detection of hotspots in the explicably thermal imagery of solar PV arrays.

Lightweight Hotspot Detection Model Fusing SE and ECA Mechanisms.

In this paper, we propose a lightweight lithography machine learning-based hotspot detection model that integrates the Squeeze-and-Excitation (SE) attention mechanism and the Efficient Channel Attention (ECA) mechanism. These mechanisms can adaptively adjust channel weights, significantly enhancing the model's ability to extract relevant features of hotspots and non-hotspots through cross-channel interaction without dimensionality reduction. Our model extracts feature vectors through seven convolutional layers and four pooling layers, followed by three fully connected layers that map to the output, thereby simplifying the CNN network structure. Experimental results on our collected layout dataset and the ICCAD 2012 layout dataset demonstrate that our model is more lightweight. By evaluating overall accuracy, recall, and runtime, the comprehensive performance of our model is shown to exceed that of ConvNeXt, Swin transformer, and ResNet 50.

Electrothermal Modeling of Photovoltaic Modules for the Detection of Hot-Spots Caused by Soiling

Solar energy plays a major role in the transition to renewable energy. To ensure that large-scale photovoltaic (PV) power plants operate at their full potential, their monitoring is essential. It is common practice to utilize drones equipped with infrared thermography (IRT) cameras to detect defects in modules, as the latter can lead to deviating thermal behavior. However, IRT images can also show temperature hot-spots caused by inhomogeneous soiling on the module’s surface. Hence, the method does not differentiate between defective and soiled modules, which may cause false identification and economic and resource loss when replacing soiled but intact modules. To avoid this, we propose to detect spatially inhomogeneous soiling losses and model temperature variations explained by soiling. The spatially resolved soiling information can be obtained, for example, using aerial images captured with ordinary RGB cameras during drone flights. This paper presents an electrothermal model that translates the spatially resolved soiling losses of PV modules into temperature maps. By comparing such temperature maps with IRT images, it can be determined whether the module is soiled or defective. The proposed solution consists of an electrical model and a thermal model which influence each other. The electrical model of Bishop is used which is based on the single-diode model and replicates the power output or consumption of each cell, whereas the thermal model calculates the individual cell temperatures. Both models consider the given soiling and weather conditions. The developed model is capable of calculating the module temperature for a variety of different weather conditions. Furthermore, the model is capable of predicting which soiling pattern can cause critical hot-spots.

Enhancing PV panel diagnosis: RT-DETR vs YOLO v8 for real-time hotspots detection

Enhancing PV panel diagnosis: RT-DETR vs YOLO v8 for real-time hotspots detection

AI-powered VM selection: Amplifying cloud performance with dragonfly algorithm

The convenience and cost-effectiveness offered by cloud computing have attracted a large customer base. In a cloud environment, the inclusion of the concept of virtualization requires careful management of resource utilization and energy consumption. With a rapidly increasing consumer base of cloud data centers, it faces an overwhelming influx of Virtual Machine (VM) requests. In cloud computing technology, the mapping of these requests onto the actual cloud hardware is known as VM placement which is a significant area of research. The article presents the Dragonfly Algorithm integrated with Modified Best Fit Decreasing (DA-MBFD) is proposed to minimize the overall power consumption and the migration count. DA-MBFD uses MBFD for ranking VMs based on their resource requirement, then uses the Minimization of Migration (MM) algorithm for hotspot detection followed by DA to optimize the replacement of VMs from the overutilized hosts. DA-MBFD is compared with a few of the other existing techniques to show its efficiency. The comparative analysis of DA-MBFD against E-ABC, E-MBFD, and MBFD-MM shows %improvement reflecting a significant reduction in power consumption 8.21 %, 8.6 %, 6.77 %, violations in service level agreement from 9.25 %, 6.98 %–7.86 % and number of migrations 6.65 %, 8.92 %, 7.02 %, respectively.

A general and accurate pattern search method for various scenarios

As chip designs grow in complexity, the optical proximity correction (OPC) process becomes increasingly time-consuming. As a result, pattern search technology is becoming a foundation stone of many tasks for manufacturing, such as lithography simulation, hotspot detection, mask optimization, and so on. The most difficult challenge of pattern search is to locate clips by specific patterns within the layout accurately and efficiently. In this paper, we present a generalized pattern search method capable in diverse scenarios, including patterns with hollow shapes, shifting edges, and multi-layer situations. Experimental results show that our method outperforms commercial tools in pattern locating accuracy and handling search problems involving complex patterns which is not directly support by commercial tools yet.

A lightweight joint metric detection approach on YOLO for hot spots in photovoltaic modules

The hot spot effect is one of the primary causes of damage to photovoltaic (PV) modules and a significant factor contributing to the decline in their power generation capacity. Thermal imaging inspection of PV modules is an indispensable aspect of PV plant operation and maintenance. This article introduces a lightweight detection algorithm for hot spots in PV modules based on an enhanced version of YOLOv8. The algorithm incorporates a lightweight convolution operator, Volumetric Grid Spatial Cross Stage Partial, as a replacement for the original C2f operator, resulting in effective reduction of the model's parameter size and improved detection speed. Additionally, the Content-Aware ReAssembly of FEatures upsampling operator is utilized instead of the nearest-neighbor algorithm during the upsampling process, thereby minimizing the loss of hot spot feature information. Furthermore, a joint metric approach combining the Complete Intersection over Union metric and the normalized Wasserstein distance metric is proposed to calculate the localization loss of the target bounding boxes, thereby enhancing the regression accuracy of small hot spot prediction boxes. Experimental results demonstrate that the improved YOLOv8 network enables fast and accurate detection of hot spots. This method achieves an impressive average precision of 98.8% and a high detection speed of 218.3 fps.

Detection of hotspots and coldspots of the incidence of hand, foot and mouth diseases: A case study in Ho Chi Minh city, Vietnam

Background: Hand, foot, and mouth disease (HFMD) is a common pediatric illness mostly caused by Coxsackievirus A16 (CV-A16) and Human Enterovirus 71 (EV-A71). HFMD has grown significantly over the last 20 years, and it has drawn a lot of attention. To better understand of the epidemiology of HFMD, this study aims to investigate the use of Getis-Ord’s 𝐺𝑖 ∗ statistic in the identification of hotspots and coldspots of HFMD in Ho Chi Minh city, Vietnam. Methods: Getis-Ord’s 𝐺𝑖 ∗ statistic was first applied to identify hotspots and coldspots of HFMD incidence including highhigh, low-low spatial clusters. HFMD cases and infection rates collected in Ho Chi Minh were then used to explore the spatial clusters of HFMD Finally, the main findings will be discussed and summarised. Results: It was found that the large number of hand, foot and mouth disease cases were mainly concentrated in the western area of Ho Chi Minh City. A total of 04 hotspot clusters and 05 coldspots of HFMD infection rate were detected in the 29th week of 2024 in the west, center and north of Ho Chi Minh City, respectively. Whereas, 06 hotspots and 04 coldspots of HFMD infection rate were detected in the 30th week of 2024 in the west, south and center of Ho Chi Minh City, respectively. Conclusions: The results of this investigation confirm the value of Getis-Ord's 𝐺𝑖 ∗ statistic in the detection of hotspots and coldspots associated with confirmed cases of HFMD.

Hotspot detection in the cultivation of organic winter wheat variety mixtures

Hotspot detection in the cultivation of organic winter wheat variety mixtures

Sága, a Deep Learning Spectral Analysis Tool for Fungal Detection in Grains-A Case Study to Detect Fusarium in Winter Wheat.

Fusarium head blight (FHB) is a plant disease caused by various species of the Fusarium fungus. One of the major concerns associated with Fusarium spp. is their ability to produce mycotoxins. Mycotoxin contamination in small grain cereals is a risk to human and animal health and leads to major economic losses. A reliable site-specific precise Fusarium spp. infection early warning model is, therefore, needed to ensure food and feed safety by the early detection of contamination hotspots, enabling effective and efficient fungicide applications, and providing FHB prevention management advice. Such precision farming techniques contribute to environmentally friendly production and sustainable agriculture. This study developed a predictive model, Sága, for on-site FHB detection in wheat using imaging spectroscopy and deep learning. Data were collected from an experimental field in 2021 including (1) an experimental field inoculated with Fusarium spp. (52.5 m × 3 m) and (2) a control field (52.5 m × 3 m) not inoculated with Fusarium spp. and sprayed with fungicides. Imaging spectroscopy data (hyperspectral images) were collected from both the experimental and control fields with the ground truth of Fusarium-infected ear and healthy ear, respectively. Deep learning approaches (pretrained YOLOv5 and DeepMAC on Global Wheat Head Detection (GWHD) dataset) were used to segment wheat ears and XGBoost was used to analyze the hyperspectral information related to the wheat ears and make predictions of Fusarium-infected wheat ear and healthy wheat ear. The results showed that deep learning methods can automatically detect and segment the ears of wheat by applying pretrained models. The predictive model can accurately detect infected areas in a wheat field, achieving mean accuracy and F1 scores exceeding 89%. The proposed model, Sága, could facilitate the early detection of Fusarium spp. to increase the fungicide use efficiency and limit mycotoxin contamination.