Real-time Early Warning System Research Articles

Exploring an early warning system for maritime security risks: An approach based on compressed sensing

Piracy has posed intractable challenges to maritime security and the global economy. This study introduces a real-time early warning system designed to monitor piracy activity in the vicinity of merchant vessels, thereby facilitating the assessment and countermeasure of potential attacks. The approach employs frequency agile radar to identify pirate ships characterized by aggregation and jamming behavior. To address the computational difficulties caused by changing frequencies, compressed sensing with a tailored target information processing procedure is applied to reconstruct the radar observation scene. The developed target detection algorithm yields precise information about nearby pirate ships, enabling a comprehensive evaluation of the risk level and the subsequent recommendation of an appropriate security scheme. Extensive experiments have shown that the proposed early warning model can provide best warning performance compared to other benchmark models, with an average accuracy of 0.964, precision of 0.966, recall of 0.955, F1 score of 0.960.

Continuous ambulatory epilepsy detection system incorporating feature engineering

Epilepsy is a prevailing disease that affects people from different age brackets and demographic backgrounds. It leads to uncontrollable onset of seizures and can result in severe neurological injuries. In this paper, we devised a novel seizure prediction system as a real-time early warning system for patients. By using real-time transmissible, portable, and wireless devices, we can acquire raw data from scalp electroencephalogram (EEG) without any pre-processing for the input. After pre-processing, the data is fed into selected prediction algorithms based on literature review and a combination of methodologies. After times of iteration, our result shows a promising performance, with an accuracy rate of 100% Bonn dataset. We further designed a hardware data acquisition apparatus (with our program built-in) to smooth and ameliorate the data acquisition process when eliminating overmuch electrodes, which may serve as a promising seizure onset detecting device in the new era.

Seribu islands in the megacities of Jakarta on the frontlines of the climate crisis

Jakarta, the biggest city in Indonesia, has one district that consists of hundreds of islands that face severe climate hazards called the Seribu Islands complex. This study explores the evidence of local climate trends, the potential impact, and its policy intervention on Seribu Islands, which are classified as small island states and widely recognized as being especially at risk from climate change, threatening their economic and social growth. Long-term in-situ climate data, satellite data, interviews with local stakeholders, and literature reviews were utilized to conduct an exploratory descriptive analysis. The result revealed that Seribu Island experienced a 2.2°C increase in minimum temperature from 1980 until 2021, 3.5-fold of the frequency of extreme temperature and precipitation, 4.17 mm/year of sea level rise, and 10.8 ha land expansion in the densest island. Moreover, about 67% of the inhabitant’s islands were occupied by built-up areas that cover more than 50% of the region. Further, under the worst-case SLR scenario, about 58.4% of the area will be affected, and about 29 islands will disappear. This evidence was also reinforced by every single local respondent’s viewpoint who felt that climate change is occurring in the region. Even though the region faces a severe threat of climate change, the issue of climate change adaptation has not been mainstreamed yet into their local policy. Therefore, the urgency of a real-time climate ground station, a real-time early warning system, and establishing a Regional Disaster Management Agency (BPBD) at the district level have yet to be addressed. Furthermore, the knowledge gained from such case studies is outlined, along with some scientific evidence that may assist small island states in better fostering the opportunities provided by climate change adaptation.

The implementation of a real time early warning system using machine learning in an Australian hospital to improve patient outcomes

BackgroundEarly Warning Scores (EWS) monitor inpatient deterioration predominantly using vital signs. We evaluated inpatient outcomes after implementing an Artificial Intelligence (AI) based intervention in our local EWS. MethodsA prior study calculated a Deterioration Index (DI) with logistic regression utilising demographics, vital signs, and laboratory results at multiple time points to predict any major adverse event (MAE—all cause mortality, ICU admission, or medical emergency team activation). The current study is a single hospital, pre-post study in Australia comparing the DI plus the existing EWS (Between the Flags-BTF) to only BTF. Data were collected on all eligible inpatients (≥16 years, admitted ≥24 hours, in general non-palliative wards). Controls were inpatients in the same hospital between January and December 2019. The DI was integrated into the electronic medical record and alerts were sent to senior ward nurse phones (July 2020–April 2021). ResultsWe enrolled 28,639 patients (median age 73 years, IQR: 60–83) with 52.3% female. The intervention and control groups did not show any statistically significant differences apart from reduced admissions via the emergency department in the intervention group (40.4% vs 41.6%, P = 0.03). Risk for an MAE was lower in intervention than control (RR: 0.81; 95%CI: 0.74–0.89). Length of hospital stay was significantly reduced in the intervention group (3.74 days, IQR 1.84–7.26) compared to the control group (3.86 days, IQR 1.86–7.86, P = 0.002) ConclusionsImplementing the DI in one hospital in Australia was associated with some improved patient outcomes. Future RCTs are needed for further validation.

Ship Collision Realtime Early Warning System Based on Automatic Identification System (AIS) Data

One of the efforts to improve shipping safety is using the Automatic Identification System (AIS) according to the International Maritime Organization (IMO) recommendations. By utilizing Automatic Identification System (AIS) data, the system can monitor AIS ship movement patterns and compare them with the latest conditions. Information such as the ship's position, destination and speed is tracked. The author will create a real-time ship collision early warning system based on AIS data in this final project. The method used is the haversine formula to calculate the distance between two coordinate points on the earth's surface based on the ship's latitude and longitude. The results of AIS data in the form of latitude, longitude, and ID numbers will be processed using PHP to determine the distance and angle of the ship in real time. From 50 tests on two ships, two conditions can cause ship collisions. The results of these tests are that the percentage of collisions is 4%. The real-time system helps know the ship's actual movement based on current conditions

Coastal Flooding Assessment Induced by Barometric Pressure, Wind-Generated Waves and Tidal-Induced Oscillations: Kaštela Bay Real-Time Early Warning System Mobile Application

Our work presents a reliable procedure to obtain real-time assessment of the sea water elevation at the Kaštela Bay site to ensure the a priori warning in the case of expected coastal flooding along the site area. In its origin, the presented procedure relies on relevant data sets which are site-specific and locally observed. Observed data sets are used within the procedure to assess sea water surface elevation when induced by barometric pressure changes and wind-generated waves. Tidal-induced changes are introduced into the assessment procedure by a pre-learned algorithm which relies on long-term sea level oscillations from the relevant tidal gauge. Wind-generated wave heights are determined in the near shore area, following the features of the depth and reflection of the shoreline subsections. By coupling three mechanisms, this paper offers a unique real-time procedure to determine the sea water elevation and assess the possibility for coastline structure to be flooded by the sea. Given information is visualized in a form of mobile application that implements the algorithm and allows end users to set the notifications based on the given ruleset.

Science, Medicine, and the Anesthesiologist

Science, Medicine, and the Anesthesiologist

Lessons in machine learning model deployment learned from sepsis

In three recent and related publications, researchers from Johns Hopkins University and Bayesian Health report results from implementing and prospectively evaluating the Targeted Real-time Early Warning System (TREWS) for sepsis at five hospitals.1-3.

Prospective, multi-site study of patient outcomes after implementation of the TREWS machine learning-based early warning system for sepsis.

Early recognition and treatment of sepsis are linked to improved patient outcomes. Machine learning-based early warning systems may reduce the time to recognition, but few systems have undergone clinical evaluation. In this prospective, multi-site cohort study, we examined the association between patient outcomes and provider interaction with a deployed sepsis alert system called the Targeted Real-time Early Warning System (TREWS). During the study, 590,736 patients were monitored by TREWS across five hospitals. We focused our analysis on 6,877 patients with sepsis who were identified by the alert before initiation of antibiotic therapy. Adjusting for patient presentation and severity, patients in this group whose alert was confirmed by a provider within 3 h of the alert had a reduced in-hospital mortality rate (3.3%, confidence interval (CI) 1.7, 5.1%, adjusted absolute reduction, and 18.7%, CI 9.4, 27.0%, adjusted relative reduction), organ failure and length of stay compared with patients whose alert was not confirmed by a provider within 3 h. Improvements in mortality rate (4.5%, CI 0.8, 8.3%, adjusted absolute reduction) and organ failure were larger among those patients who were additionally flagged as high risk. Our findings indicate that early warning systems have the potential to identify sepsis patients early and improve patient outcomes and that sepsis patients who would benefit the most from early treatment can be identified and prioritized at the time of the alert.

Factors driving provider adoption of the TREWS machine learning-based early warning system and its effects on sepsis treatment timing.

Machine learning-based clinical decision support tools for sepsis create opportunities to identify at-risk patients and initiate treatments at early time points, which is critical for improving sepsis outcomes. In view of the increasing use of such systems, better understanding of how they are adopted and used by healthcare providers is needed. Here, we analyzed provider interactions with a sepsis early detection tool (Targeted Real-time Early Warning System), which was deployed at five hospitals over a 2-year period. Among 9,805 retrospectively identified sepsis cases, the early detection tool achieved high sensitivity (82% of sepsis cases were identified) and a high rate of adoption: 89% of all alerts by the system were evaluated by a physician or advanced practice provider and 38% of evaluated alerts were confirmed by a provider. Adjusting for patient presentation and severity, patients with sepsis whose alert was confirmed by a provider within 3 h had a 1.85-h (95% CI 1.66-2.00) reduction in median time to first antibiotic order compared to patients with sepsis whose alert was either dismissed, confirmed more than 3 h after the alert or never addressed in the system. Finally, we found that emergency department providers and providers who had previous interactions with an alert were more likely to interact with alerts, as well as to confirm alerts on retrospectively identified patients with sepsis. Beyond efforts to improve the performance of early warning systems, efforts to improve adoption are essential to their clinical impact and should focus on understanding providers' knowledge of, experience with and attitudes toward such systems.

Characteristics of the disastrous debris flow of Chediguan gully in Yinxing town, Sichuan Province, on August 20, 2019

On August 20, 2019, at 2 a.m., a disastrous debris flow occurred in Chediguan gully in Yinxing town, China. The debris flow destroyed the drainage groove and the bridge at the exit of the gully. In addition, the debris flow temporarily blocked the Minjiang River during the flood peak, flooding the Taipingyi hydropower station 200 m upstream and leaving two plant workers missing. To further understand the activity of the debris flow after the Wenchuan earthquake, the characteristics of this debris flow event were studied. Eleven years after the Wenchuan earthquake, a disastrous debris flow still occurred in the Chediguan catchment, causing more severe losses than those of earlier debris flows. In this paper, the formation mechanism and dynamic characteristics of this debris flow event are analysed based on a drone survey, high-definition remote sensing interpretations and other means. The catastrophic debris flow event indicates that debris flows in the Wenchuan earthquake area are still active. A large amount of dredging work in the main gully could effectively reduce the debris flow risk in the gully. In addition, it is also important to repair or rebuild damaged mitigation measures and to establish a real-time monitoring and early warning system for the high-risk gully.

A Sustainable Early Warning System Using Rolling Forecasts Based on ANN and Golden Ratio Optimization Methods to Accurately Predict Real-Time Water Levels and Flash Flood.

Remote monitoring sensor systems play a significant role in the evaluation and minimization of natural disasters and risk. This article presents a sustainable and real-time early warning system of sensors employed in flash flood prediction by using a rolling forecast model based on Artificial Neural Network (ANN) and Golden Ratio Optimization (GROM) methods. This Early Flood Warning System (EFWS) aims to support decision makers by providing reliable and accurate information and warning about any possible flood events within an efficient lead-time to reduce any damages due to flash floods. In this work, to improve the performance of the EFWS, an ANN forecast model based on a new optimization method, GROM, is developed and compared to the traditional ANN model. Furthermore, due to the lack of literature regarding the optimal ANN structural model for forecasting the flash flood, this paper is one of the first extensive investigations into the impact of using different exogenous variables and parameters on the ANN structure. The effect of using a rolling forecast model compared to fixed model on the accuracy of the forecasts is investigated as well. The results indicate that the rolling ANN forecast model based on GROM successfully improved the model accuracy by 40% compared to the traditional ANN model and by 93.5% compared to the fixed forecast model.

Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal.

The aim of the study was to inspect the acoustic properties and sleep characteristics of a preapneic snoring sound. The feasibility of forecasting upcoming respiratory events by snoring sound was also investigated. Participants with habitual snoring or a heavy breathing sound during sleep were recruited consecutively. Polysomnography was conducted, and snoring-related breathing sound was recorded simultaneously. Acoustic features and sleep features were extracted from 30-second samples, and a machine learning algorithm was used to establish 2 prediction models. A total of 74 eligible participants were included. Model 1, tested by 5-fold cross-validation, achieved an accuracy of 0.92 and an area under the curve of 0.94 for respiratory event prediction. Model 2, with acoustic features and sleep information tested by Leave-One-Out cross-validation, had an accuracy of 0.78 and an area under the curve of 0.80. Sleep position was found to be the most important among all sleep features contributing to the performance of the 2 models. Preapneic sound presented unique acoustic characteristics, and snoring-related breathing sound could be deployed as a real-time apneic event predictor. The models, combined with sleep information, serve as a promising tool for an early warning system to forecast apneic events. Wang B, Yi X, Gao J, etal. Real-time prediction of upcoming respiratory events via machine learning using snoring sound signal. J Clin Sleep Med. 2021;17(9):1777-1784.

A study on the impact of the 2017 early monsoon flash flood: Potential measures to safeguard livelihoods from extreme climate events in the haor area of Bangladesh

The northeastern part of Bangladesh is characterized by a unique wetland ecosystem called haor, which experiences regular monsoon flooding starting from late May/June. But change in rainfall pattern over last few decades has resulted in increased risk of early monsoon flash floods like the flood of 2017, which wreaked havoc in the haor at the beginning of April. Through a cross sectional survey covering 1845 flood affected households, the present study aims at understanding this flood's effects on their livelihoods and recommending policies to safeguard livelihoods in the face of changing climate. Analysis of rainfall and river water level data revealed that heavy rainfall in the haor and adjacent Indian sub-catchments were the main reasons behind this unprecedented event. Loss of income because of this flood was common, but was more prevalent among the older respondents, people having lower educational attainment and persons dependent on natural resources-based occupations. Most of the people who were forced to switch occupations experienced reduction in average daily income, irrespective of whatever occupations they switched to. Food insecurity was common among all groups, but was predominant among fishermen and day labourers. Integrated measures including development of real-time early warning system, community-based early response, proper management of crop protection embankments and regular dredging of waterways are essential to reduce the vulnerability of the people to flood. Furthermore, rapid expansion of newly developed short duration rice varieties (e.g. BRRI dhan 81), employment of case seedlings/raised-bed seedlings in elevated lands to ensure high quality timely transplantation immediately after recession of stagnant water, introduction of other profitable crops, poultry and dairy farming (e.g. sheep farming), creation of skill-based alternative livelihood options to adapt the agriculture sector with the changing climate. An equitable investment by the government is crucial to foster the economic growth by ending poverty by 2030, and to secure decent income, better livelihood and food security of the haor inhabitants.

An auto-detection network to provide an automated real-time early warning of rock engineering hazards using microseismic monitoring

The hazards associated with the extraction of ores and the construction of underground geoengineering structures should be mitigated to prevent the loss of human life. Microseismic monitoring (MM) has been used extensively in underground engineering around the world to assess the condition of underground structures and to improve personnel safety. However, MM cannot be currently used as a real-time early warning system because arrival detection still requires manual processing. Although studies have forwarded automated algorithms to perform this task, they are far less accurate than a human expert and therefore provide inaccurate source locations. In this study, we propose a workflow to specifically obtain the state-of-the-art performance of arrival detection. First, to avoid severe class-imbalance problems and speed up convergence in training, a new label form is proposed. We then propose an innovative self-attention mechanism in deep learning to efficiently model distant correlation (“segmentation-integration self-attention”). This new self-attention mechanism can effectively encode the distant inner dependencies by processing signals in several segments and by building correlations between the corresponding descriptors for the segments. Additionally, a new neural network architecture (“extractor-encoder-generator”) for automated arrival picking is proposed based on the above self-attention. This architecture extracts functional features of the signals, models correlations by an attention-related sublayer and generates a score map defining the confidence of the first break. The combination of the above technology is referred to as the “auto-detection” network. The new architecture outperforms all other algorithms by a considerable margin, provides a similar accuracy to that of a human expert, and is well balanced in terms of inference speed and accuracy. The auto-detection network allows for a practical and automated MM technique that can be used to provide an accurate real-time early warning system.

Sensing technology based quality control and warning systems for sleeve grouting of prefabricated buildings

Sleeve grouting is widely used for connecting prefabricated buildings (PB) components to form key nodes. Although some nondestructive testing techniques have been tried for the inspection of grouting defects, the application effect is not satisfactory due to its detection lag, poor anti-interference, and program complexity. The quality level of PB nodes is still uncertain, and early warning has not been achieved. This paper develops a quality control model and carries on real-time early warning system application to quality management of PB nodes. Toward this aim, sensing technology is applied to collect quality data, including grouting compactness and mechanical properties inside the sleeves for a prefabricated uniformed layer. This quality control model is verified and validated through a real-world PB project. The results show that it can effectively improve the inspection efficiency and accuracy in PB node quality, specifically in the real-time detection, monitoring and early warning of the quality level.

MACROECONOMIC PERSPECTIVE ON CONSTRUCTING FINANCIAL VULNERABILITY INDICATOR IN CHINA

This paper attempts to develop a financial vulnerability indicator for China as a barometer for the state of financial vulnerability in the Chinese financial market, possibly for real-time application. Twelve variables from different sectors are utilised to extract a common vulnerability component using a dynamic approximate factor model. Through the implementation of a Markovswitching Bayesian vector autoregression (MSBVAR) model, the empirical results indicate that a high-vulnerability episode is associated with substantially lower economic activity, but a low-vulnerability episode does not incur substantial changes in economic activity. Notably, the constructed indicator can serve as a real-time early warning system to signify vulnerabilities in the Chinese financial market.

A Fast Deploying Monitoring and Real-Time Early Warning System for the Baige Landslide in Tibet, China.

Landslide early warning systems (EWSs) have been widely used to reduce disaster losses. The effectiveness of a landslide EWS depends highly on the prediction methods, and it is difficult to correctly predict landslides in a timely manner. In this paper, we propose a real-time prediction method to provide real-time early warning of landslides by combining the Kalman filtering (KF), fast Fourier transform (FFT), and support vector machine (SVM) methods. We also designed a fast deploying monitoring system (FDMS) to monitor the displacement of landslides for real-time prediction. The FDMS can be quickly deployed compared to the existing system. This system also has high robustness due to the usage of the ad-hoc technique. The principle of this method is to extract the precursory features of the landslide from the surface displacement data obtained by the FDMS and, then, to train the KF-FFT-SVM model to make a prediction based on these precursory features. We applied this fast monitoring and real-time early warning system to the Baige landslide, Tibet, China. The results showed that the KF-FFT-SVM model was able to provide real-time early warning for the Baige landslide with high accuracy.

Mapping Forest Fire Risk and Development of Early Warning System for NW Vietnam Using AHP and MCA/GIS Methods

Forest fires constitute a major environmental problem in tropical countries, especially in the context of climate change and increasing human populations. This paper aims to identify the causes of frequent forest fires in Son La Province, a fire-prone and forested mountainous region in northwest Vietnam, with a view to constructing a forest fire-related database with multiple layers of natural, social and economic information, extracted largely on the basis of Landsat 7 images. The assessment followed an expert systems approach, applying multi-criteria analysis (MCA) with an analytical hierarchy process (AHP) to determine the weights of the individual parameters related to forest fires. A multi-indicator function with nine parameters was constructed to establish a forest fire risk map at a scale of 1:100,000 for use at the provincial level. The results were verified through regression analysis, yielding R2 = 0.86. A real-time early warning system for forest fire areas has been developed for practical use by the relevant government authorities to provide more effective forest fire prevention planning for Son La Province.

Multiple Instance Learning for Predicting Necrotizing Enterocolitis in Premature Infants Using Microbiome Data.

Necrotizing enterocolitis (NEC) is a life-threatening intestinal disease that primarily affects preterm infants during their first weeks after birth. Mortality rates associated with NEC are 15-30%, and surviving infants are susceptible to multiple serious, long-term complications. The disease is sporadic and, with currently available tools, unpredictable. We are creating an early warning system that uses stool microbiome features, combined with clinical and demographic information, to identify infants at high risk of developing NEC. Our approach uses a multiple instance learning, neural network-based system that could be used to generate daily or weekly NEC predictions for premature infants. The approach was selected to effectively utilize sparse and weakly annotated datasets characteristic of stool microbiome analysis. Here we describe initial validation of our system, using clinical and microbiome data from a nested case-control study of 161 preterm infants. We show receiver-operator curve areas above 0.9, with 75% of dominant predictive samples for NEC-affected infants identified at least 24 hours prior to disease onset. Our results pave the way for development of a real-time early warning system for NEC using a limited set of basic clinical and demographic details combined with stool microbiome data.