Short Lead Times Research Articles

Lead-time-dependent calibration of a flood forecasting model

The hydrological models used for flood forecasting purposes are usually first calibrated in simulation mode and then applied for flood forecasting in combination with some data assimilation procedure allowing to update and correct the model in real-time. Such a two-step procedure may not be the best choice to train model parameters for forecasting purposes. An alternative approach is to calibrate the hydrological model separately for each target lead time in the presence of the updating procedure. However, it is unclear whether this approach can provide the most efficient and the most robust forecasts. We compared in this paper three approaches to calibrate the parameters of a flood forecasting model: calibration in simulation mode (i.e. without data assimilation), lead-time-dependent calibration (i.e. with data assimilation), and a procedure combining both approaches. An hourly flood forecasting model (a parsimonious hydrological model combined with a state updating assimilation procedure) was used to produce forecasts for three lead times on 687 catchments and 40,411 flood events in metropolitan France. The performance of the model calibrated with the three approaches was evaluated according to catchment response times and flood rise times. An analysis of parameter robustness was also carried out. The results showed that lead-time-dependent calibration improves performance for catchments characterised by slow temporal dynamics. However, for catchments with fast temporal dynamics, it leads to degraded performance at short lead times and reduces parameter robustness. We found that the combined calibration approach is the best compromise between parameter robustness and performance at all lead times and for all catchment and flood types.

Statistical Response of ENSO Complexity to Initial Condition and Model Parameter Perturbations

Abstract Studying the response of a climate system to perturbations has practical significance. Standard methods in computing the trajectory-wise deviation caused by perturbations may suffer from the chaotic nature that makes the model error dominate the true response after a short lead time. Statistical response, which computes the return described by the statistics, provides a systematic way of reaching robust outcomes with an appropriate quantification of the uncertainty and extreme events. In this paper, information theory is applied to compute the statistical response and find the most sensitive perturbation direction of different El Niño–Southern Oscillation (ENSO) events to initial value and model parameter perturbations. Depending on the initial phase and the time horizon, different state variables contribute to the most sensitive perturbation direction. While initial perturbations in sea surface temperature (SST) and thermocline depth usually lead to the most significant response of SST at short and long ranges, respectively, initial adjustment of the zonal advection can be crucial to trigger strong statistical responses at medium range around 5–7 months, especially at the transient phases between El Niño and La Niña. It is also shown that the response in the variance triggered by external random forcing perturbations, such as the wind bursts, often dominates the mean response, making the resulting most sensitive direction very different from the trajectory-wise methods. Finally, despite the strong non-Gaussian climatology distributions, using Gaussian approximations in the information theory is efficient and accurate for computing the statistical response, allowing the method to be applied to sophisticated operational systems. Significance Statement The purpose of this work is to better understand how El Niño–Southern Oscillation (ENSO) responds to changes in its initial state and internal dynamics or external forcings. A statistical quantification of this response allows for the comprehension of the triggering conditions and the effect of climate change on the occurrence frequency and strength of each type of ENSO event. Such a study also allows to detect the most sensitive perturbation directions, which has practical significance in guiding anthropogenic activities. The approach used to study the response in this work is through the framework of information theory, which allows for an unbiased and robust assessment of the statistical response that is not affected by the turbulent dynamics of the system.

Predictability of the 2020 Strong Vortex in the Antarctic Stratosphere and the Role of Ozone

AbstractThe Antarctic vortex of October–December 2020 was the strongest on record in the satellite era for the season in the mid‐ to lower stratosphere. However, it was poorly predicted by the Australian Bureau of Meteorology's operational seasonal climate forecast system of that time, ACCESS‐S1, even at a short lead time of a month. Using the current operational forecast system, ACCESS‐S2, we have, therefore, tried to find a primary cause of the limited predictability of this event and conducted forecast sensitivity experiments to understand the potential role of ozone in the event and its associated anomalies of the Southern Annular Mode (SAM) and rainfall over south–eastern Australia and western Patagonia. Here, we show that the 2020 strong vortex event did not follow the canonical dynamical evolution seen in previous strong vortex events in spring but suddenly appeared as a result of the record‐low upward propagating wave activity in September 2020. The ACCESS‐S2 forecasts significantly underestimated the negative wave forcing in September even at zero lead time, irrespective of the ozone configuration, therefore falling short in predicting the record strength of the polar vortex in late spring 2020. Nevertheless, ACCESS‐S2 with prescribed realistic ozone that had large anomalies in the Antarctic stratosphere significantly better predicted the strong vortex and the subsequent positive SAM and related rainfall anomalies over south–eastern Australia and western Patagonia in the austral summer of 2020–21. This highlights the potentially important role of ozone variations for seasonal climate forecasting as a source of long‐lead predictability.

Choosing between additive and conventional manufacturing of spare parts: On the impact of failure rate uncertainties and the tools to reduce them

Recently, there has been great interest in using additive manufacturing (AM) to produce spare parts: allowing to produce spare parts with short lead times and close to the point of use, AM reduces the need for large inventories otherwise required by conventional manufacturing (CM) techniques to deal with intermittent spare parts demand. However, using AM to produce spare parts is limited by two main drawbacks: high production costs and uncertain failure rates arising from AM still being a relatively new production technique. While the former can be counterbalanced by inventory cost reductions, it is unclear how the latter impacts the sourcing option decision (i.e. AM or CM). We aim to fill this gap by studying a periodic review model in which spare parts demands follow a Poisson process. To make our analysis accurate and reliable, we leverage a material science approach to obtain realistic values for the failure rate uncertainties. From the results, it emerges that AM is heavily penalized by failure rate uncertainties much higher than those of CM. Consequently, we then focus on some recent tools developed to reduce AM failure rate uncertainties: porosity assessment and in-situ monitoring. We find that the failure rate uncertainty should be reduced by five and six percentage points to make it worthwhile to invest in porosity assessment and in-situ monitoring, respectively. Finally, we determine under which circumstances each tool is preferred over the other and found that porosity assessment is typically the most competitive uncertainty reduction tool.

EXAMINING THE INTEGRATION OF INDUSTRY 4.0 TECHNOLOGIES IN MANUFACTURING

The integration of Industry 4.0 technologies is transforming manufacturing, enhancing efficiency, flexibility, and sustainability. This research paper explores the application of key technologies such as artificial intelligence (AI), the Internet of Things (IoT), and big data analytics in the sector. By analyzing a comprehensive dataset that includes supply chain metrics from major corporations like Apple, Amazon, and Google, the study examines the impact of Industry 4.0 practices on critical supply chain performance indicators, including inventory turnover ratio, lead time, and customer satisfaction. Industry 4.0 technologies are reshaping supply chain management (SCM) strategies such as lean manufacturing, agile SCM, and cross-docking. AI and block chain are shown to significantly enhance supply chain agility and resilience. Companies that integrate these technologies tend to have more streamlined operations, shorter lead times, and higher customer satisfaction rates. Environmental sustainability practices are becoming increasingly important as businesses strive for more eco-friendly manufacturing processes. The study finds that organizations employing agile supply chain practices and leveraging advanced technologies often demonstrate superior operational efficiency and financial performance. The integration of these technologies also introduces challenges, such as increased supply chain complexity and risk management concerns. This paper provides a comprehensive overview of how Industry 4.0 technologies are revolutionizing manufacturing and supply chain dynamics. It highlights the future trajectory of smart manufacturing and emphasizes the need for continued innovation to address emerging challenges. The research offers insights into how companies can adapt their supply chain strategies to thrive in the evolving landscape of digital transformation and automation.

Predictability Limit of the 2021 Pacific Northwest Heatwave From Deep‐Learning Sensitivity Analysis

AbstractThe traditional method for estimating weather forecast sensitivity to initial conditions uses adjoint models, which are limited to short lead times due to linearization around a control forecast. The advent of deep‐learning frameworks enables a new approach using backpropagation and gradient descent to iteratively optimize initial conditions, minimizing forecast errors. We apply this approach to the June 2021 Pacific Northwest heatwave using the GraphCast model, yielding over 90% reduction in 10‐day forecast errors over the Pacific Northwest. Similar improvements are found for Pangu‐Weather model forecasts initialized with the GraphCast‐derived optimal, suggesting that model error is an unimportant part of the perturbations. Eliminating small scales from the perturbations also yields similar forecast improvements. Extending the length of the optimization window, we find forecast improvement to about 23 days, suggesting atmospheric predictability at the upper end of recent estimates.

An integer mathematical model proposal for an agile external milkrun system

A milkrun system is a cyclical material delivery system that enables frequent small-batch deliveries with short lead times and low inventories. The literature classifies milkrun systems into external and internal categories. External milkrun systems in agile production environments, on the other hand, have not received much attention in the literature thus far. This study presents an integer mathematical model for the cost minimisation of external milkrun tours and discusses agile external milkrun systems. In addition to minimising costs, the proposed mathematical model can determine the types and number of milkrun vehicles. This is an important advantage for logistics professionals. The proposed model was coded using the Lingo software, and its effectiveness was shown using both a real-world problem and randomly generated test problems. The proposed model found the best solution in a few seconds for some real-world and most test problems. The results have shown that the proposed mathematical model for external milkrun systems in agile environments is useful.

A Flow Shop Scheduling Method Based on Dual BP Neural Networks with Multi-Layer Topology Feature Parameters

Nowadays, the focus of flow shops is the adoption of customized demand in the context of service-oriented manufacturing. Since production tasks are often characterized by multi-variety, low volume, and a short lead time, it becomes an indispensable factor to include supporting logistics in practical scheduling decisions to reflect the frequent transport of jobs between resources. Motivated by the above background, a hybrid method based on dual back propagation (BP) neural networks is proposed to meet the real-time scheduling requirements with the aim of integrating production and transport activities. First, according to different resource attributes, the hierarchical structure of a flow shop is divided into three layers, respectively: the operation task layer, the job logistics layer, and the production resource layer. Based on the process logic relationships between intra-layer and inter-layer elements, an operation task–logistics–resource supernetwork model is established. Secondly, a dual BP neural network scheduling algorithm is designed for determining an operations sequence involving the transport time. The neural network 1 is used for the initial classification of operation tasks’ priority; and the neural network 2 is used for the sorting of conflicting tasks in the same priority, which can effectively reduce the amount of computational time and dramatically accelerate the solution speed. Finally, the effectiveness of the proposed method is verified by comparing the completion time and computational time for different examples. The numerical simulation results show that with the increase in problem scale, the solution ability of the traditional method gradually deteriorates, while the dual BP neural network has a stable performance and fast computational time.

Development of Readily Available & Robust High Heat Flux Gardon Gauges

Concentrated solar power (CSP) technologies deliver concentrated solar energy as a heat source to industrial processes, power generation cycles, and chemical cycles. CSP systems require accurate and reliable high flux measurements, and next generation CSP systems will require flux measurement up to 1000 W/cm2. Existing flux measurement devices do not comprehensively meet the flux rating, cycle life, cost, and lead-time needs of stakeholders, necessitating the development of an improved flux sensor. In this study, Sandia National Laboratories (SNL) partnered with Hukseflux Thermal Sensors to develop a low-cost, short lead-time, and robust flux sensor rated to 250 W/cm2. Three prototype circular foil gauge designs were assessed for performance at the National Solar Thermal Test Facility (NSTTF) at SNL. Each gauge design measured flux up to 250 W/cm2 with <5% measurement error. Following baseline error quantification, gauges were exposed to flux above 500 W/cm2 to assess gauge failure mechanisms. Gauges physically survived >500 W/cm2 flux exposure, but measurement error was found to increase after foil coatings reached 400 °C. The results of this study suggest that coating optical properties change at excessive temperatures and that foil coating temperature, rather than heat flux level, dictates the acceptable gauge measurement range.

A boundary perturbation approach for regional wave ensemble forecast

In this study, we developed and validated two wave ensemble prediction systems (WEPS) to forecast wave conditions along the southeastern coast of Australia. Using the SWAN model (GEN3 ST6), we integrated complex bathymetric features with an unstructured grid and validated model outputs against buoy observations from Sydney, Port Kembla, and Batemans Bay. The two WEPS, SWAN-WW3 and SWAN-Pert, utilize different methodologies: SWAN-WW3 derives boundary conditions from NCEP’s Global Wave Ensemble System, while SWAN-Pert employs Latin Hypercube Sampling for boundary perturbations based on historical data. Our results demonstrate that both systems effectively predict significant wave height (Hs), with SWAN-Pert showing improved forecast accuracy in certain metrics compared to SWAN-WW3. Despite underdispersion in spread-skill diagrams, both WEPS exhibited good agreement with observed data. Additionally, rank histograms revealed that SWAN-Pert is more reliable at shorter lead times. This study highlights the potential of integrating statistical sampling methods and ensemble systems for enhancing regional wave forecasting accuracy.

Reducing operational costs in a manufacturing system that incorporates quality assurances, probabilistic failures, overtime and outsourcing

This research focuses on the dynamic landscape of today's competitive market, where production management must meet clients' expectations for high-quality products and shorter lead times while managing internal disruptions due to inevitable defects and unpredictable equipment failures. Achieving these operational goals without compromising product quality, missing deadlines, or experiencing production disruptions is essential for minimizing operational expenses. The study examines the dynamics of the system's operating costs through the development of models, mathematical formulations, optimization techniques, and algorithm proposals. It demonstrates the system's convexity and establishes the optimal batch time for implementing the proposed methodologies. The research results show relevant failure costs of 3.51 %, overtime added costs of 4.57 %, outsourcing setup costs of 0.73 %, outsourcing variable costs of 41.82 %, quality-related costs of 2.98 %, in-house variable costs of 40.42 %, and in-house holding costs of 3.55 %. The study develops strategies for optimal overtime use to meet production targets without excessive labor costs and provides a structured framework for making informed outsourcing decisions that balance cost savings with quality and reliability considerations. Overall, the research provides a robust framework for reducing operational costs while maintaining or improving the quality and reliability of manufacturing processes

Optimal procurement decisions based on smoothing strategy for dual sourcing with different lead times

This study establishes a discrete-time inventory model with a smoothing procurement strategy in dual sourcing. Specifically, an assembly manufacturer purchases components from two suppliers: one overseas supplier with a long lead time but a low price, and one local supplier with a short lead time but a high price. A smoothing procurement approach taking into account previous ordering record and current demand is adopted by the assembly manufacturer. We start by exploring the scenario where storage space is not a constraint. The cost-minimization conditions for single sourcing and dual sourcing are studied. Assuming demand follows a normal distribution, the optimal order quantities and inventory levels are subsequently derived and shown to follow the same distribution. By comparing the inventory costs, it is found that dual sourcing is superior to single sourcing. Then, we examine the scenario where an assembly manufacturer with limited storage space collaborates with a third-party logistics company. The optimal order quantities, inventory levels, smoothing procurement coefficient and storage space are derived for the assembly manufacturer. Finally, numerical experiments with practical data are constructed to verify our findings.

Validation of an AI-based solution for breast cancer risk stratification using routine digital histopathology images

BackgroundStratipath Breast is a CE-IVD marked artificial intelligence-based solution for prognostic risk stratification of breast cancer patients into high- and low-risk groups, using haematoxylin and eosin (H&E)-stained histopathology whole slide images (WSIs). In this validation study, we assessed the prognostic performance of Stratipath Breast in two independent breast cancer cohorts.MethodsThis retrospective multi-site validation study included 2719 patients with primary breast cancer from two Swedish hospitals. The Stratipath Breast tool was applied to stratify patients based on digitised WSIs of the diagnostic H&E-stained tissue sections from surgically resected tumours. The prognostic performance was evaluated using time-to-event analysis by multivariable Cox Proportional Hazards analysis with progression-free survival (PFS) as the primary endpoint.ResultsIn the clinically relevant oestrogen receptor (ER)-positive/human epidermal growth factor receptor 2 (HER2)-negative patient subgroup, the estimated hazard ratio (HR) associated with PFS between low- and high-risk groups was 2.76 (95% CI: 1.63–4.66, p-value < 0.001) after adjusting for established risk factors. In the ER+/HER2- Nottingham histological grade (NHG) 2 subgroup, the HR was 2.20 (95% CI: 1.22–3.98, p-value = 0.009) between low- and high-risk groups.ConclusionThe results indicate an independent prognostic value of Stratipath Breast among all breast cancer patients, as well as in the clinically relevant ER+/HER2- subgroup and the NHG2/ER+/HER2- subgroup. Improved risk stratification of intermediate-risk ER+/HER2- breast cancers provides information relevant for treatment decisions of adjuvant chemotherapy and has the potential to reduce both under- and overtreatment. Image-based risk stratification provides the added benefit of short lead times and substantially lower cost compared to molecular diagnostics and therefore has the potential to reach broader patient groups.

Corrosion Evaluation of Metal Binder-Jetted Stainless Steel Parts: Influence of Printing Parameters on Performance

Additive manufacturing is a rapidly expanding fabrication process in both research and industry, capable of constructing parts layer by layer from a 3D CAD model. ASTM recognizes seven distinct additive manufacturing processes, with the majority capable of producing metal components. Among these technologies, binder jetting stands out, offering the potential to create a variety of metallic parts for applications in the biomedical and energy sectors at reduced costs and shorter lead times. Achieving acceptable corrosion resistance is crucial for such applications. Binder jetting exhibits advantages such as relatively high part density (~99%) and good surface roughness (~6μm for as-sintered parts), making it an ideal candidate for investigating corrosion behavior, given the general correlation between low surface roughness, high density, and improved corrosion performance. Despite its promise, binder jetting lacks standardized procedures for producing parts with specific properties, necessitating users to determine parameters through trial and error. The corrosion behavior of parts created through binder jetting remains uncertain due to various influencing parameters. This study focuses on two key printing factors: layer thickness and binder saturation, chosen due to their impact on surface roughness, density, and thereby on corrosion. Existing literature suggests that increased layer thickness may reduce powder bed density, potentially affecting green part density and causing more significant shrinkage during sintering. Simultaneously, binder saturation is recognized as a critical factor influencing final part quality. Predicting the combined impact of layer thickness and binder saturation on corrosion performance proves challenging. To unravel this relationship, we conducted experiments using an ExOne Innovate Plus metal binder jetting machine and 316L stainless steel metal powder (mean powder size of 22μm) which is widely recognized as the most common powder in research. A number of test runs were conducted by varying layer thickness and binder saturation, offering insights into their nuanced interplay in the context of corrosion performance. Initially, to address insufficient green strength, printer drying time and curing cycle time were optimized by conducting a series of experiments changing the drying time and curing time. Subsequent printing involved changing binder saturation parallel to the layer thickness variations, resulting in distinct parts with varying properties. Then the predetermined curing cycle was executed, followed by de-powdering, and sintering in a furnace. Sintering is imperative to increase the part density and enhance the overall strength of the final product. However, it's essential to acknowledge that sintering may adversely impact corrosion resistance due to the remaining porosity and surface fissures. The choice of sintering profile becomes crucial, necessitating careful consideration to minimize elemental migration during sintering and mitigate porosity. Scanning electron microscopy and optical microscopy are used to analyze sintered parts and interpret corrosion mechanisms. Electrochemical tests were conducted to identify corrosion rates, corrosion potential, passivation behavior, pitting potential, etc. in contrast with conventionally manufactured 316L in M NaCl solution. This study offers insights and recommendations aimed at enhancing the corrosion resistance of binder jetted stainless steel parts.

Strategies for digital transformation of cold chain logistics of fresh agricultural products based on Internet of things

The rapid development of cities and urbanization in China has forced the growth of new channels for buying agricultural products. The purpose of this research is to examine how Internet of Things (IoT’s) technologies can digitize a traditional fresh food supply chain. Comparative and descriptive analysis methods are used to highlight the major pain points in the traditional supply chains and assess how digital transformation could help. We delve into every part of digital transformation, which includes establishing an information platform based on IoT and developing smart storage options. Our findings revealed that through end-to-end digital integration, supply chain efficiency is improved with shorter lead times and leaner inventories that yield reduced costs as well as fewer losses while ensuring product quality and traceability. In sum, such an approach would enhance sustainability within the fresh food value chain. As such, our article highlights key aspects of transitioning towards a digital environment in this sector for those planning similar ventures.

Product Development Anxiety: A Contingency Planning Model for Innovative Production Companies

This paper investigates the possible failure modes of the product development process in production companies that are active in the B2C markets with a focus on household products. Since these cases require short lead times and are difficult to differentiate, in many instances the result will not be the desired one and could affect profitability for a season or for good. A model of these possibilities is created and an approach to plan contingencies for their solutions is proposed in the article. The main guideline is to switch from failure probability determination to accepting failure as inevitable and using digital solutions to reinforce the development process to offset its impact. For this goal, an Industry 5.0 Abatement Factor (abbreviated IFAF) is introduced in the contingency planning approach, which factors in the evaluation of the low cost of digital instruments and the proper mix of Technology, Humans, and AI (abbreviated THAI). The new working procedure based on these concepts and their interlinkages is discussed based on specific examples.

Evaluation of Performance Enhancement in Garment Industry Using Lean Tools

The comprehensive idea of lean manufacturing emphasizes the identification and removal of wastes related to production. Increased productivity, shorter lead times, improved line balancing, lower work in progress (WIP), higher quality, more adaptable designs, lower costs, etc. are all major objectives of applying lean tools. This study aims to investigate how lean manufacturing techniques might enhance manufacturing efficiency in Bangladeshi RMG sector. The sewing section of a small scale garment industry was the focus of this study. In the study, the existing line layout of a trouser style was examined using value stream mapping which contributed to the identification of numerous waste in the existing layout. In order to change the current status of the sewing section, the workplace was organized by 5S tools to remove wastes identified by VSM followed by altering the conventional longer line layout into three cell of work station as suggested by cellular manufacturing. After implementation of these lean tools, results observed include, production output is increased by 17.80%, efficiency is increased by 12.46%, SMV of product is standardized to 18.68 min., defects per hundred units is reduced to 10% and order change over time is reduced by 37%. The conclusions reached during the execution of this implementation study are realistic and applicable in industries with similar structures. GUB JOURNAL OF SCIENCE AND ENGINEERING, Vol 10(1), 2023 P 27-36

Trade-offs between stump-to-roadside lead time and harvesting cost, when using different number of operators in a harvester-forwarder system

AbstractFor customer-oriented wood supply, buffering is required for flexibility to handle interactions in the wood procurement system. This includes balancing lead-time and operational cost by using stocks and production capacity as buffers. Despite the well-known challenge to balance the interactions between harvesting and forwarding in Nordic mechaniced CTL-operations, there has been limited research on how the machine groups can be staffed to enable flexibility and more focus on other measures to create flexibility. Therefore, this study explored trade-offs between wood lead-time and harvesting cost in the stump-to-roadside part of the wood supply chain by altering the numer of full-time working operators in the harvesting groups. This was done using discrete-event simulations implemented in Anylogic software. Input data included information about operational conditions in 1500 forest stands. The results revealed that the best balance was to have sufficient harvesting capacity to adjust wood lead times at the expense of increased harvesting costs. Of the tested options, the best balance was achieved when staffing a two-machine group with three operators, and thereby allocating 50% of the used work-shifts to regulating the field wood stock between the two machines. This resulted in the shortest lead times and the smallest harvesting cost increase. Compared to the option with no flexibility for stock adjustment (4 operators), the average lead-time could be reduced to one tenth at a cost increase of 3.4%. These findings have the potential to improve decisions of how harvesting groups are staffed to balance specific objectives of desired lead times and costs, which migh prove to be a valuable addition to the already used measures to manage wood flow.

The varied clonal trajectory of liver and lung metastases of colorectal cancer

BackgroundThe liver and lungs are the most common sites of colorectal cancer (CRC) metastases. Their involvement can take five different clinical scenarios: lung metastases only, liver metastases only, lung metastases before liver metastases, liver metastases before lung metastases and simultaneous lung and liver metastases. Using clinical and morphological data we studied the clonal trajectory of these scenarios. Materials and methodsA total of 465 (CRC) patients with 7952 liver and 6406 lung metastases were evaluated. Metastases clinical route was deciphered from metastases number, timing, and linear/parallel ratio (LPR)- a computerized parameter used for deducing clonal trajectories. LPR of +1 suggest pure linear dissemination and −1 pure parallel. ResultsLung-only metastases: high percentage of metachronous disease with a long lead time and a low LPR suggest parallel dissemination. Liver-only metastases: Rare metachronous disease with a short lead time, and a high LPR suggest linear spread. Lung-before-liver metastases: rare solitary metastasis, a median gap of 21 months between the organs, high lung and low liver LPRs suggest linear progression to the lungs and parallel dissemination to the liver. Liver-before-lung metastases: low liver and high lung LPRs and a median gap of 16.5 months between the organs suggest parallel dissemination to the liver and linear spread from the liver to the lungs. Simultaneous liver and lung metastases: rare solitary metastasis and similar and high LPRs suggest simultaneous linear progression to both organs. ConclusionsCRC metastases have different dissemination trajectories in different clinical scenarios. This information can potentially impact on clinical management.

Long Short-Term Memory Networks' Application on Typhoon Wave Prediction for the Western Coast of Taiwan.

Huge waves caused by typhoons often induce severe disasters along coastal areas, making the effective prediction of typhoon-induced waves a crucial research issue for researchers. In recent years, the development of the Internet of Underwater Things (IoUT) has rapidly increased the prediction of oceanic environmental disasters. Past studies have utilized meteorological data and feedforward neural networks (e.g., BPNN) with static network structures to establish short lead time (e.g., 1 h) typhoon wave prediction models for the coast of Taiwan. However, sufficient lead time for prediction remains essential for preparedness, early warning, and response to minimize the loss of lives and properties during typhoons. The aim of this research is to construct a novel long lead time typhoon-induced wave prediction model using Long Short-Term Memory (LSTM), which incorporates a dynamic network structure. LSTM can capture long-term information through its recurrent structure and selectively retain necessary signals using memory gates. Compared to earlier studies, this method extends the prediction lead time and significantly improves the learning and generalization capability, thereby enhancing prediction accuracy markedly.