Hasil untuk "Disasters and engineering"

Mark Looi

The rapid advance of Generative AI into software development prompts this empirical investigation of perceptual effects on practice. We study the usage patterns of 147 professional developers, examining perceived correlates of AI tools use, the resulting productivity and quality outcomes, and developer readiness for emerging AI-enhanced development. We describe a virtuous adoption cycle where frequent and broad AI tools use are the strongest correlates of both Perceived Productivity (PP) and quality, with frequency strongest. The study finds no perceptual support for the Quality Paradox and shows that PP is positively correlated with Perceived Code Quality (PQ) improvement. Developers thus report both productivity and quality gains. High current usage, breadth of application, frequent use of AI tools for testing, and ease of use correlate strongly with future intended adoption, though security concerns remain a moderate and statistically significant barrier to adoption. Moreover, AI testing tools' adoption lags that of coding tools, opening a Testing Gap. We identify three developer archetypes (Enthusiasts, Pragmatists, Cautious) that align with an innovation diffusion process wherein the virtuous adoption cycle serves as the individual engine of progression. Our findings reveal that organizational adoption of AI tools follows such a process: Enthusiasts push ahead with tools, creating organizational success that converts Pragmatists. The Cautious are held in organizational stasis: without early adopter examples, they don't enter the virtuous adoption cycle, never accumulate the usage frequency that drives intent, and never attain high efficacy. Policy itself does not predict individuals' intent to increase usage but functions as a marker of maturity, formalizing the successful diffusion of adoption by Enthusiasts while acting as a gateway that the Cautious group has yet to reach.

Jingyue Li, André Storhaug

With the advancement of Agentic AI, researchers are increasingly leveraging autonomous agents to address challenges in software engineering (SE). However, the large language models (LLMs) that underpin these agents often function as black boxes, making it difficult to justify the superiority of Agentic AI approaches over baselines. Furthermore, missing information in the evaluation design description frequently renders the reproduction of results infeasible. To synthesize current evaluation practices for Agentic AI in SE, this study analyzes 18 papers on the topic, published or accepted by ICSE 2026, ICSE 2025, FSE 2025, ASE 2025, and ISSTA 2025. The analysis identifies prevailing approaches and their limitations in evaluating Agentic AI for SE, both in current research and potential future studies. To address these shortcomings, this position paper proposes a set of guidelines and recommendations designed to empower reproducible, explainable, and effective evaluations of Agentic AI in software engineering. In particular, we recommend that Agentic AI researchers make their Thought-Action-Result (TAR) trajectories and LLM interaction data, or summarized versions of these artifacts, publicly accessible. Doing so will enable subsequent studies to more effectively analyze the strengths and weaknesses of different Agentic AI approaches. To demonstrate the feasibility of such comparisons, we present a proof-of-concept case study that illustrates how TAR trajectories can support systematic analysis across approaches.

Fulvio D. Lopane, Richard J. Dawson

ABSTRACT Infrastructure systems provide crucial services to human settlements. Extreme weather events, especially flooding, can disrupt these vital services. Temporary and demountable flood protections (TDFPs) are increasingly used to protect infrastructure assets and provide resilience. Budget constraints mean that TDFP are typically deployed to multiple sites from a single warehouse. Identifying optimal locations to maximise coverage and minimise costs is a complex spatial problem not yet tackled in the literature. To address this, a Spatial Resource Allocation Optimisation (SRAO) framework, using a genetic algorithm (GA), has been developed. The SRAO framework is applied to a case study in the Humber Estuary (UK) where 133 strategic infrastructure assets serve over 400,000 people in the floodplain. Eight scenarios assess how cost, TDFP availability, transport and asset prioritisation for protection influence warehouse size and sites. The SRAO identifies optimal strategies that, relative to other strategies, reduce annual costs by 40%–50% and deployment times by 60%–70%. Furthermore, 8 ‘hotspot’ sites appear in over 60% of optimal solutions; these can be considered robust to model uncertainties and scenario assumptions, providing decision‐makers with locations performing well under varied conditions. The methodology benefits local authorities, infrastructure operators and emergency management agencies, reducing costs and improving resilience for communities.

Roberto Verdecchia, Justus Bogner

From its first adoption in the late 80s, qualitative research has slowly but steadily made a name for itself in what was, and perhaps still is, the predominantly quantitative software engineering (SE) research landscape. As part of our regular column on empirical software engineering (ACM SIGSOFT SEN-ESE), we reflect on the state of qualitative SE research with a focus group of experts. Among other things, we discuss why qualitative SE research is important, how it evolved over time, common impediments faced while practicing it today, and what the future of qualitative SE research might look like. Joining the conversation are Rashina Hoda (Monash University, Australia), Carolyn Seaman (University of Maryland, United States), and Klaas Stol (University College Cork, Ireland). The content of this paper is a faithful account of our conversation from October 25, 2025, which we moderated and edited for our column.

Dipin Khati, Yijin Liu, David N. Palacio et al.

Applications of Large Language Models (LLMs) are rapidly growing in industry and academia for various software engineering (SE) tasks. As these models become more integral to critical processes, ensuring their reliability and trustworthiness becomes essential. Consequently, the concept of trust in these systems is becoming increasingly critical. Well-calibrated trust is important, as excessive trust can lead to security vulnerabilities, and risks, while insufficient trust can hinder innovation. However, the landscape of trust-related concepts in LLMs in SE is relatively unclear, with concepts such as trust, distrust, and trustworthiness lacking clear conceptualizations in the SE community. To bring clarity to the current research status and identify opportunities for future work, we conducted a comprehensive review of $88$ papers: a systematic literature review of $18$ papers focused on LLMs in SE, complemented by an analysis of 70 papers from broader trust literature. Additionally, we conducted a survey study with 25 domain experts to gain insights into practitioners' understanding of trust and identify gaps between existing literature and developers' perceptions. The result of our analysis serves as a roadmap that covers trust-related concepts in LLMs in SE and highlights areas for future exploration.

Yanqing Men, Zixuan Zhang, Jing Wang et al.

In recent years, geological disasters such as water inrush during drilling and blasting operations have posed significant challenges in tunnel engineering. This paper presents a novel continuous-discrete coupling method based on LS-DYNA, combining the finite element method (FEM) and smoothed particle hydrodynamics (SPH), to simulate the water inrush phenomenon in blasting engineering. The proposed FEM-SPH model effectively captures the propagation of explosion shock waves, simulates small deformation areas with solid grids, and models water behavior using SPH. This study systematically investigates the dynamic evolution of water inrush, divided into three distinct phases: the rupture of the water-resistant rock layer, the emergence of fluid-conducting channels, and the onset of large-scale water influx. Results indicate that under blasting load, the stress of the surrounding rock increases sharply, leading to instantaneous water inrush. The FEM-SPH model demonstrates superior performance in simulating the complex interactions between blasting stress waves, water pressure, and rock mass damage. This research provides new insights and methods for water control in tunnel engineering and offers significant potential for preventing water inrush disasters in underground construction.

Alexandra Seawell, Hayley J. Fowler, Stephen Blenkinsop et al.

ABSTRACT The temporal distribution of rainfall is a key driver of flood response. Yet, flood estimation methods are frequently based on symmetrical design profiles. Recent research using sub‐hourly rainfall data from Great Britain indicates that a significant proportion of observed rainfall events are non‐symmetrical. This paper investigates how different rainfall profiles affect river flow hydrographs for a set of small, flash‐flooding catchments. Results show that rainfall profiles affect observed hydrograph peak flow and timing. Most importantly, back‐loaded rainfall profiles lead to higher peak flows than symmetrical or front‐loaded profiles. These observations are compared to current design practice, using the Revitalised Flood Hydrograph (ReFH2.3) model to simulate flows from different rainfall profiles. Simulated events reproduce the observed response of peak magnitude but differ for peak time. A comparison of modelled flows with catchment descriptors indicates that steep, low permeability, wet catchments are most sensitive to rainfall profile shape. These are also the most vulnerable catchments to flash flooding. We recommend that different rainfall profile shapes should be considered for flood risk assessments in rapid response catchments, particularly since global warming is increasing the number of intense, short‐duration downpours.

Yugeng Zhang, Heng Zhang, Hongwei Lian et al.

Abstract The stability of flooded coal pillars has long been a challenging issue in coal pillar research, especially during the recovery of production after water inrush disasters. Due to the harsh on-site conditions, it is difficult to directly collect samples for analysis, which brings numerous challenges in both design and production. Based on the engineering background of the 1314 working face of Xiaoyun Coal Mine, this paper proposes a clustering optimization algorithm and successfully uses microseismic data collected on-site to identify the boundary of the flooded coal pillar, validating the results through simulation comparisons. The study found that the flooded state within the coal pillar can be classified into saturated flooded zones, unsaturated flooded zones, and dry zones. The characteristics of the flooded coal pillar during the early stage of mining are more complex, with irregular variations in the flood boundary and local phenomena of sudden changes. Through the analysis of stress and delamination data, the primary controlling factors of this phenomenon are identified and the causes are explained. The research not only demonstrates the feasibility of using microseismic data to identify the flooding status of coal pillars but also provides valuable insights for analyzing the flooded state of coal pillars during the recovery of production after water inrush incidents. This study, particularly regarding coal pillar monitoring and safety control, presents new challenges.

Hehua Zhu, Jinxiu Yan, Wenhao Liang

Social development has led to the placement of high standards on ultra-long and ultra-deep mountain tunnels. Disasters may be encountered during the construction and maintenance of such mountain tunnels due to high geostress, high geotemperature, high hydraulic pressure, and special adverse strata, in addition to various other problems caused by engineering activities. To deal with uncertain geological conditions during mountain tunnel construction, comprehensive geological prediction, refined monitoring, and dynamic design and construction methods based on information technology should be adopted. For the operation and maintenance of ultra-long tunnels, the concepts of dynamic evacuation rescue, active protection, energy conservation, and environmental protection should be fully embodied in order to address significant problems related to ventilation, rescue situations, and energy consumption. Moreover, integrated construction and maintenance should be carried out to achieve digital sensing and intelligent maintenance. New ideas and technologies should be adopted to improve the quality and efficiency of the whole process of construction and operation, and to enable the construction of environmentally friendly tunnels, thus achieving the ultimate goals of safety, efficiency, greenness, and intelligence for ultra-long and ultra-deep rock tunnels. With the development of society, the economy, and transportation networks, the construction of ultra-long and ultra-deep tunnels through mountains has become increasingly inevitable. Ultra-long and ultra-deep tunnels are generally defined as tunnels that have a length exceeding 10 km and a depth exceeding 500 m [1]. Mountain tunnels mainly consist of road tunnels, railway tunnels, and hydraulic tunnels. Although an ultra-long and ultra-deep tunnel potentially features the advantages of safety, environmental friendliness, and speed, the cost and difficulty of project establishment, construction, and operation are considerable. Table 1 lists the ultra-long and ultra-deep mountain tunnels that have already been built or are under construction around the world. According to an incomplete survey, there are 56 ultra-long and ultra-deep mountain tunnels in China and 21 abroad. Among these tunnels, the 57.1 km Gotthard Basis Tunnel is the longest and deepest in the world, the 32.7 km New Guanjiao Tunnel is the world’s longest tunnel above 3000 m, the 18 km Highway Tunnel of Qinling Zhongnan Mountain is the world’s longest double-line highway tunnel, and the 16 km Qinling Tianhuashan Tunnel is Asia’s longest single-hole two-lane high-speed rail tunnel. With the increasing demand for and ongoing progress in tunnel construction technologies, the construction of ultra-long and ultra-deep mountain tunnels will usher in new development opportunities. Due to the high geostress, high geotemperature, and ultra-long construction and operation, these complex tunnel projects must handle unprecedented challenges in terms of design, construction, operation, and maintenance, which demand new ideas and engineering measures.

Zhen Huang, Weizhen Zeng, Qixiong Gu et al.

Abstract Natural rock is a very common building material, and its stability plays an important role in the safety and durability of tunnels, rock slopes, and other projects. However, tunnels and other underground projects often face the threat of disasters such as fires, and during their service, the surrounding rocks are also subject to erosion by acidic groundwater. Therefore, it is of great engineering significance to study the damage of the surrounding rocks under the threat of fire and acid groundwater. In this study, the variations in chromatic aberration, mass, P-wave velocity, porosity, thermal conductivity, and tensile strength of granite, sandstone, and marble induced by high temperature and chemical erosion are investigated from room temperature to 1000 °C. The results show that the rock materials exhibit different degrees of variation in physical and mechanical properties under the effect of high temperatures and acidic solution corrosion. Then, according to the variation of physical and mechanical properties, the formation mechanism of damage is discussed and the damage process verified by the deterioration of mechanical properties. For rocks heated to different temperatures and eroded by acid solution, the main reason for damage is the increase of cracks which destroy the structure and weaken the bonding between particles.

Ebube Alor, Ahmad Abdellatif, SayedHassan Khatoonabadi et al.

Software engineering (SE) chatbots are increasingly gaining attention for their role in enhancing development processes. At the core of chatbots are Natural Language Understanding platforms (NLUs), which enable them to comprehend user queries but require labeled data for training. However, acquiring such labeled data for SE chatbots is challenging due to the scarcity of high-quality datasets, as training requires specialized vocabulary and phrases not found in typical language datasets. Consequently, developers often resort to manually annotating user queries -- a time-consuming and resource-intensive process. Previous approaches require human intervention to generate rules, called labeling functions (LFs), that categorize queries based on specific patterns. To address this issue, we propose an approach to automatically generate LFs by extracting patterns from labeled user queries. We evaluate our approach on four SE datasets and measure performance improvement from training NLUs on queries labeled by the generated LFs. The generated LFs effectively label data with AUC scores up to 85.3% and NLU performance improvements up to 27.2%. Furthermore, our results show that the number of LFs affects labeling performance. We believe that our approach can save time and resources in labeling users' queries, allowing practitioners to focus on core chatbot functionalities rather than manually labeling queries.

Zehra Koyuncu, Ömer Ekmekcioğlu

The main objective of the current study is to generate a flood hazard map by using the machine learning algorithms hybridized with the geographic information systems (GIS). In this regard, the province of Istanbul, which is the metropolitan city of Turkey, was selected as the focal region within the scope of the study. The class imbalance was tackled through the commonly used random under sampling (RUS) technique in order to create a fair comparison datum line. It is worth mentioning that this is the first time this approach has been used for flood hazard mapping studies in Turkey. Random forest (RF), stochastic gradient boosting (SGB), and XGBoost algorithms were used. The best predictive performance was obtained with the XGBoost algorithm, followed by SGB and RF, respectively. The RF and SGB models showed a 90.67% success rate in determining the inundation points, while the XGBoost model outperformed its counterparts with a 92.00% success rate in determining the inundation points. In this research, the importance levels of the flood triggering variables were further investigated in order to enliven the comprehensibility of the obtained results. Thus, the most important variable was the precipitation, followed by the distance to the drainage network and the number of curves, respectively. Finally, it is suggested that flood vulnerability mapping attempts can be considered as promising approaches against increasing flood incidents over the years.

Zhenjun WANG, Qingsong ZHANG, Bing HUI et al.

Water inrush disasters in karst areas have caused great losses to underground engineering construction, so it is urgent to control water gushing disasters in karst pipelines. In this paper, solution polymerization is used to prepare a grouting and sealing expanded matrix material to control disasters. Noncovalent weak interactions were used to improve the surface properties of the expanded matrix material, the effects of the natural polymer content on the properties of the matrix material were studied, and a modified expanded matrix material with an optimal response rate was prepared. A cross-linked curing agent (CCA) was developed and synthesized, and a new cross-linked expansive grouting and sealing material (WIS grouting material) with various particle sizes was synthesized with noncovalent interactions such as hydrogen bonding, static electricity, van der Waals forces, etc. The results showed that a 4 % solution polymer content (accounting for the particle mass of the expanded matrix) was the optimal dosage, and the optimal ratio of the modified expansion matrix material to the crosslinking curing agent was 1:1. The early compressive strength exceeded 0.2 MPa, and the water absorption rate reached 170 times. There was a power function relationship between the water absorption rate and time, and the rate was controlled by adjusting the particle sizes. The mechanism through which the WIS grouting material underwent expansion and crosslinking was explained at the microscopic level. The gel formed in response to water resisted dispersion in dynamic water and rapidly sealed karst pipe water gushers. This paper proposes a novel approach to utilizing the expansion characteristics of polymer chemical synthetic materials for crosslinking to seal karst pipe water gushers, effectively addressing the issue of poor resistance to dynamic water dispersion in traditional grouting materials used in karst areas. These results provide a scientific basis for the development and application of new materials to control water inrush in karst pipes.

Jia He, Wenjing Duan, Yuxuan Zhou et al.

Abstract Disaster information content is an objective mapping of disaster situations, social response, and public opinions. Social response to emergency is an important mechanism for implementing and guaranteeing emergency management of major natural hazard-related disasters. Understanding how disaster information content affects social response to emergencies is helpful for managing risk communication and efficient disaster response. Based on the 2008 freezing-rain and snowstorm disasters in southern China, this study used Python to extract 7,857 case-related media reports and applied natural language processing for text analysis. It used three typical cases to identify and analyze disaster media report content and the relationship between these reports and the social response to the emergency. Eight categories of disaster response—such as prewarning and forecasting, announcements by the authorities, and social mobilization—appeared in the disaster information in the media, along with disaster impact information, that is, real-time disaster status. Disaster response information and an appropriate amount of disaster impact information played important roles in prewarning, disaster relief, public opinion guidance, and social stability maintenance and can serve important functions in communicating with all stakeholders of emergency management, assisting or influencing emergency departments or individuals in decision making, and eliminating “information islands.” Empathy caused the general public to become “disaster responders” through receiving information. Rumors and an excess of negative information may have a perverse amplification effect on public opinion and increase the unpredictability of the disaster situation and the risk of social crisis.

Jinxuan Zhou, Shucheng Tan, Jun Li et al.

China is actively promoting the construction of clean energy to reach its objective of achieving carbon neutrality. However, engineering constructions in mountainous regions are susceptible to landslide disasters. Therefore, the assessment of landslide disaster susceptibility is indispensable for disaster prevention and risk management in construction projects. In this context, the present study involved conducting a field survey at 42 landslide points in the selected planned site region. According to the geological and geographical conditions of the study region, the existing regulation, and the influencing factors of landslides, the assessment in the field survey was performed based on 11 impact factors, namely, the slope, slope aspect, curvature, relative relief, NDVI, road, river, fault, lithology, the density of the landslide points, and the land-use type. Next, based on their respective influences, these impact factors were further divided into subfactors according to AHP, and the weights of each factor and subfactor were calculated. The GIS tools were employed for linear combination calculation and interval division, and accordingly, a landslide susceptibility zone map was constructed. The ROC curve was adopted to test the partition evaluation results, and the AUC value was determined to be 0.845, which indicated the high accuracy of the partition evaluation results.

Yue Wang, Deliang Sun, Haijia Wen et al.

To compare the random forest (RF) model and the frequency ratio (FR) model for landslide susceptibility mapping (LSM), this research selected Yunyang Country as the study area for its frequent natural disasters; especially landslides. A landslide inventory was built by historical records; satellite images; and extensive field surveys. Subsequently; a geospatial database was established based on 987 historical landslides in the study area. Then; all the landslides were randomly divided into two datasets: 70% of them were used as the training dataset and 30% as the test dataset. Furthermore; under five primary conditioning factors (i.e., topography factors; geological factors; environmental factors; human engineering activities; and triggering factors), 22 secondary conditioning factors were selected to form an evaluation factor library for analyzing the landslide susceptibility. On this basis; the RF model training and the FR model mathematical analysis were performed; and the established models were used for the landslide susceptibility simulation in the entire area of Yunyang County. Next; based on the analysis results; the susceptibility maps were divided into five classes: very low; low; medium; high; and very high. In addition; the importance of conditioning factors was ranked and the influence of landslides was explored by using the RF model. The area under the curve (AUC) value of receiver operating characteristic (ROC) curve; precision; accuracy; and recall ratio were used to analyze the predictive ability of the above two LSM models. The results indicated a difference in the performances between the two models. The RF model (AUC = 0.988) performed better than the FR model (AUC = 0.716). Moreover; compared with the FR model; the RF model showed a higher coincidence degree between the areas in the high and the very low susceptibility classes; on the one hand; and the geographical spatial distribution of historical landslides; on the other hand. Therefore; it was concluded that the RF model was more suitable for landslide susceptibility evaluation in Yunyang County; because of its significant model performance; reliability; and stability. The outcome also provided a theoretical basis for application of machine learning techniques (e.g., RF) in landslide prevention; mitigation; and urban planning; so as to deliver an adequate response to the increasing demand for effective and low-cost tools in landslide susceptibility assessments.

Yong-gang Zhang, Yuanlun Xie, Yan Zhang et al.

Y. Niu, E. Wang, Zhonghui Li et al.

Diyuan Li, Zida Liu, D. J. Armaghani et al.

The occurrence of rockburst can cause significant disasters in underground rock engineering. It is crucial to predict and prevent rockburst in deep tunnels and mines. In this paper, the deficiencies of ensemble learning algorithms in rockburst prediction were investigated. Aiming at these shortages, a novel machine learning model, deep forest, was proposed to predict rockburst risk. The deep forest combines the characteristics of deep learning and ensemble models, which can solve complex problems. To develop the deep forest model for rockburst prediction, 329 real rockburst cases were collected to build a comprehensive database for intelligent analysis. Bayesian optimization was proposed to tune the hyperparameters of the deep forest. As a result, the deep forest model achieved 100% training accuracy and 92.4% testing accuracy, and it has more outstanding capability to forecast rockburst disasters compared to other widely used models (i.e., random forest, boosting tree models, neural network, support vector machine, etc.). The results of sensitivity analysis revealed the impact of variables on rockburst levels and the applicability of deep forest with a few input parameters. Eventually, real cases of rockburst in two gold mines, China, were used for validation purposes while the needed data sets were prepared by field observations and laboratory tests. The promoting results of the developed model during the validation phase confirm that it can be used with a high level of accuracy by practicing engineers for predicting rockburst occurrences.

Ke Ma, Guoyang Liu