22 results
Now showing1 - 10 of 22
- Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Prescriptive maintenance of sewer systems based on large language models(2026-07); ; ; Sewer systems constitute urban infrastructure critical for public health, environmental protection, and flood prevention. Decision-making in sewer systems maintenance often fails to effectively and efficiently interpret heterogeneous and multimodal inspection data into actionable maintenance strategies. To improve decision-making in sewer systems maintenance, this paper introduces a prescriptive maintenance (RxM) framework, in which a pre-trained large language model (LLM) is adapted to the domain of sewer systems using adapter-based fine-tuning. As will be shown in this paper, the LLM is capable of interpreting heterogeneous, multimodal inspection data and recommending stepwise repair procedures of sewer systems, following the paradigm of RxM.Publicationtype: Conference Paper8 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , A decentralized digital twinning approach for assessing civil engineering structures(2026-07); ; In wireless structural health monitoring (SHM), the focus is usually placed on reducing the power-consuming data transmission for preserving the limited resources of wireless sensor nodes. In this direction, embedded computing has been leveraged for distributing SHM data analysis tasks to the sensor nodes instead of transmitting raw data to centralized servers. This paper reports on decentralizing SHM tasks using reduced-order numerical models of structures, embedded into wireless sensor nodes as decentralized digital twins that use sensor data to assess the structural condition. The applicability of the proposed approach is demonstrated via laboratory validation tests.Publicationtype: Conference Paper7 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Digital twins in civil engineeringA plenitude of digital twin reviews have been published in recent years, most of which, however, focus on project-specific digital twin implementations or specific application areas within civil engineering. The architecture, including the composition of internal elements, and schemas of digital twins have not been studied yet in a universally applicable and systematic manner. In this study, a multivocal review of digital twins in civil engineering is conducted, combining a systematic literature review and a gray literature review to capture both peer-reviewed literature and industry practices. To achieve a shared understanding of digital twins in civil engineering, a systematic survey of definitions, schemas, system architectures, and internal elements of digital twins is conducted. Drawing from the results of the review, (i) a definition, (ii) a pattern for a schema, and (iii) a reference architecture for digital twins are proposed, serving as a blueprint for facilitating digital twin implementations in civil engineering.Publicationtype: Journal Article19 7 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Decentralized sensor fault diagnosis for wireless structural health monitoring systems using Artificial Intelligence of Things(DEStech Publications, 2025-09); ; ; ; Structural health monitoring (SHM) is a non-destructive evaluation technique that utilizes sensor data for assessing the condition of civil infrastructure. Sensors in SHM systems may experience faults, which may influence the accuracy, reliability, and performance of SHM systems. The timely detection, isolation, and accommodation of sensor faults in SHM systems has been the focus of sensor fault diagnosis (FD) approaches, which have increasingly been employing artificial intelligence (AI) algorithms due to the effectiveness of AI in sensor FD. However, current AI-based FD approaches require transmitting large amounts of raw sensor data to centralized servers for offline analysis, resulting in inefficiencies as well as computational burdens on centralized servers. This paper introduces a decentralized sensor fault diagnosis (DSFD) approach for wireless SHM systems using Artificial Intelligence of Things (AIoT). In particular, AI-based FD models are embedded into wireless sensor nodes of SHM systems to detect, isolate, and accommodate sensor faults. By embedding the FD models into the wireless sensor nodes, only high-level information, specifically the status of the sensors, is transmitted to centralized servers. As a result, data transmission inefficiencies as well as computational burdens on centralized servers are reduced. The proposed DSFD approach is validated in a controlled laboratory experiment, in which custom- built wireless senor nodes are installed on a test structure that is dynamically excited using a shake table. After training and embedding the AI-based FD models into the custom-built wireless sensor nodes, sensor faults are artificially injected into the sensor data, demonstrating the ability of the DSFD approach to diagnose sensor faults in a decentralized manner. The results of the validation test corroborate the capability of the proposed approach to efficiently ensure the accuracy, reliability, and performance of SHM systems.Publicationtype: Conference Paper5 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Decision making in structural health monitoring using large language models(DEStech Publications, 2025-09); ; ; ; Structural health monitoring (SHM) is fundamental in decision making for damage identification and prescriptive maintenance of civil infrastructure. Traditional decisionmaking methods often fall short in integrating heterogeneous sensor data, inherent to SHM, and the natural language processing required to interpret, e.g., inspection reports, expert assessments, and historical documentation relevant to prescriptive maintenance. To address these limitations, this study introduces a framework for integrating large language models (LLMs) into SHM workflows, facilitating decision making in damage identification and prescriptive maintenance. The proposed framework links convolutional neural networks (CNNs) with generative LLMs. Unlike approaches based solely on prompt engineering, this study applies task-specific fine-tuning via low-rank adaptation (LoRA) to the Mistral-7B-Instruct-v0.1 model, using CNN-generated output and damage metadata as input. The results demonstrate – apart from the proof of concept – a successful generalization to previously unseen CNN-generated output, enabling context-sensitive damage identification. In conclusion, integrating LLMs into SHM workflows allows synthesizing heterogeneous sensor data and natural language, thus enhancing decision making for damage identification and prescriptive maintenance of civil infrastructure.Publicationtype: Conference Paper20 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Digital-twin-based management of sewer systems: research strategy for the KaSyTwin project(Multidisciplinary Digital Publishing Institute (MDPI), 2025-01-22); ; ; ; Sewer infrastructure is vital for flood prevention, environmental protection, and public health. As part of sewer infrastructure, sewer systems are prone to degradation. Traditional maintenance methods for sewer systems are largely manual and reactive and rely on inconsistent data, leading to inefficient maintenance. The KaSyTwin research project addresses the urgent need for efficient and resilient sewer system management methods in Germany, aiming to develop a methodology for the semi-automated development and utilization of digital twins of sewer systems to enhance data availability and operational resilience. Using advanced multi-sensor robotic platforms equipped with scanning and imaging systems, i.e., laser scanners and cameras, as well as artificial intelligence (AI), the KaSyTwin research project focuses on generating digital twin-enabled representations of sewer systems in real time. As a project report, this work outlines the research framework and proposed methodologies in the KaSyTwin research project. Digital twins of sewer systems integrated with AI technologies are expected to facilitate proactive maintenance, resilience forecasting against extreme weather events, and real-time damage detection. Furthermore, the KaSyTwin research project aspires to advance the digital management of sewer systems, ensuring long-term functionality and public welfare via on-demand structural health monitoring and non-destructive testing.Publicationtype: Journal Article19 23 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Automated defect detection in fused filament fabrication coupling deep learning and computer visionFused filament fabrication (FFF) is an additive manufacturing technique, popular due to its versatility and cost-effectiveness. However, FFF machines, such as 3D printers, are prone to runtime errors, wasting time and material, while requiring constant human supervision. This paper presents a defect detection approach for FFF processes based on artificial intelligence, combining convolutional neural networks and computer vision. The defect detection approach is validated using 3D prints designed to mimic common FFF defects. The results demonstrate the capability of the proposed approach to automatically detect defects in FFF processes, thereby reducing time and material waste as well as the need for human supervision.Publicationtype: Conference Paper20 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Maschinelles Lernen zur Schadenserkennung bei Leichtbaubrücken mit beweglichen Massen(Technische Universität Hamburg, Institut für Digitales und Autonomes Bauen, 2024-09-18); ; Leichtbaubrücken sind Fahrzeugverkehr ausgesetzt, bei dem die Gewichte der Fahrzeuge – insbesondere bei Schwerverkehr – häufig mit dem Eigengewicht der Brücke vergleichbar sind. Diese beweglichen Massen führen zu einem komplexen dynamischen Verhalten, das über herkömmliche Strategien zur Schadenserkennung nur unzureichend analysiert werden kann. Dieser Beitrag beschreibt ein Konzept zur Schadenserkennung bei Leichtbaubrücken unter Verwendung von maschinellem Lernen (ML). Es wird ein ML-Modell zur Klassifizierung von Bauwerksschäden entwickelt, unter Verwendung von Simulationen traniniert und mittels realen Messdaten in einem Laborversuch validiert. Die Validierung zeigt, dass das ML-Modell geeignet ist, Bauwerksschäden an Leichtbaubrücken zu identifizieren, auch unter variierenden dynamischen Bedingungen, die durch die beweglichen Massen verursacht werden.Publicationtype: Conference Paper24 45 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: without files , Digital twins, architectures, and elements in civil engineering – A multivocal literature review(Springer, 2024-08); ; ; ; A plenitude of digital twin reviews have been published in recent years, most of which, however, solely review project-specific digital twin implementations or focus on specific application areas within civil engineering. Digital twin architectures, including the composition of the internal elements of digital twins, have received little attention so far. A common understanding of digital twin architectures and the composition of their elements would advance digital twin implementations, enhancing the reliability and performance of digital twins. In this study, a multivocal review of digital twins in civil engineering is conducted, aiming to provide a scientific basis for reaching a consensus on the abstract architecture and the composition of the internal elements of digital twins. Methodologically, the review is conducted in a twofold manner, accounting for both peer-reviewed, indexed literature, as well as for material that may report on research and development projects and industrial applications not indexed in well-accepted citation databases. The main contribution of this study is a systematic overview of existing architectures of digital twins implemented in civil engineering, including the internal elements. Furthermore, a generic reference architecture of digital twins is proposed, and digital twins are defined in both an informal and formal (i.e., mathematical) manner. It is expected that the reference architecture and the definitions proposed in this study will serve as a blueprint for digital twin implementations in civil engineering, providing a common understanding and enhancing the reliability and performance of digital twins.Publicationtype: Conference Paper45 - Some of the metrics are blocked by yourconsent settings
Item type:Publication, Access status: with files , Identification of combined sensor faults in structural health monitoring systems(IOP Publishing, 2024-07-19); ; ; ; Fault diagnosis (FD), comprising fault detection, isolation, identification and accommodation, enables structural health monitoring (SHM) systems to operate reliably by allowing timely rectification of sensor faults that may cause data corruption or loss. Although sensor fault identification is scarce in FD of SHM systems, recent FD methods have included fault identification assuming one sensor fault at a time. However, real-world SHM systems may include combined faults that simultaneously affect individual sensors. This paper presents a methodology for identifying combined sensor faults occurring simultaneously in individual sensors. To improve the quality of FD and comprehend the causes leading to sensor faults, the identification of combined sensor faults (ICSF) methodology is based on a formal classification of the types of combined sensor faults. Specifically, the ICSF methodology builds upon long short-term memory (LSTM) networks, i.e. a type of recurrent neural networks, used for classifying ‘sequences’, such as sets of acceleration measurements. The ICSF methodology is validated using real-world acceleration measurements from an SHM system installed on a bridge, demonstrating the capability of the LSTM networks in identifying combined sensor faults, thus improving the quality of FD in SHM systems. Future research aims to decentralize the ICSF methodology and to reformulate the classification models in a mathematical form with an explanation interface, using explainable artificial intelligence, for increased transparency.Publicationtype: Journal Article10 167