Nov 2026· Journal of computing in civil engineering· 0 citations· 28 references
TL;DR
This work presents an alternative, result-oriented, data-driven method based on generative AI to assist engineers in the conceptual design phase of bridge construction, and demonstrates that result-oriented, data-driven generative models can support early-stage bridge topology exploration under controlled conditions.
Abstract
The planning process for bridge construction is time-consuming, prone to errors, and cost-intensive, because bridges always have to be individually adapted to specific boundary conditions, for example, the respective topography situation. Despite sharing the same fundamental physical principles, bridge structures can take different forms due to regional variations in engineering expertise. Existing computational design approaches for bridges are predominantly rule-based or parametric. To support early-stage decision-making, successful practices from around the world could be analyzed using image-based, data-driven artificial intelligence (AI) methods. Hence, this work presents an alternative, result-oriented, data-driven method based on generative AI to assist engineers in the conceptual design phase of bridge construction. First, the research background and related works in structural design, with a focus on bridge engineering, are elaborated. Subsequently, we propose a multimodal generative model combining pix2pix and BERT to map valley topographies and textual design specifications to conceptual bridge topology configurations. The approach leverages image-based terrain information and text-based design descriptors to infer structurally plausible configurations from precedent data, rather than relying on explicitly formulated design rules. The results show that the model can reproduce multiple topology patterns consistent with the training data and generate constraint-consistent configurations for topographies with similar geometric characteristics. The findings demonstrate that result-oriented, data-driven generative models can support early-stage bridge topology exploration under controlled conditions, providing a complementary alternative to parametric design workflows. The paper concludes by outlining methodological implications and directions for extending the approach toward broader applicability.
Implementing Historic Building Information Modelling (HBIM) for heritage structures is challenged by incomplete knowledge of hidden or inaccessible elements, as well as limited information on construction history, original design, and structural details, making geometric definition inherently uncertain. Simultaneously, applications such as structural analysis often require the same missing information. This study proposes an adaptive parametric Scan-to-BIM-to-FEM workflow that explicitly incorporates geometric uncertainty by generating multiple plausible and complete reconstructions from survey data and typological inference, enabling their use in parametric structural analysis. Starting from TLS survey, adaptive families were used to link measured and inferred dimensions through geometric constraints. The methodology is applied to the 19th-century masonry arch bridge of Montoggio (Genoa, Italy), currently characterized by a hybrid structural system resulting from subsequent retrofitting. Eight geometrical configurations were tested by varying uncertain parameters, including vault thickness and backing height, and were analyzed through modal and static finite element simulations. The results show a limited but non-negligible sensitivity of the structural response to these parameters, highlighting the influence of geometric uncertainties on analysis outcomes. In this light, the proposed framework provides a bridge between survey, modelling, and structural analysis, enabling HBIM to support interpretative and predictive structural assessment.
G. Sacco, Matilde Ridella, C. Calderini· Buildings· 0 citations
The accelerating digitalization of the construction sector has brought into sharp focus the inefficiencies and labor-intensive nature inherent in traditional craftsmanship for historical architecture. In response, this study proposes an AI-driven generative design workflow that translates semantic inputs into 2D imagery and 3D models, enabling the systematic learning and replication of stylistic features from a quintessential southern Chinese architectural ornament—the Lingnan stucco relief. This approach moves beyond the conventional reliance on individual expertise and heuristic empirical methods. The structural performance of the AI-generated designs was evaluated through combined material testing and numerical simulations. Concurrently, hybrid additive manufacturing techniques were explored for rapid physical prototyping. Collectively, this integrated framework illustrates the feasibility of revitalizing traditional craft practices within a contemporary intelligent paradigm, offering a technological pathway for the preservation and renovation of historic built heritage.
Yiwei Yin, Jiale Cheng, Jiashao Zhou et al.· AI in Civil Engineering· 0 citations
Existing BIM-based workflows for complex spatial steel bridges often support 3D visualization and documentation, but they still lack a formal requirement-to-function traceability mechanism and a reliable automated link from special-shaped surface modeling to fabrication-oriented data. To address this gap, this study develops and validates a Model-Based Systems Engineering (MBSE)-oriented design–fabrication integration workflow for special-shaped steel bridges. The workflow combines requirement decomposition, functional architecture modeling, ENOVIA-based collaborative data management, skeleton-driven parametric modeling, User-Defined Feature (UDF) templates, Engineering Knowledge Language (EKL) batch instantiation, an IFC-based manufacturing information extension, ProNest nesting, and model-driven NC-code generation. The method was implemented for the Q7 North Pedestrian Bridge, a spatially twisted special-shaped steel landscape bridge. In the case study, the proposed workflow reduced typical repetitive component modeling time by 70.8%, shortened drawing generation time by 80.0%, increased nesting material utilization from 84.6% to 91.8%, and controlled the maximum coordinate-transformation deviation of formwork points within 1.42 mm. Field validation showed a mean fabrication deviation of 1.6 mm and a maximum site assembly closure deviation of 4.5 mm. The results indicate that the proposed MBSE-oriented digital thread improves design consistency, reduces manual data re-entry, and strengthens traceability from requirements to manufacturing and assembly. The study provides a reproducible case-study framework for model-driven steel bridge design–fabrication integration and identifies the limitations of UDF-library construction cost, software-specific learning requirements, and single-project validation.
Xiang Guo, Yongyi Yang, Wei Liu et al.· Metals· 0 citations
Efficient bridge scanning and documentation are crucial for creating reliable digital 3D models. However, scanning workflows often rely on implicit practitioner experience rather than standardized protocols. This paper presents practical insights derived from a Multiple Case Study (MCS) of ten heterogeneous, real-world bridges in Germany. The study evaluates Terrestrial Laser Scanning (TLS), Mobile Laser Scanning (MLS) and Unmanned Aerial Systems (UAS) photogrammetry. The findings isolate distinct performance trade-offs. TLS offers high accuracy but suffers from shadowing occlusions. Conversely, UAS provides operational flexibility but introduces geometric vulnerabilities, including photogrammetric reconstruction noise on fine structures and SLAM trajectory drift on vibrating spans. To unify these insights, a generalized, BPMN-compliant process model mapping the complete data acquisition lifecycle under legal and spatial constraints is defined. This research provides an actionable, practical guide to optimize data quality and efficiency in structural engineering workflows.
Monika Lederer, Christoph Stahl, Jan-Iwo Jäkel et al.· Remote Sensing· 0 citations
This paper shows an AI-assisted workflow supported by a Large Language Model for the geometric and informative reconstruction of masonry arch bridges from multi-source input data. The proposed methodology combines point cloud preprocessing, vegetation filtering, AI-based or semi-automatic segmentation, geometric feature extraction, 3D mesh generation, and parametric Heritage Building Information Modeling integration. The study specifically aims to examine how the quantity and type of input information influence the reconstruction process and the reliability of the resulting HBIM model. The input data considered include Point Clouds (PC), photographic images, and descriptive information related to the bridge geometry, construction features, and visible architectural components. Particular attention is given to the initial geometric characteristics of the point cloud, including its point density, spatial distribution, level of completeness, and overall number of points, to assess how point cloud numerosity affects the accuracy and level of detail of the reconstructed geometry. Starting from a dense 3D survey, the workflow identifies and reconstructs the main architectural and structural components of a masonry arch bridge, including the arch, intrados, parapets, masonry walls, roadway surface, abutments, and cutwaters. The extracted geometry is converted into a clean 3D mesh and subsequently structured into parametric HBIM objects suitable for documentation, conservation, structural assessment, and future monitoring activities. A Cloud-to-Mesh comparison is performed to evaluate the geometric accuracy of the reconstructed model with respect to the original point cloud. The results demonstrate the potential of hybrid AI and geometric approaches to improve the efficiency, repeatability, and reliability of Scan-to-BIM processes for historical masonry bridge heritage. Furthermore, they show that the geometric quality of the HBIM model depends primarily on the density, spatial distribution and completeness of the structural points, rather than on their total number. Well-distributed point clouds, in fact, allow for more reliable reconstructions than larger datasets characterised by uneven coverage.
V. Alfio, Massimiliano Pepe, Donato Palumbo et al.· Applied Sciences· 0 citations
Integrating artificial intelligence (AI) into the built environment has greatly improved building automation and performance optimization. However, the potential of AI to inform early spatial configuration, user-oriented planning, and its role in shaping architectural decisions at the conceptual design stage have not been well studied. This study aims to compare traditional architectural design processes with AI-informed design approaches at the conceptual stage. An experimental comparative design model is employed in which two parallel design scenarios are developed for the same prototype: a conventional, architect-led concept design and an AI informed concept design. The comparative evaluation framework is structured around five multidimensional criteria: Space Utilization Efficiency, Daylight Performance, Circulation Optimization, User Scenario Compatibility, and Spatial Adaptation Capacity. This study examines whether AI-informed conceptual design generates spatial configurations that differ measurably from conventional approaches. The findings will contribute to the development of data-driven, adaptive, and user-centered architectural methodologies.
Sanam Rezaeifam, Seyed Babak Ehsani Oskouei, Gökçen Firdevs Yücel Caymaz· Proceedings of the internati...· 0 citations
The professor of physics and inaugural director of the NSF AI Institute for Artificial Intelligence and Fundamental Interactions will lead LNS and continue his research in particle physics.
