Pablo Araya-Santelices
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Pablo Araya-Santelices has a total of 12 co-authors. Here are the top co-authors by number of shared publications:
Edison Atencio
2 shared publications- Edison Atencio
School of Civil Engineering, Pontificia Universidad Católica de Valparaíso, Valparaíso 2362804, Chile
2 shared publications - Fidel Lozano-Galant
Department of Civil Engineering, Universidad de Castilla-La Mancha, Av. Camilo Jose Cela s/n, 13071 Ciudad Real, Spain
2 shared publications Zacarías Grande
2 shared publications- Felipe Muñoz-La Rivera
School of Civil Engineering, Pontificia Universidad Católica de Valparaíso, Valparaíso 2340000, Chile
1 shared publication - Francisco Oyarce
School of Civil Engineering, Pontificia Universidad Católica de Valparaíso, Av. Brasil 2147, Valparaíso 2340000, Chile
1 shared publication
Latest Publications
- Open Access
- 16 June 2026
- Published by Elsevier in SSRN Electronic Journal
Synthetic Dataset Generation with Multimodal Generative Image Models for Reinforced-Concrete Surface Crack Segmentation
Pixel-level crack segmentation in reinforced concrete is constrained by scarce annotations and limited transferability of public datasets to real inspection domains. This study proposes a workflow for generating synthetic crack datasets using multimodal generative image models, real concrete textures, spatial masks, parameterised prompts, morphological control and photographic variation. The generated dataset was benchmarked against the Concrete Crack Conglomerate dataset and a 50/50 Hybrid scenario combining public and synthetic images. Three convolutional neural network configurations were trained under equivalent conditions and evaluated on a real multi-bridge test set acquired with unmanned aerial vehicles, comprising one texture-related bridge and four external bridges. The Synthetic scenario achieved the highest observed average performance in all-bridges and external-bridges crack-positive subsets, with macro F1/mIoU (mean Intersection over Union) values of 0.809/0.729 and 0.793/0.712. Results indicate that real-texture-conditioned synthetic data can partly mitigate domain mismatch and support domain-adapted dataset preparation for semantic crack segmentation.
- 16 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessSynthetic Dataset Generation with Multimodal Generative Image Models for Reinforced-Concrete Surface Crack Segmentation
Pixel-level crack segmentation in reinforced concrete is constrained by scarce annotations and limited transferability of public datasets to real inspection domains. This study proposes a workflow for generating synthetic crack datasets using multimodal generative image models, real concrete textures, spatial masks, parameterised prompts, morphological control and photographic variation. The generated dataset was benchmarked against the Concrete Crack Conglomerate dataset and a 50/50 Hybrid scenario combining public and synthetic images. Three convolutional neural network configurations were trained under equivalent conditions and evaluated on a real multi-bridge test set acquired with unmanned aerial vehicles, comprising one texture-related bridge and four external bridges. The Synthetic scenario achieved the highest observed average performance in all-bridges and external-bridges crack-positive subsets, with macro F1/mIoU (mean Intersection over Union) values of 0.809/0.729 and 0.793/0.712. Results indicate that real-texture-conditioned synthetic data can partly mitigate domain mismatch and support domain-adapted dataset preparation for semantic crack segmentation.
- Open Access
- 29 May 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (11) , 6107
BIM-GIS-Based Approach for Quality Management Aligned with ISO 9001
Quality management during construction is critical to ensuring compliance with technical specifications and quality standards. Traditional practices often rely on manual, paper-based documentation, leading to inefficiencies, data fragmentation, and poor traceability. This study presents QualiSite, a novel digital workflow that integrates Building Information Modeling (BIM) and Geographic Information Systems (GIS), aligned with ISO 9001:2015 requirements, to enhance quality management in building projects. The research is framed under the Design Science Research Method (DSRM), guiding the iterative development and validation of the tool. QualiSite was tested in a real-world case study involving the construction of reinforced concrete walls. The results demonstrated functional improvements in inspection traceability, consistency of quality records, and coordination between field data and BIM elements. Using structured digital forms contributed to more consistent data capture and greater efficiency in recording, organizing, and visualizing quality control statuses within the 3D environment. These outcomes enabled transparent inspection processes and clear visualization of quality status across construction elements. The digital workflow also facilitated the identification of nonconformities and streamlined communication between field inspectors and model managers. This approach advances traditional quality management by embedding inspection records into a Cyber-Physical Systems (CPS) framework, contributing to the digital transformation of the Architecture, Engineering, and Construction (AEC) industry and supporting the vision of Smart Industry.
- 29 May 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (11) , 6107
Open AccessBIM-GIS-Based Approach for Quality Management Aligned with ISO 9001
Quality management during construction is critical to ensuring compliance with technical specifications and quality standards. Traditional practices often rely on manual, paper-based documentation, leading to inefficiencies, data fragmentation, and poor traceability. This study presents QualiSite, a novel digital workflow that integrates Building Information Modeling (BIM) and Geographic Information Systems (GIS), aligned with ISO 9001:2015 requirements, to enhance quality management in building projects. The research is framed under the Design Science Research Method (DSRM), guiding the iterative development and validation of the tool. QualiSite was tested in a real-world case study involving the construction of reinforced concrete walls. The results demonstrated functional improvements in inspection traceability, consistency of quality records, and coordination between field data and BIM elements. Using structured digital forms contributed to more consistent data capture and greater efficiency in recording, organizing, and visualizing quality control statuses within the 3D environment. These outcomes enabled transparent inspection processes and clear visualization of quality status across construction elements. The digital workflow also facilitated the identification of nonconformities and streamlined communication between field inspectors and model managers. This approach advances traditional quality management by embedding inspection records into a Cyber-Physical Systems (CPS) framework, contributing to the digital transformation of the Architecture, Engineering, and Construction (AEC) industry and supporting the vision of Smart Industry.
- 9 April 2025
- Published by Elsevier in Automation in Construction
- Vol. 175
Bridge management with AI, UAVs, and BIM
- 9 April 2025
- Published by Elsevier in Automation in Construction
- Vol. 175
Bridge management with AI, UAVs, and BIM
- Open Access
- 9 September 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (18) , 9063
Towards the Integration and Automation of the Design Process for Domestic Drinking-Water and Sewerage Systems with BIM
The use of building information modelling (BIM) in construction projects is expanding, and its usability throughout building lifecycles, from planning and construction to operation and maintenance, is gaining increasing proof. In the design of domestic drinking-water and sewerage systems (DDWSSs), BIM focuses on coordinating disciplines and their design. Despite studies promoting BIM environments for DDWSSs that take into account the regulatory frameworks of corresponding countries, these efforts do not include the use of parametric tools that enhance the efficiency of the design process. Therefore, engineers still use conventional 2D design, which requires many rounds of iteration, and manual work is also generally still used. In this research, we developed and validated an intuitive methodology for solving a specific DDWSS problem, using a design science research method (DSRM) as an applied science approach. This was addressed by developing an artefact and validating it through two case studies. The obtained solution combines BIM models and parametric tools to automate the manual activities of the traditional design method. This article aims to bring abstract BIM concepts into practice and encourage researchers and engineers to adopt BIM for DDWSSs.
- 9 September 2022
- Published by MDPI AG in Applied Sciences
- Vol. 12 (18) , 9063
Open AccessTowards the Integration and Automation of the Design Process for Domestic Drinking-Water and Sewerage Systems with BIM
The use of building information modelling (BIM) in construction projects is expanding, and its usability throughout building lifecycles, from planning and construction to operation and maintenance, is gaining increasing proof. In the design of domestic drinking-water and sewerage systems (DDWSSs), BIM focuses on coordinating disciplines and their design. Despite studies promoting BIM environments for DDWSSs that take into account the regulatory frameworks of corresponding countries, these efforts do not include the use of parametric tools that enhance the efficiency of the design process. Therefore, engineers still use conventional 2D design, which requires many rounds of iteration, and manual work is also generally still used. In this research, we developed and validated an intuitive methodology for solving a specific DDWSS problem, using a design science research method (DSRM) as an applied science approach. This was addressed by developing an artefact and validating it through two case studies. The obtained solution combines BIM models and parametric tools to automate the manual activities of the traditional design method. This article aims to bring abstract BIM concepts into practice and encourage researchers and engineers to adopt BIM for DDWSSs.