This paper explores how model updates can be effectively facilitated for integration into models of foreign disciplines. Graph transformations are used to capture and describe the model changes introduced by one domain. These transformations are then interpreted to suggest appropriate updates for models in other disciplines. Repetitive and predictable model updates can be automated, with the necessary modifications presented to users as actionable alteration proposals. The proposed workflow reduces the burden on model authors by automating tedious yet essential tasks, allowing them to focus on solving complex design coordination challenges.
Mining construction in Chile faces significant productivity challenges, largely due to inefficiencies in equipment operation. Therefore, improving operators training is essential. This paper proposes the use of virtual reality (VR) technology as an innovative training tool for mining construction operators. This approach aims to enhance operator proficiency and ultimately improve productivity in underground mining projects. The proposed VR-enabled training program is implemented in two case studies covering two underground mining construction equipment. The results obtained demonstrated that the proposed program yield positive impacts on operator’s productivity, indicating that VR could be a useful tool for boosting job performance.
Solid waste generated from construction activities pose serious environmental challenges and adversely affect the sustainability of the construction industry. Advancements in data-driven approaches, including computer vision (CV), offer a potential approach to automating the process of waste quantification on construction sites. This study proposes a system that utilises CV to quantify waste generated onsite. The proposed framework consists of analysis video streams of a waste container and estimating the waste material based on predefined waste categories. This study advances automated construction waste detection and identification and ultimately promoting efficient construction waste management, circular economy, and sustainable development.
Planning modern facade systems is complex, requiring optimization across multiple domains.This paper proposes an AI-enhanced workflow for facade planning, harnessing computer vision and human input via a Large Language Model.A generative AI system then guides a parametric model to produce 3D facade designs. Automated checks provide feedback to a Reinforcement Learning system, to iteratively determine optimal solutions.These solutions are verified and finalized by human expertise, ensuring improved outcomes with reduce planning time and effort.The approach illustrates how combining advanced AI methods with human expertise can address the multifactorial challenges of facade design within current industry practices.
Cost estimation in building industry largely relies on manually extracting and classifying textual descriptions, a process susceptible to human error. Although recent advancements in Large Language Models (LLMs) hold promise, their application in this domain requires further investigation.This study proposes a methodology to optimize LLM performance validated through the development of a tool that classifies cost descriptions into a three-level hierarchical taxonomy and extracts relevant information organising the data in a database as output. Results demonstrate a F1 score of 0.96 on classification tasks contributing to cost estimation automation, reducing manual processing, and enhancing knowledge management within the domain.
The analog nature of bridge inspections hinders long term prognosis of structural health. Accordingly, we propose a comprehensive digitization concept to enhance the effectiveness of bridge inspections: historical inspection data is consolidated into digital damage models with machine learning. These models can then be combined with a BIM model for the creation of a digital twin, which paves the way for an evaluation of damage progression. The model furthermore enables inspectors to easily locate existing damages via images and point clouds, and to record changes into a new damage model. This significantly improves maintenance and retrofitting efforts.
Railway maintenance, particularly in curved sections, is a complex and costly operation requiring effective solutions to minimise wear. Machine Learning (ML) models were developed to monitor curved tracks using accelerometer data for lubrication level prediction, sensor fault detection, and outlier analysis. To support this, a monitoring system designed within a Machine Learning Operations (MLOps) framework was implemented. While vibration alone proved inconclusive in predicting lubrication levels, a significant increase in outliers during lubrication system inactivity indicated potential rail issues. As a result, the solution offers near real-time insights, helping end users make data-driven maintenance decisions by presenting model outcomes.
Due to the need for effective track and trace systems to enable circular construction supply chains (CCSC), this study evaluated the performance of radio-frequency identification (RFID), near-field communication (NFC) chips, quick response (QR) codes and Direct Product Marking (DPM) in identifying and reading material information. Key metrics - detection speed, error rates, and user experience - were assessed and gave consideration to the differential impacts of controlled and uncontrolled experimental conditions. Findings indicate that, relative to DPM, RFID and NFC offered improved usability and reduced reading times, with minimal differences in error rates; results were strongly influenced by usability and experimental context.
Automated Compliance Checking (ACC) facilitates rapid and objective review of building permits. The present study, based on in-depth qualitative expert interviews with regulators and ACC experts worldwide, provides an overview of the current ACC advances in place within or linked to regulatory bodies. The interviewees highlighted key challenges, including ambiguous terminology, inconsistencies, and the balance between human oversight and algorithmic decision-making. A comparative analysis of current practices across different countries offers insights into their lessons learned, future plans, and additional research needs.
