An Ontology to Enable Semantic BIM-Based Data Quality Assurance and Quality Control
Keywords:QA & QC, ontology, Artificial intelligence, knowledge-based system, Protégé platform, automated compliance checking, ETS pavilion F project, BIM One
Quality assurance (QA) and quality control (QC) are two central and concurrent processes for data quality management. Together, they enable construction project participants to ensure high quality output by meeting client requirements and complying with a range of building regulations. QA is a proactive process used iteratively from the beginning of the project, while QC is reactive to be triggered to verify the product quality regarding the delivery requirements. When implemented in a BIM strategy, QA-QC aim at ensuring the quality of BIM data throughout the entire lifecycle of a construction project (Leygonie, et al. 2022). Over the last fifty years, several research works, methodologies and tools have been developed to optimize QA-QC processes and automate their execution, thus avoiding manual, time consuming, and error-prone implementation which can result in financial losses and project failure (Doukari et al. 2022). Most of these approaches have relatively succeeded, in different domain applications, to satisfy a set of desired compliance checking characteristics and enable syntactic data quality checks, such as data (objects, relations, attributes and values) completeness, structure and naming verifications. However, to the best of our knowledge, there is no comprehensive framework for data quality validation which proposes efficient semantic QA-QC processes that should automatically enable semantic compliance checking, such as data relevance, and ensure the overall accuracy of the project data. Conventional practice still consists in involving human experts while conducting semantic QA-QC processes. Moreover, there is little research that addresses how QA-QC can best be coupled and used with Artificial intelligence (AI) technologies to optimise and develop more effective strategies for better project data management (Sacks et al. 2019). A combination of BIM and AI could exploit the massive and increasing amount of data and knowledge (e.g., project feedback, domain expertise) that are increasingly obtainable. A prerequisite for doing so is the development of a dedicated ontology that would enable the formalisation of domain knowledge, including associated concepts, relations, and constraints that are specific to QA-QC processes (Amorocho and Hartmann 2021). This knowledge is acquired directly or indirectly from domain experts in different expertise areas. Furthermore, ontologies enable process automation and tools development as they provide a machine-readable representation of knowledge. This project proposes the development of a QA-QC ontology to support BIM-based projects and ensure both syntactic and semantic high quality BIM data. Also, it aims at demonstrating the ontology application by developing a knowledge-based system for application within the context of a new construction which is the ‘ETS Pavilion F’ project in Montreal. Knowledge-based systems are known to be ‘intelligent’ due to their inherent ability to mimic the human decision process while reasoning and using domain specific knowledge (Doukari and Greenwood 2020). The digital tool proposed will provide support to project participants lacking QA-QC knowledge, and so substitutes for human experts. In addition, this system will be equipped with learning capabilities so that its knowledge can be updated and extended continuously.
Copyright (c) 2022 University of New Brunswick
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors/employers retain all proprietary rights in any process, procedure or article of manufacture described in the Work. Authors/employers may reproduce or authorize others to reproduce the Work, material extracted verbatim from the Work, or derivative works for the author’s personal use or for company use, provided that the source is indicated.