BIM, machine learning and computer vision techniques in underground construction: current status and future perspectives

M. Q. Huang, J. Ninić, Q. B. Zhang

Research output: Contribution to journalArticleResearchpeer-review

Abstract

The architecture, engineering and construction (AEC) industry is experiencing a technological revolution driven by booming digitisation and automation. Advances in research fields of information technology and computer science, such as building information modelling (BIM), machine learning and computer vision have attracted growing attention owing to their useful applications. At the same time, population-driven underground development has been accelerated with digital transformation as a strategic imperative. Urban underground infrastructures are valuable assets and thus demanding effective planning, construction and maintenance. While enabling greater visibility and reliability into the processes and subsystems of underground construction, applications of BIM, machine learning and computer vision in underground construction represent different sets of opportunities and challenges from their use in above-ground construction. Therefore, this paper aims to present the state-of-the-art development and future trends of BIM, machine learning, computer vision and their related technologies in facilitating the digital transition of tunnelling and underground construction. Section 1 presents the global demand for adopting these technologies. Section 2 introduces the related terminologies, standardisations and fundamentals. Section 3 reviews BIM in traditional and mechanised tunnelling and highlights the importance of integrating 3D geological modelling and geographic information system (GIS) databases with BIM. Section 4 examines the key applications of machine learning and computer vision at different stages of underground construction. Section 5 discusses the challenges and perspectives of existing research on leveraging these emerging technologies for escalating digitisation, automation and information integration throughout underground project lifecycle. Section 6 summarises the current state of development, identified gaps and future directions.

Original languageEnglish
Article number103677
Number of pages40
JournalTunnelling and Underground Space Technology
Volume108
DOIs
Publication statusPublished - Feb 2021

Keywords

  • Building information modelling
  • Computer vision
  • Machine learning
  • Tunnelling
  • Underground construction

Cite this