Back to Member Hub

Construction Data Standards and AI: Ontologies and Crosswalks (IFC, Uniclass, COBie)

🔒 This course requires registration

To access this course and all our learning materials, please register for the AI Fluency programme.

Register Now →

Construction data does not join up. IFC, Uniclass 2015, COBie and BOT each describe the same building in different vocabularies, and the same word means different things in each.

This course explains those standards in plain language, then shows how to build and check a crosswalk between them, including the false friends that break AI classification silently.

15 lessons with a graded quiz on every lesson. Free and open to access.

What you will learn

  • Read IFC, Uniclass 2015, COBie and BOT without a specification in front of you
  • Recognise a good crosswalk and the false friends that wreck a bad one
  • Understand why falsifiability matters more than a clean validation report
  • See where AI classification fails on construction data without a safety net
  • Apply an open, published crosswalk to your own project

Course curriculum

15 lessons, each with a graded quiz.

Module 1: Why the data does not join up

The joining problem, and what an ontology means in construction terms.

Module 2: The four standards

IFC, Uniclass 2015, COBie and BOT, each in plain language.

Module 3: Crosswalks and how they fail

What a good crosswalk looks like, the false friend problem, and falsifiability.

Module 4: AI on construction data

The zero percent problem, part versus kind, and the element, system and product strata.

Module 5: Using the open crosswalk

How the crosswalk was built and checked, what it means for AI on your project, and how to apply it.

Course Content

1 of 2