BIM is old. Considerably older than most of the people selling it. The idea that a building should be described to a computer as objects with properties rather than as lines on a sheet was fully articulated in the 1970s — before the personal computer, before the mouse was commonplace, and roughly four decades before most governments got around to requiring it.
What follows is the sequence, with sources at the bottom. We have flagged the places where the popular version of this story is simply wrong, because those errors get repeated in procurement documents and tender responses where they do real damage.
1963 — Sketchpad, and the idea that drawings could be data
Ivan Sutherland's Sketchpad, developed at MIT, is where interactive computer graphics starts. It let a user draw on a screen with a light pen and — critically — apply constraints: make this line parallel to that one, keep this joint attached. Change one element and the related geometry updated.
That is the ancestor of every parametric feature in every BIM tool you use. Sketchpad was not about buildings, but it established the principle that a drawing could be a structured model with rules, rather than a picture.
1975 — Charles Eastman describes BIM before the word exists
Charles Eastman, then at Carnegie Mellon, published "The Use of Computers Instead of Drawings in Building Design" and developed a prototype called the Building Description System. Read the description today and it is uncomfortably familiar: a single integrated database serving both visual and quantitative analysis, parametric design, and high-quality computable 3D representations.[1][2]
Eastman's argument was that drawings are a lossy, redundant way to describe a building — the same information restated across dozens of sheets, each of which has to be updated by hand and each of which can disagree with the others. His answer was to describe the building once, as elements, and derive the drawings from it.
Fifty years on, the core BIM sales pitch is still the argument Eastman made in 1975. The industry did not lack the idea. It lacked the hardware, and then it lacked the incentive.
He is widely called the father of BIM, and on the evidence of that paper the title is deserved. What he did not do was name it.
1986 — "Building model" enters the literature
The vocabulary arrived in stages. The term building model, in something close to today's sense, appears in a 1985 paper by Simon Ruffle published in 1986, and in a 1986 paper by Robert Aish.[3] Aish's work at the time is also where much of the thinking about building modelling for construction sequencing originates.
1987 — ArchiCAD puts 2D and 3D on a desk
Graphisoft's ArchiCAD launched in 1987, described as the first CAD product on a personal computer able to create both 2D drawings and 3D geometry from the same model.[3] This matters more than it sounds: until a modelling tool ran on hardware an ordinary practice could afford, BIM remained a research topic. Graphisoft moved it onto the desk.
1992 — the term is finally written down (and not by whom you think)
Here is the correction worth carrying with you. The phrase "Building Information Model" first appears in a 1992 paper by G.A. van Nederveen and F.P. Tolman of TU Delft.[3] Not Autodesk. Not a vendor. Two Dutch researchers, in an academic paper, a decade before the industry noticed.
A myth worth killing
You will frequently read that the term was coined by Jerry Laiserin, sometimes described as "a British architect", in 1992. This is wrong twice over. Laiserin is an American industry analyst and writer, not a British architect — and his contribution came a decade later.
What Laiserin actually did was more useful than coining a word. In 2002–2003 he hosted a public exchange between Autodesk, Bentley Systems and Graphisoft, and in doing so helped standardise BIM as the common industry name for what each vendor had been calling something different.[3] Naming a thing is easy. Getting three competitors to use the same name is the hard part.
2002 — the vendors align and BIM becomes a market
2002 is the hinge year. Autodesk published a white paper titled Building Information Modeling[3], having acquired the maker of Revit — a young parametric building modeller — earlier that year. Bentley and Graphisoft were pushing comparable ideas under their own banners. Laiserin's exchange gave the category a single label.
From this point BIM stops being a research programme and becomes a product category, with everything that implies: real investment, real capability, and a great deal of marketing that outran what the software could actually do.
1995–2013 — IFC, or how models learned to talk to each other
A model locked inside one vendor's file format solves one office's problem and creates the industry's. The Industry Foundation Classes (IFC) were developed to let building data move between tools, under the organisation that began as the International Alliance for Interoperability and is now buildingSMART.
IFC's significance is institutional as much as technical: it has been an international standard, ISO 16739, since 2013.[3] That is the point at which open exchange of building data stopped being an aspiration and became something a client could specify in a contract.
2016 — the UK stops asking and starts requiring
Adoption curves in construction bend when someone with money insists. The UK government announced its intention to require collaborative 3D BIM on its projects by 2016[3] — the "Level 2" mandate. Whatever one thinks of how it was implemented, it did what a decade of vendor marketing had not: it made BIM capability a condition of winning public work, and the supply chain responded accordingly.
2018 — ISO 19650, and BIM grows up
The UK's PAS 1192 series was internationalised as ISO 19650, whose first two parts carry the 2018 designation — BS EN ISO 19650-1:2018 and -2:2018 — with publication and national adoption running into January 2019.[3]
ISO 19650 is where BIM stops being about software at all. It is an information management standard: who requires what information, in what structure, at which point, and who checks it. OIR, AIR, PIR and EIR; the common data environment and its approval states; naming and metadata conventions. Not a single mention of which modeller you use.
The through-line of sixty years: BIM began as a way to describe a building to a computer, and ended up as a way to govern who is accountable for information. The geometry was never the hard part.
Where this leaves you
Two practical conclusions fall out of the history.
First, the modelling problem was solved a long time ago; the information problem was not. Every generation of tooling has been better at geometry than the last, and the recurring failure has stayed constant — nobody defined what the model was for, so it got built to no particular standard and served no particular decision. That is precisely the gap ISO 19650 addresses, and it is why governance work usually pays back faster than more modelling capacity.
Second, the industry's own vocabulary is unreliable. If a widely repeated claim about who invented BIM can be wrong in two respects at once and still circulate freely, treat confident assertions about BIM maturity, ROI and "levels" with the same suspicion. Ask for the source.
Next in this series: the ten buildings that proved BIM at scale, and what each one actually demonstrated.
Sources
- Eastman, C. The Use of Computers Instead of Drawings in Building Design (1975), as summarised in Charles M. Eastman — Wikipedia.
- Rebels of Construction — Charles M. Eastman, Beck Technology.
- Building information modeling — Wikipedia, for the van Nederveen & Tolman (1992) first use, the Ruffle (1985/86) and Aish (1986) papers, ArchiCAD (1987), Autodesk's 2002 white paper, Laiserin's 2003 role, IFC as ISO 16739 (2013), the UK 2016 mandate, and ISO 19650.
- Laiserin, J. Building Information Modeling Two Years Later (PDF).