The BIM Prophet — Charles M. Eastman
Umberto RinaldiHow Prof. Eastman imagined the future of design fifty years before it existed
The room in 1975
Imagine a desk. Paper, pencils, maybe a cup of coffee gone cold. Outside the window, the campus of Carnegie Mellon University. It's 1975 and Charles Eastman, professor and researcher, is writing a paper that will be published in the AIA Journal under a deceptively sober title: "The Use of Computers Instead of Drawings in Building Design".
He is not describing software. He is not proposing an update to something that already exists. He is building an entire world — with its own rules, its own language, its own internal logic — without knowing whether anyone would ever have the tools to inhabit it.
Thinking about it today, one is reminded of Tolkien elaborating Middle-earth: the languages of the Elves, the genealogies of Kings, the history of ages no reader would ever directly experience. A complete, coherent system, inhabited by its own logic. Eastman was doing the same thing — but instead of a fantasy world, he was laying the foundations of what we now call BIM, in 1975 Eastman himself called It BDS (Building Description System).
The problem no one had yet named
To understand the scope of what Eastman had intuited, one must understand what design looked like before him — and in part, what it still looks like today for those working only with 2D CAD.
Every building lived on paper in three separate versions: the floor plan, the elevation, the section. Three distinct drawings, produced by hand or with CAD, that had to tell the same story from different angles. The problem was that they were not connected. If you moved a window in the floor plan, you had to remember to update it in the elevation. And in the section. And in the window schedule. And in the bill of quantities.
The error was not in the designer. It was in the system. Eastman grasped this with an almost unsettling clarity: the problem was not the quality of the drawing, but the redundancy of information. The same window existed in four different places, and every time something changed, the coherence of the project depended entirely on the memory and discipline of whoever was working on it.
His solution was elegant in its radicalism: eliminate redundancy at the root. Not four representations of the same window — a single object, with all its properties, from which representations are generated automatically. The floor plan is not a drawing. It is a projection of the model. The elevation is a projection of the model. The section is a projection of the model.
The drawing ceases to be the final product and becomes a by-product of the database.
The language before the world
But the most extraordinary part of the 1975 paper is not the solution to the redundancy problem. It is the last page.
Here Eastman no longer speaks of architects and drawings. He broadens his gaze and describes something that closely resembles what we now call the BIM ecosystem — and does so with a precision that, read today, is almost chilling.
Manufacturers as data providers. Eastman imagines that companies producing building components — windows, doors, structural systems — would no longer deliver paper catalogues, but would supply their elements directly as digital objects, with all their technical properties already embedded. In 2026 we call them BIM objects or Revit families. In 1975, Eastman was already describing them.
Automated building code compliance checking. Imagine a digital model being submitted to a public authority, which does not read it manually but analyses it automatically to verify regulatory compliance. This is exactly what is now called BIM Validation — a frontier the industry is still working on, fifty years later.
The construction site as the recipient of the model. The model does not only serve design. It serves construction, logistics management, material calculation, and execution phase planning. What we now call 4D and 5D BIM — time and cost integrated into the model — Eastman had already laid out as a natural consequence of his system.
It was a manifesto. Not a technical proposal, not an operational manual. A manifesto of the dematerialisation of drawing — the idea that value lies not in the stroke on paper, but in the quality and accessibility of spatial data.
Thirty years of silence (and submerged work)
So why did it take thirty years to become standard?
The honest answer is threefold: the hardware was missing, the software was immature, and — when both finally arrived — there was still cultural and commercial resistance to overcome.
In 1975, computers were mainframes that filled entire rooms. Eastman himself, in the paper, spoke cautiously about managing "hundreds of thousands of elements" and the need for tape or disk memory. He had the theory. The machine to run it was missing.
The first to take up the baton was Gábor Bojár, a Hungarian physicist who graduated from Eötvös Loránd University, who in 1982 left his job at the state geophysics institute and founded Graphisoft together with István Gábor Tari, an assistant professor at BME Budapest. To launch Graphisoft in 1982, the partners needed a professional computer, then extremely expensive in communist Hungary. Bojár recounted in several interviews and in his book The Graphisoft Story that he and his partner sold their wives' jewellery to raise the necessary funds (around $30,000, an enormous sum at the time). To work on the software that would become ArchiCAD, Gábor Bojár had to resort to creative and risky methods to circumvent the technological restrictions of the era, going as far as personally smuggling several computers across the Iron Curtain. In particular, he brought four Apple Macintosh computers into the country hidden in the boot of his car — it was the only way to test his software on graphically capable machines. In 1987 ArchiCAD was launched commercially: the first BIM software on a personal computer, capable of generating 2D and 3D geometry. It was not even the only attempt of those years — Sonata, released in 1986, was by Bojár's own admission technically more advanced than ArchiCAD at that moment, already anticipating what would become the mature definition of BIM. But licence costs were still prohibitive, and the workstations required were beyond the reach of most practices.
The decisive leap came in the late 1990s. Leonid Raiz and Irwin Jungreis, former developers of Pro/Engineer — the parametric modelling software used in the mechanical industry — left everything to found a new company with a precise goal: to bring parametric logic to the building sector. In 2000 their software was called Revit. In 2002 Autodesk acquired it.
But the acquisition was not the end of the story — it was the beginning of a new, complicated chapter. Autodesk already had its flagship product: AutoCAD, with a consolidated global client base and a business model that worked. Revit was technically superior for BIM design, but pushing its adoption meant, to some extent, cannibalising the product it already sold. The result was a slow transition, not always driven with conviction, and a far from uniform diffusion on a global scale. Different markets, different professional cultures, different regulations — each country found its own path, with its own timing and methods. In many practices, AutoCAD and Revit coexisted for years on the same machines, used in parallel on different projects.
Eastman's language had found the machine to run on. But not everyone was ready to switch it on at the same moment.
The prophecy has come true. What now?
Today we work every day with the tools Eastman had imagined sitting at his desk in 1975. Parametric families, BIM catalogues from manufacturers, automatic coordination between views, the model as a single source of truth — all of it is there, described with an almost prophetic clarity nearly half a century ago.
But there is an interesting detail. Eastman had not simply predicted BIM. He had predicted a process — a different way of working, of collaborating, of thinking about a building as a system of information before it is a form in space. And that part of the vision is still being realised.
The integration of artificial intelligence into BIM tools is the next chapter of this story. Tools like Revit are already incorporating AI functions to automate repetitive operations, suggest design solutions, and analyse models in real time. We are probably facing a new "moment zero" — not unlike the one Eastman experienced in 1975, when the vision was clear but the tools were still struggling to keep pace.
We will talk about this in the next article. Because the story has not stopped — it has only accelerated.
Have you read Eastman's original paper? Or did you discover BIM from a different direction? Tell me in the comments — every starting point reveals something different about how this world spread.
I'll leave you the link to the official journal where Prof. Eastman's article was published (PDF, pp. 46–50) 👇