This Is The First Real World 3D Object Ever Scanned: A VW Beetle

I actually wrote about this over a decade ago, but I found myself thinking about this once again, the fact that, in 1972, the first real-world object was “scanned” and translated into three-dimensional data to become the very first actual, physical world object to be turned into a computer 3D model. That real world object was a 1967 Volkswagen Beetle owned by the Marsha Sutherland, the wife of Ivan Sutherland, the computer graphics pioneer.

What makes this all really great is how that Beetle was scanned. Remember, this was in 1972, and literally nothing had ever been 3D scanned before. Professor Sutherland was teaching a graduate-level computer graphics class at the University of Utah, and issued a challenge to his students to find an object to attempt to render in a realistic manner on a computer. It had to be something very recognizable, something iconic, and something large enough that could be worked on as a group.

A Volkswagen Beetle was a perfect choice.

Remember, this was 1972: one year after the very first microprocessor was released, the Intel 4004, and the year that Atari’s Pong hit the arcades. That was the level of mainstream computer graphics at the time: moving squares on a screen.

But Sutherland and his students could see beyond Pong. They were working on real three dimensional objects that could be stored as data in a computer, then moved and rotated just like an actual object, and rendered, with light and shading. But to do this, they needed data. There were no 3D modeling programs, no scanning devices – but there was a class of students with yardsticks and yellow marking crayons.

As you can see in the image above, the Beetle was scanned by measuring points and vectors drawn on its surface, all by hand, and that coordinate data was entered into a text file.

This was a group project, groups of students assigned to different sections of the car, taller students taking the roof, shorter ones getting the hood, fenders, and other lower body parts.

Here’s what one of the students, Robert McDermott, remembers of the process:

We used yardsticks to measure the x, y, and z coordinates of the painted points on the car surface. The Beetle was assumed to have left to right symmetry so we measured only half of the car. The process was slow and tedious, taking many class sessions to complete. Marsha was a wonderful sport as she drove the car around town festooned with our markings.

I love that Mrs.Sutherland drove the Beetle around covered in the lines and numbers, which I bet looked pretty cool.

The wheels weren’t measured, nor the bumpers or many other details, but it’s worth noting the turn indicators atop the fenders were measured, as were the chrome headlight bezels, which were new for ’67.

Once the car was measured into so many polygons, there was still a lot of work to do to make a coherent whole; the various sections didn’t line up perfectly – a yardstick used to measure metal with compound curves isn’t exactly the most precise methodology – and duplicating and mirroring the data to make the other side – longitudinal symmetry was assumed, so they only measured half the car – wasn’t easy either.

By the time all of the human effort going into measuring the car, refining the data, and the computer time was tabulated up, it was estimated that the cost of the project was likely worth more than the Beetle itself.

The final results weren’t displayed on a monitor, like we would today, as monitors capable of displaying shaded images simply didn’t exist. To render flat polygon-shaded or smoother Gouraud-shaded images (Henri Gouraud was a PhD student there as well), hardware was built to render images directly onto Polaroid film and later, 4×5 photo negatives.

You can see rendered examples above there, with a modern flat polygon rendered Beetle, using that original data file, from the Mac laptop I’m typing this on now.

When I wrote that first article about this years ago, I couldn’t find the exact original 3D data file that Sutherland and his students created, but I’m happy to say that’s no longer the case. Here is a link to that original 1972 hand-scanned 3D file. I love that this data still exists and is available; it is literally the direct result of those students hand-measuring lines on an actual Beetle, way back in 1972.

I don’t think it’s too much of an exaggeration to say this crude model is the start of all of the incredible 3D CGI used in so many things today, and it all started with a humble little Beetle.

It’s strange to think of a 3D file as a genuine historical artifact, but I think we can safely say this qualifies.