My 3 years in architecture offices made clear my design skill and my affinity for making maximum use of computer-aided design tools. Also, having had a good look at The Profession, I had little desire to pursue a professional architecture license.
I decided I would specialize in design and CAD, at the time the very leading edge of digital design. The Master of Science program at the M.I.T. School of Architecture and Planning seemed the place to be.
environment design | user tools | simulations | 3D design | architecture | urban studies
Master of Science in Architecture Studies (S.M.Arch.S)
Massachusetts Institute of Technology, Cambridge MA
A two year program, and a class of 21 - all but three of us from other countries. Most of the interest among my class had to do with cultural subjects. I was the only one pursuing digital technology. (This was two years before Bill Mitchell became dean and changed all that.)
The program was very open-ended, with only two required courses. They also encouraged working with other departments at M.I.T.
I wanted to research how CAD could assist architectural designers. In my 5 years of practice to that date, I made great use of AutoCAD and saw a huge potential. What could I come up with?
expert systems, or not
I looked into design theory and at the ideas behind expert systems, where details could be worked out by the machine and freeing the designer to investigate more possible solutions. I that progress so far on ES was pretty disappointing. Given my limited coding experience, I got the feeling making such a tool was going to be a very tall order
At the same time, I fell in love with urban design. I’ve always been drawn to look at the larger context, and living in Boston was a great first-hand experience. Urban studies also were more statistical by nature - a broader scope - than architecture, which dealt more in human and subjective issues
"Explorations of Computer Tools for Urban Design Projects” was undertaken with an open mind. I worked with MiniCAD on a Macintosh SE because it came with Pascal scripting capabilities. I would attempt to design urban places with it, noting where I though automation would help me out
What I built was a design environment. The development site was the workspace. The building program, the square footages of different types of use, were represented as raw volumes in an inventory. The designer could grab chunks of these volumes and arrange and stack them on the site, mass-modeling buildings and the spaces between them.
big idea, little resources
More interesting, but what I wasn’t able to build at the time, was a plan to instance automatically upon the designer’s mass model basic urban spaces of varying types - generating very quickly plausible visualizations of the designer’s work for presentation and review. My thesis paper explored in some depth basic urban space types - plazas, parks, streetscapes, alleys, and parking - and codified variations based on dimensions and relationships.
Decades later, I saw much the same idea had been realized by ESRI with their CityEngine application. So cool