On Friday, November 13th I joined several of my fellow members of Art and Engineering Design on a trip up to Chicago to visit the downtown offices of
Skidmore, Owings & Merrill, an architecture firm involved in buildings such as the
Sears (not Willis) Tower,
Trump Tower, and the ridiculous
Burj Dubai. They had invited our group up for a day of talks about various aspects of the company, and I, with an amateur interest in huge skyscrapers, the sense that I should be more involved in AED events, and an interest in going up to Chicago, agreed to take part in the trip.
I got to the train station in Champaign at 6:45 to get my ticket and meet up with the rest of the group and boarded the Greyhound at 7:10 for an uneventful ride up.
We pulled into the Greyhound station in Chicago and bought a pass for the subway, which we took most of the way to the SOM offices, which are located on Michigan Avenue. We walked the remaining few blocks and arrived at the super classy offices in time for our scheduled arrival at 10:00.
Once we arrived at the offices were were given a quick tour, an explanation of the different departments in the company, and shown examples of different projects that SOM had been involved in.
Once this was over we were taken to a conference room where we listened to three in depth talks about various aspects of the company and designing buildings, all of which proved to be very interesting.
The first talk was about optimizing the design of buildings from a structural standpoint. Math and engineering were at the forefront, and the implication was that building design is first and foremost based on these mathematical designs, and the architects only tinker a little bit to create the desired aesthetic. I'm sure that not quite enough credit was given to the architects here, but given the natural beauty of the mathematical based models, it wasn't too hard to believe that they took the forefront in design. This talk ended up giving me the impression that most professions that at all involve math are filled with people who think that the golden ratio is pretty much the coolest thing ever. In this case, optimized building designs were based on the logarithmic spiral. This means that designs for new buildings are easily distinguishable from older buildings. Old buildings tend to be boring, unnatural and ugly boxes, while new buildings have an organic quality to them, being much more filled with curves, which also makes them much more appealing to the eye.
The second talk was about designing sustainable buildings and what that means. SOM has an internal standard that all of the new buildings it designs will be at least "lightly green." Different examples of sustainable buildings were given, from allowing comfortable work in a desert environment to designing a building to have zero emissions. It was discussed how there are two popular trends in buildings these days that are at odds with each other, and which most clients want. First, they want their buildings to be energy efficient, to be green, to even be certified as such. They also want their buildings to have a sexy all-glass exterior. This however means that a lot of energy will lost. In more temperate climates it is possible to have an energy efficient building that is all glass, though more difficult than one that isn't all glass. However, in hotter regions, such as southeast asia and the middle east it is simply impossible to make a green building in the same style.
Finally, we were given a talk about what SOM's Black Box group does. As a Computer Science major I wasn't expecting anything in these presentations to be relevant to my major, but to my surprise what the Black Box group is doing is completely relevant, and pretty awesome to boot. What the group does is solve various building, architecture and city planning problems with various computer models. One of the coolest methods they utilize is called genetic algorithms, in which they describe an object with a sequence of numbers. Each number represents an aspect of the object, and together, the sequence makes a sort of DNA for the specific instance of the object. Many random objects of the same type, but with different numerical values are generated and grouped together to form a "generation" for the object. Then the entire generation is put through a simulation to see how well each object performs. There will be a specific goal or two in mind: light let in a window, material used, surface area the sun shines on, etc. Some of the objects will perform better than others. The top 10% or so of a generation will be allowed to move on to and help create the next generation. Some of the next generation will come about through random mutations of the survivors' genetic code, and others will come about by combining the code of two survivors, simulating mating. This process will continue for many generations until the objects seem to converge on an optimal design.
I find this all extremely cool, and the fact that there is a strong computer science element even here makes me feel as if I could go into almost any field with a Computer Science degree, as almost everything uses computers these days.