My goal for this course is to acquire the tools I need to analyze and visualize the quantitative aspects of Architecture, and through such quantification develop hypotheses about the qualitative aspects of Architecture. I am inspired largely by Systematic Paleontology, the formal methodology for describing a specimen of the natural world and making a case for its likely relationships.
In the Fall of 2014 I engaged with my first attempt at such an analysis in Francesca Torello’s course History of Architecture in the Islamic World. Using preexisting phylogenetic analysis software I scored 20 specimens (buildings) according to 20 characteristics. The resultant phylogenetic tree was shockingly coherent, with both geographic, temporal, and programatic trends made visible despite having entered only superficial information.
The procedure is straight forward: gather the information, quantify and organize the information, compare selected parts of the information to other selected parts of the information, and highlight key data points that seem to have a relationship.
What I would like to build is a simple comma separated value database that can be infinite extensible in both number of characters and number of specimens. A simple interface allows for a visitor to the database’s website to view the data in its raw format but also to select different viewing options that translate the data into compelling and interactive visualization.
There are three components that need to exist. The first component is the database itself. By using Google Sheets, the comma separated values can be updated from and made accessible to any web browser. The second component is the read/write access to the spreadsheet. This is easily accomplished through the Temboo Google Sheets API. Commands and examples are made available by Temboo in a variety of languages. The third component is the visualization. From the workshops that Eddy has made available to us, I can see a great many options for visualization. As I am proposing an infinitely extensible database of characters, I would like to be able to visualize and contrast in novel and interactive ways. Using 3JS I could create a simulated environment in which every data point is represented. By changing the first person perspective of the view, the data is dynamically rearranged according to different analysis types.
Data acquisition / quantification is to be performed manually at this time. I do however envision a future in which a drone might explore a building, dig into its depths, probe its walls, and use every manner of imaging to document every inch. We are already using 3d laser scanning to construct BIM models. In the future, these data points could be compared to the data points from other buildings. Recently a paper in the Anatomical Record (Boyer et al “A New Fully Automated Approach for Aligning and Comparing Shapes”) used just such a procedure to automate the analysis and comparison of a matrix of bones. As processing power becomes cheaper and cheaper perhaps objects as large as buildings and even cities, regions, and countries could be subjected to such analysis.