Some time ago, I successfully finished my first studio project since enrolling into the Master of Design Engineering(MDE) program at Harvard University. There many learnings while I approached the assignment, and wanted to write down a little as a record.
MDE is a two years masters program. At Harvard, I need to complete four semesters worth of courses to graduate. There is a basic outline of how to complete these semesters. In the case of the MDE program, approximately half of the courses are core classes, and the other half consists of elective courses. The studio class I’m working on is one of the two core classes which every MDE student must take.
Mobility Visualization Project
Before coming to Harvard, the theme of this studio class was announced to the incoming students, and that was mobility. As mobility being the core theme of this studio class, all of the assignments are structured. For this particular assignment, the students were asked to create a visualization that relates to the theme of mobility. As a program aiming to teach students how to come up with integrative problem solving utilizing both design and engineering skills, the goal of this class was to equip students with necessary skills. Particularly for this assignment, the focus was on being able to tackle data to derive with a conclusion through a visualization.
Hell Breaks Loose
When I first came across the studio theme, I imagined vehicle products that major mobility companies would produce, and I wanted to work on this assignment by avoiding that typical association. After some thoughts, I ended up proposing on working on visualizing the route of nuclear wastes in the United States, which is not a visualization you would see on a regular basis. Later on, I came to experience a great deal of misery by making this choice…
To begin with, I had a hard time trying to find the right data. Come to think about it, it’s not the type of information you want it out in public, so this is fairly reasonable and I made a bad decision on picking the topic. After investing some time in intensive googling, it wasn’t exactly what I imagined, but I came across something that is interesting to visualize. One major problem was that this data wasn’t digitized, and I had to transcribe some of the information manually, and that pushed back the schedule. After I finished the project, I was just few hours away from the deadline.
What was the Takeaway?
The assignment was due Monday, and I wasn’t even done with preparing the data by the previous Friday, and it was just hell. Initially, I was intending on presenting the assignment as an interactive project using D3.js or leaflet.js, but given the situation, the TA had advised me to steer away from this ambition. The assignment being interactive wasn’t mandatory, so I decided other ways to complete the assignment.
What was particularly interesting about this assignment for me was that the students were given the opportunity to use a variety of tools to complete the project. We could have worked on as an interactive project using D3.js, but also could have been done by utilizing Tableau, inVision, or Sketch. The emphasis was on how to understand the target data, and what kind of argument we would make through the data and the visualization method. Given that the data is a real thing, the instructor kept on saying that we can “just fake it” (in a good way). So, by taking this advice, I “faked” my presentation by pulling several all-nighters.
At the end of the day, I didn’t end up using D3.js, but I wrote a KML file through Google MyMaps and exported the file to GIS software and then integrated all of the data on Illustrator. I could have used openFrameworks to make the result slightly interactive, but I had to give it up due to the time constraint.
Looking Back
According to Ben Fry, a data visualization guru, he says that data visualization goes through the following seven stages:
Acquire
Parse
Filter
Mine
Represent
Refine
Interact
Note that representation comes towards the end and understanding the data comes in the first part. However I analyze the data, I should have spent more time following what the data is more. It would have been even better if I could iterate the process of acquiring the data several times. People say that designers need to “fail fast” and this was exactly what I was missing in this assignment. I should have iterated and tried to refine my ideas several times. Regardless of the precision of the hypothesis itself, use whatever tool that serves the purpose of testing the hypothesis and iterate that. This is really important. Up until this point, I was thinking that making a mockup automatically means to write a program, but instead, it’s about testing the hypothesis regardless of the method.
