Assginment 1: Documentation

Shark Attack and the Economy {Australia): Can you see the relation?
Chloe Yeung
307140857

Data Characteristics
The two datasets represented in my data sculpture is Shark Attacks and Unemployment, inspired by the article ‘Shark attacks drop as recession cuts beach time’ [1] where George Burgess proposed that shark attacks are declining due to the current worldwide economic recession. I was interested to see whether he’s statement holds true for past years so I collected datasets for the U.S. and Australia, choosing the attributes of year (1990 to 2007); unemployment rate [2] (averaged per year) as the means to represent recession and total shark attacks [3] (fatal + non fatal yearly).
[My excel spread sheet can be seen here:Combined_datasets.xls

Looking broadly, both the U.S. and Australia have experienced increased shark attacks despite declines in the unemployment rate, suggesting that Burgess’s hypothesis may not be accurate. I can only theorize why this trend is apparent: in times of low unemployment, meaning people are employed and earning a steady income, are more they are inclined to spend money for luxury activities – such as beach holidays, exposing them to shark habitats increasing the chances of a shark attack occurring.

Design Rationale
‘”Where is the correlation between your datasets?” ‘The trends between your datasets are not clear.” These questions became the driving force behind my data sculpture. To begin my data mapping I first decided upon a form to work off. A spherical form was chosen because of its circular, flowing structure, a simple form that seems to entice audiences to touch, but more so to illustrate the miss-match of my two datasets.

A spherical form meant that I was able to segregate its form into 18 sections to represent the attribute: year. I had different ideas on how I could differentiate between the final two attributes: unemployment and shark attacks; my initial idea was to have ‘blocked areas’ and ‘see-through’ areas that would represent each dataset.

But when user interaction came into the equation I had to modify this idea, so I morphed my spherical form into a ‘step sphere’, allowing me to indent one centimetre into each circular section to represent unemployment whilst the unindented areas represented shark attacks. The reason behind indenting unemployment was more for structural reasons, because of the spherical form the top and bottom sections are obviously smaller, so I had to ensure that, particularly the base, had maximum surface area in order to hold the structure. This was also the reason behind why I decided to arrange the years in chronological order: top – 1990 and the bottom – 2007.

It is to be noted that the size of the indented sections are not random. I had to calculate and recalculate to ensure that I was displaying the data as truthful as possible. However because of scaling there is still some obscurity of my datasets, but it seems to enhance the ‘unclear’ nature of my datasets. I further pushed the notion of ‘unclear trends’ through the form of user interaction. Each individual section can be moved freely 360 degrees supported by a notched axis, so the audience can turn and arrange each section as a means of exploration and data interpretation, as well as altering the sculptures visual aesthetics.

Labeling the sculpture was one of the design requirements to be considered. I am against physically writing on my sculpture ‘Unemployment’ and ‘Shark attacks’ because to me that felt too obvious, so to ‘label’ my sculpture I decided to sand down the indented edges of unemployment. This not only visually changes its appearance, as these areas are now smooth and has a ‘frosted’ look; it also provides a tactile label, where a person will be able to ‘feel’ the differences between the datasets, made clear with a key provided. However, I did discreetly label the date on each section.

Technical Challenges:
Technical challenges for my data sculpture came at me from left, right and centre throughout the design process:
Maya Modeling: NURBS modeling-
This is where I hit my first wall of frustration, where I spent an excessive amount of time creating inconsistent sections. To make up for time I modeled in polygons to see how my sculpture will fit together then recreated the templates through illustrator and Rhino.

Fabrication mistakes-
I am no engineer so trying to work how my sculpture would fit together was a challenge and it was here in which my design almost failed. My proposed idea for user interaction was moveable sections, so my idea was to have each section stacked on top of one another through an axis to create the form. What I didn’t take into consideration was the surface friction that would occur between the sections when stacked, which in effect would prevent the sections from moving individually. Luckily Andrew made this problem aware and was solved by separating each section with circular ‘rings’ that minimized the friction.
Just when I thought that my fabrication process would be finished when my design was laser cut onto 3mm Perspex the nightmares continued. I was unaware at the time that one of my pieces had its inner circle off scale, which meant that when I attempted to put them together it didn’t fit. A second laser cutting session was booked to fix this problem.

Glue exploration: What can glue Perspex without leaving a ‘frosted’ trail?-
Working with 3mm Perspex meant that I had to combine two layers to create a sculpture of a reasonable size. My problem was how am I going to combine two layers without seeing the glue? I decided to do a little online research and found that Selley’s Plastics Glue [4] was a good option. So, going to the local hardware store I decided to further buy three more other adhesive alternatives: Loctite Super Glue, Selley’s Araldite, and Selley’s Araldite Ultra Clear.

In the end Selley’s Araldite Ultra Clear was the strongest and the clearest glue to use, with minimal residue. However, air bubbles and slight inconsistencies within the glue can still be seen.

Data labeling: Sanding, sanding and sanding-
To label my sculpture I decided to create ‘tactile’ labels, where the audience will not only be able to see the different datasets, but also feel it. Little did I know that sanding Perspex was such a big task. I first started sanding by hand, but two hours later and I had only sanded two section pieces. Not Good. My only solution was to sand with a hand held drill fitted with a stone grinder. In all it took seven hours of mechanical and hand sanding to sand all 18 pieces.