This is the web page for Michael DeLaurier's 2021 Indepdent Study with Sam Cortez. The purpose of the study is to use Rhinocerous, Grasshopper, and Elk to transform numerical data sets into thought-provoking physical sculptures. Throughout the study, I'll explore the facinating connections between data and scuplture as I render these data models and 3D print them. Below, updates on my progress and final models will be posted every Monday.
I'm super pleased with the result of my project. I printed the nine counties with white PLA, but then used LEDs to give them a colorful look. I also photographed them on glass so I could also capture their reflection.
I ran into an issue with my original final project idea: it's difficult to find COVID data for the specific longitude latitude squares I found from this website. Instead, I'm going to use boolean difference to carve the squares into the Bay Area's counties. This will make data aquisition and printing much faster. Below is the final rendering and Grasshopper code for the project.
Note: It's absolutely curcial to do all moving and extruding within Grasshopper, not on Rhino. If you try to extrude the topographical surfaces in Rhino, everything will take much longer.
All the data I gathered was from the New York Times. They have accessible, up-to-date data that lets you select between total cases and per capita cases.
This is the rendering of my final project. The height of the pillars will be based on regional data (mostly likely COVID cases) and the tops of the pillars will be the topography of the area. No actual data was used to create this model; this is just a placeholder.
For my first test print, I extruded a single topography square and then printed it with minimal infill:
I liked the idea, but the topographic elements weren’t exaggerated enough. For my next test, I extruded the surface in the z axis:
I liked this test print a lot. For my next one I’m going to exaggerate the z stretch even more. These are three size tests I did to determine how my 3D printer handles changes in size detail. The medium size will be my best bet since I’ll need to print quite a few squares.
Sam found a way to make the topography model work! The solution was to download .hgt (height) files from this website. The website contains several files that span the entire state of California. Routing that file into the Grasshopper code below successfully previewed the topography in Rhino. Grasshopper code below:
I’m definitely going to use this instead of the OSM data. It’s much more visually compelling and easier to print.
The path I'm currently persuing is Elk's OSM Module. While the OSM Module only interprets roads and highways, I'm hoping to manipulate their height to create a 3D sculpture. Right now, my current obstacle is Open Street Map's export capability. It's not powerful enough to export the entire city at once, so I need to figure out how to access the larger file.
Below is a working script that uses the OSM Module.
Update: I got the topography module to work! The info below is helpful, but the updated information is above.
For multiple weeks, I've been trying to work out Elk's topography module. I followed countless tutorials, searched through tons of online forums, and tried several versions of Elk, but nothing's worked. I think the root of the cause might be that I'm using a Mac instead of a PC. Below is a detailed description of most all the steps I took. Hopefully someone else can figure this out!
I started with this tutorial, it provides detailed information on how to use websites that let you extract topographical data. I followed it as closely as possible, but no luck.
Here's the link.
Below are the two sites I used for topographical data: USGS Earth Explorer and the National Map Viewer. The video above details how to use each, but the files I grabbed from each website didn't produce any results in Rhino.
Here's the Elk script which never worked; it's copied exactly from the video but yielded no results. For reference, the coordinates needed for the domains are, top to bottom, South, North, West, then East (this wasn't that clear in the video). Another helpful tip, the coordinate bounding box you specify in Elk must be smaller than the region you selected in Earth Explorer/National Map Viewer.
That's as far as I got with this topogarphy model. My only path forward from this point would be to try the exact same script on a PC.
COVID Map
Voting data
Consumption of coca cola
Global Warming Trends
Grocery Shopping: 2020 Edition
Most pollutant contries
TikTok data set (comments, ages, consumption behavior, trending videos)
Social media platform frequency. Light features
Music popularity across the world or us
Trump’s twitter insults
Political Social Media Posts (Classify partisan bias, audience, and goal based on politicians' social media)
To start, I'm going to use a Bay Area COVID data set as a proof of concept. Before diving into Elk and Grasshopper, I'll manually manipulate topographical maps in Rhino to demonstrate data visualization.
Initial tests using "Extrude Surface to Point"
Material: Glass
Height: Random
I'm not super pleased with this result, I need to do more research to see how I can smooth out the county lines without sacrificing map clarity.