Amazon’s drone delivery PR stunt brought the idea of commercial and public use drones to the masses. It’s still a very long time though until Amazon quadcopters are delivering groceries.

However, various government agencies and universities already have licenses for drones, and thanks to the EFF we have transparency into some of the use cases. The EFF has filed and won freedem on information act requests with the FAA, the federal agency responsible for granting certificates of aviation (COAs) for domestic drone flights (flights over 400ft). An ongoing map of COAs in the public domain is below.

I ended up looking through some of the COA files and was fascinated by the maps and annotation tools used to document flights. Below I walk through a handful of the maps that have been filed.

Department of the Army, Hawaii

I would love to know what software this is

City of Herington, Kansas

Powerpoint for annotation?

Miami-Dade, Florida Police Department

Satellite imagery is incredibly valuable as a base map for flight planning, and pre Skybox and Planet Labs, Digital Globe was the main provider. Both images use Digital Globe imagery for the base map. The image on the right also references the US Geological Survey and NOAA as additional data layers.

Cornell University

Google dominates, but it really isn’t a great enterprise mapping application

Georgia Tech Police Department

Google and Tele Atlas base map data

Houston, Texas Police Department

More Google and Tele Atlas and Europa Technologies base map data

Mississippi Department of Marine Resources

Utilizing Google for the basemap but annotation done outside of Google

Polk County, Florida Sheriff’s Department

A lonely Microsoft map, and Filestar for Windows…wow

Seattle, Washington Police Department

Do we really have to use Microsoft Office Picture Manager to annotate a map?

A Google Earth Engine shot combining base map data from Europa and the US Geological Survey with Google’s annotation tools

USDA Agricultural Research Service

Google and Tele Atlas, Digital Globe, GeoEye, and INEGI data for the base map

A well annotated but entirely static map

Air Force

What version of windows is this?

All shapes drawn over an image in powerpoint

Texas A&M University

Google with Digital Globe and Tele Atles FTW again

Pennsylvania National Guard

Anyone able to make sense of this?

Louisiana National Guard

California Fire Services

Hays County, Kansas Emergency Services

Ogden, Utah Police Department

University of Wisconsin

End Note

A lot of the COAs are from 2008-10, so it is not definitive that these are still the tools being used.

Back in December a friend asked me what I would suggest to start learning R. I put together the list below, and am finally getting round to sharing it here. If you have other suggestions of great resources for beginners do let me know so we can add to this.

  1. TryR by CodeSchool

    This is all in the browser but will offer a frictionless basic introduction to R syntax, data frames etc.

  2. Download R Studio

    This is the simplest IDE for R, and what I use for all my projects.

  3. Take an online course

    Roger Peng’s Coursera course Computing for Data Analysis is a solid introduction. As is Jeff Leek’s Coursera course Data Analysis. Both also put their videos on Youtube: Roger’s Youtube channel and Jeff’s. Google also has a set of courseware to learn R, and you could read R-Project’s Introduction to R notes.

  4. Start a basic project

    The first project I did using R was a basic exploration of Yale’s internal Facebook that looked at a distribution of majors etc. It wasn’t complicated, and didn’t involve any analysis beyond descriptive statistics, but it forced me to start mastering the R syntax etc. Having an achievable project in mind (a basic data set you want to explore) is the best motivation for really learning the language.

  5. Blogs

    R Bloggers has a ton of tutorials and resources. The Simply Statistics blog, written by the afore mentioned Roger and Jeff, is a great resource for both stats, R, and inspiration. StatsChat is a useful blog. I also find a bunch of interesting stuff on the Revolutionary Analytics Blog. Hadley Wicham is a prolific contributor to R and the author of ggplot which is a beloved graphics library, and he is worth following on Twitter.

Kickstarter just had the billionth dollar pledged on the platform. The company published a fun infographic of some of their data to celebrate. They included a map of how much each country had pledged total, but they didn’t show where all the money had gone to. So I visualized that as 1. By days over time 2. Cumulative over time and 3. By category over time.

Details

  • The data set I scraped shows $957,512,698 total pledged (I am missing a few campaigns that threw errors when scraping) across 131,348 total campaigns
  • The data set has $840,599,488 successfully pledged across 56,270 successful campaigns

Methodology

  1. Ruby script to collect all the data from Kickstarter campaign pages
  2. Google Apps Script that geocoded location addresses uploaded into a Google spearsheet into lat/lon for mapping. Thanks to @baygross for helping me with this script.
  3. R script to pull out the addresses for geocoding, and merge the result