I recently found out that information about the number of drones that are registered with the FAA is freely available at this website. The FAA calls them Small Unmanned Aircraft Systems (sUAS) but most people just say “drones.”
I spent some time going through the data, looking at the town I live in and places near me.
The data is pretty interesting. It is divided by hobbyist/non-hobbyist and then lists the number of drones that are registered in every town in the United States (not just states, but also places like Puerto Rico, American Samoa and military bases).
After looking at this massive spreadsheet for a while, I began to wonder if somebody had put this information into a map to make it easier to digest. Searching around the internet I found a few maps but nothing had exactly what I was looking for. Time Magazine’s map was close … but it was using old data.
So I set out to make my own. More on that later … Let me get to the maps and then anybody who is interested in how I made them and some of my assumptions can keep reading below.
Choropleth Maps of Hobby Drones by County
One thing I didn’t know before making these maps is that they are called “choropleth” maps.
All the maps and stats below are dated February 14, 2017.
This map above shows the total number of registered hobby drones in each county in the U.S. Please note that the range of values that each color represents changes as the colors get darker. The only thing interesting about this map is how uninteresting it is. Counties with higher populations appear to have a higher number of drones registered.
Because of this I wanted to make a per capita map, as well.
As you can see the per capita map is much more homogeneous. This is to be expected based on how the map of the total number of drones seemed to correlate with population. The Midwest counties seem to have higher per capita counties but that could be a result of some of them having very low populations.
Also interesting is that the counties that border oceans tend to have higher per capita drone registration, even within the same state. For example, Washington and Oregon tend to have higher per capita registration in the counties that are next to the ocean than the counties that are inland.
Interesting Facts about Hobby Drone Registrants
Total # of Drone Registrations: 667,637
State with the Highest # of Drone Registrants: California, with 78,890 registrants
State with the Highest per capita: Hawaii, with 16.92 registrants per 10,000 residents
County with the Highest # of Drone Registrants: Los Angeles County, California with 17,560 registrants
County with the Highest per capita: Prairie County, Montana with 82.95 registrants per 10,000 residents
# of counties with 0 Drone Registrants: 7
How I made the Maps and Processed the Data
As I said above, the data came from here (Excel file). It has columns for country, state, city, zip code and number of drone registrants from each zip code. (One note about this … the FAA requires people to be registered, not drones. So if one person has 5 drones, they only have to register once. )
I used Python to process the data and make the maps. I’ve fully automated it so I should be able to update these maps and statistics as soon as new data is available.
I used this map as a template. It is an SVG file, which means its an XML based image file. XML can be edited with any text editor or any number of programming language libraries. In my case, I used the Beautiful Soup Python library to read and edit the SVG file.
The map template has an ID tag for each county. The ID uses the FIPS county code. The first two numbers in the FIPS code identifies the state while the last three numbers identify the county.
This posed a problem for me because the drone registration data was given based on zip codes, not FIPS county codes. Fortunately, I was able to find data that relates zip codes to FIPS codes.
This led to 2 further issues I had to resolve. First, some zip codes are located in multiple counties. I solved this by assuming that the drones that were registered to any particular zip code were distributed based on the population. For example, let’s say 100 drones are registered in a zip code 12345. And zip code 12345 is split across 2 counties, with 60% of the population in county A and 40% of the population in county B. In this case, I would assign 60 registered drones to county A and 40 registered drones to county B. The zip code-to-FIPS file that I linked to above actually has a column in it that tells the percentage of the population of a zip code that is found each county that it is located in. This is one of the things I did differently than what Time Magazine did.
The second issue I came across was that because the drone registration data is from 2017 and the zip code-to-FIPS data I had was from 2010, I had a number of new zip codes that had been created since 2010. I had no easy way of converting these to FIPS so – for now – I am leaving these zip codes out of the calculations for the map.
In order to calculate the per capita information I needed the population of data of each county. I took that from this file (.csv file). This is estimated population of each county in 2015, based off of the 2010 census.
Another issue I found with the data was that the zip code didn’t always match the city and/or state that was listed. For example, there was one entry where the City and State columns of the data listed Ames, Iowa but the zip code was 43205, which is a zip code for Indianapolis, Indiana. For the maps and stats above, I assumed the zip code was correct in those cases, not the city/state columns.
Let me know if there is any other data you’d like to see. If it’s fairly easy for me to process I will update the post to include it. For example, I only included people who registered as “hobbyists.” I could make maps for the “non-hobbyist” registrants, as well, if there is any desire.