One of the objectives of the Atlas EV Hub is to demonstrate the effectiveness of collaboration. We believe that through crowdsourcing and other data sharing practices, organizations can be more effective by spending their time on consuming and interpreting data rather than collecting it.
Public policy professionals should have access to electric vehicle registration data in a way that protects privacy and enables effective public policies and programs
Data on electric vehicle registrations is the single-most requested piece of data we get for the EV Hub. Good public policy on transportation electrification demands that we know where the vehicles are located and how that’s changing over time. With this in mind, we’re working with local stakeholders nationwide to identify a process to collect this information in a way the protects privacy and share it publicly.
You’ll need access to the EV Hub in order to participate in this effort; if you’re interested in helping out, please send an email to support@atlasevhub.com.
“We participate in this effort because we believe easy access to vehicle registration is essential to our program design and evaluation.”
Datasets
About the Electric Vehicle Registration Data Format
With the sharing of these data, we have to balance the (potentially competing) needs to make the data accessible, consistent across states, and flexible. States are not consistent in how they can share these data, so we’re not able to make the datasets completely apples-to-apples.
Vehicle Identification Number
We can sometimes share part of the vehicle identification number (VIN) for an individual registration. In other cases, the state DMV interprets the VINs before sharing the data.
For the EV Hub, we’ll only post the first eight to eleven digits of the VIN to avoid sharing anything that could be considered personably identifiable. The VIN is the most reliable way to make sure you’re interpreting the vehicle registration correctly: is it a conventional version of a vehicle or the electric version? So, we give you part of the VIN data directly along with our VIN decoder, which is a simple table of plug-in electric vehicles available in the United States.
Vehicle Location
Our target audience for the EV Hub are professionals in public policy so we don’t have personally identifiable information like the address of a vehicle. Instead, we aggregate the total number of vehicles by make and model to the ZIP code level.
Time Series and Vehicle Registration Dates
Vehicle registration datasets are typically snapshots in time of the vehicles “on the road” in a state. Thus, we need multiple snapshots to piece together changes in the market over time. We also need the complete VIN in order to determine when the vehicle first entered the state’s market (new or imported from elsewhere). That is, the first time it appears in the database, we consider it an “original” registration and subsequent occurrences of that VIN are treated as a “renewal” registration. If we don’t have the full VIN, then we can’t reliably track metrics like original or renewal registrations unless the DMV shares that information directly.
Regarding time series analyses, we’ll ideally get snapshots of the vehicle registration database frequently enough to be able to understand how the market has changed over time. With each state vehicle registration data file, we include a table describing the DMV snapshots to make it easier to interpret the data.
Some states aren’t able to share these snapshots frequently or they don’t share the vehicle’s registration date. For states where we don’t know the vehicle registration date, we’ll assign the registration date to the day the snapshot was taken or delivered to us.