1. Building the National Address Database AGIC – September 22, 2016

2. BREAKING NEWS!! FGDC ANNOUNCES ESTABLISHMENT OF NEW ADDRESS THEME After 26 years, addresses were finally added as a layer of the National Spatial Data Infrastructure on 9/13/2016 DOT and Census will serve as co-theme leads How did we get here? 2

3. Vision 3 “The National Address Database is an authoritative and publicly available resource that provides accurate address location information to save lives, reduce costs, and improve service provision for public and private interests” Source: The Need for a National Address Database, a report submitted to the Federal Geographic Data Committee by the National Geospatial Advisory Committee, December 2012.

4. The Mandate In February 2015, GAO published the report Geospatial Data: Progress Needed on Identifying Expenditures, Building and Utilizing a Data Infrastructure, and Reducing Duplicative Efforts – One of the recommendations from the report was to Create an address data theme with associated subcommittees and working groups to assist in furthering a national address database. – The FGDC Steering Committee is the responsible party, but DOT and the Census Bureau have taken the lead. 4

5. First Step: The National Address Database Summit Held April 8-9, 2015 at the Maritime Institute in Linthicum, MD Funded by DOT Objective: To identify and discuss possible options for developing a National Address Database (NAD) Broad stakeholder representation – Government (Federal, State, Local, and Tribal) – Private Sector – Non-Profits and Trade Organizations 5

6. Summit Attendees 58 Participants 10 Federal agencies 16 State governments 17 Local governments 2 Tribal representatives 8 Private companies 5 Non-profit / trade organizations Plus 25 Observers 6

7. Summit Outcomes: A Shared Vision Local authorities are the authoritative source for address assignment and are data set originators State authorities should be statewide aggregators – Many are already in this role, primarily to support public safety and emergency response (e.g. NG-9-1-1) Tribal authorities must be included in data flow – Authoritative source and aggregator of addresses on tribal lands Federal leadership and support is needed for there to be a sustainable national approach – Address the needs of the “have nots” 7

8. How it would work 8

9. High Level Recommendations On Leadership – Identify a high-level Federal Champion – State, Local, and Tribal Champions also needed On Approach – Reach out to “have” States and build on momentum – Recognize that public safety and NG911 are key factors – Identify standards and minimum content – Conduct a pilot – Just do it! 9

10. Pilot Launch October 2015 Advisory Group Established – Key Stakeholders DOT Funded Goals Determine minimum content guideline Explore workflows Understand best practices for address roll-up Assess technical feasibility Keep NAD in the public domain 10

11. Pilot Results in Brief Identified the “haves” and “have nots” – 31 states plus DC have programs with varying degrees of completeness – Also Gila River Indian Community, Navajo Addressing Authority, Dept of Navy 1 Research on existing systems e.g. OpenAddresses.io, Community TIGER Identified minimum content guideline Schema comparison -FGDC and CLDXF -State schemas: AR, AZ, MA, NC, NY, RI, UT, VA, VT, plus DC & counties Identified best geocoding & address list data sources 11 AZ AR 1 https://www.nsgic.org/public_resources/2014_09_17_08_SPAWAR-Jansen.pdfhttps://www.nsgic.org/public_resources/2014_09_17_08_SPAWAR-Jansen.pdf

12. Minimum Content Guideline – 3 Components 12 The Address itself Address Number Street Name Subaddress City/Town/Place County State Zip Geographic Location of the address Lat/Long National Grid Coordinates Metadata about the address Address authority Address source Address date Unique ID Type (residential, commercial, etc.) Placement (rooftop, driveway access, etc.)

13. Review of the Minimum Content Guideline Round 1: NSGIC/Census project steering committee Round 2: All Summit attendees – Received 11 sets of written comments Guideline was revised/refined in response to each round of comments 13

14. Feedback on the Minimum Content Guideline Overall, feedback has been mostly positive: – “The ‘low barrier to participation’ is likely an excellent idea to encourage greater data coverage.” – “Simplicity in parsing roll up tools is critical.” – “… CLDXF maintains the applicable components of the FGDC and PIDF-LO standards while addressing the needs of NG9-1-1…” – “… the summary captures what is critical yet allows flexibility so that data can be updated and upgraded iteratively.” From a position paper released by NSGIC in April: – “As a point of emphasis, we strongly concur with the direction and recommendations made in the National Address Database Draft Minimum Content Standard (v8, March 2016) document under development by the USDOT.” 14

15. 15 FGDC/CLDXF Location Metadata

16. Pilot Participants Compiled Into NAD Schema 16

17. “Have Not” Status Goal was to find agencies (likely counties or tribes) that haven’t yet created their addresses Wanted entity that was interested, motivated, and willing to work with us. We did not want to create addresses that will then sit on a shelf. Jackson County, AR AGIO was a helpful partner, they want to finish statewide addresses by plugging few remaining holes 17

18. Jackson County, AR - Data Sources Countywide E911 Address List – 18k records – Some missing zip/city info – Some basic data scrubbing needed Countywide centerlines existed – No data scrubbing needed! Countywide parcels – 79% had some address info – Data standardization was needed E.g., for city name, address field, etc. 18

19. Jackson County, AR Geocoding Approach Multiple geocoding sources were used: – Melissa Data (commercial geocoding service) – County Parcels – County road centerlines – Census road centerlines If an address wasn’t matched in one source, the next source was used. Achieved a 77% overall match rate from the 18,469 records 19

20. Final Jackson County Geocoding Results 20 Source Total Records Matched % Matched* MelissaData7,07338% Parcel Centroids1,7009% County Centerline4,11223% Census/Tiger Centerlines1,3477% Totals:14,23277%

21. 21 Jackson County, AR

22. Preliminary Pilot Findings Tribal participation is going to be a challenge – Lots of outreach, no explicitly contributed data – Gila River data is part of AZ statewide collection Data sharing agreements to make data publically available could be a challenge Aggregating existing statewide/have collections was straight forward The schema will likely evolve, but needs to remain consistent with leading address schemas to allow for streamlined ETL 22

23. Coalition of the Willing Since the release of the minimum content guideline and schema, 15 additional address programs volunteered to develop their own ETLs – District of Columbia – New Jersey – Ohio – Utah – Virginia – 9 additional counties and 1 city from Missouri (Locals Helping Locals) Montana is on deck, seeking other volunteers through NSGIC 23

24. 16.8 Million Addresses 24

25. What’s Next Continue the Coalition of the Willing Choose platforms for development and production Identify funding for continued development Make the data available! Launch Data Challenge for “have nots” 25

26. Data Challenge Goal: develop an app to gather crowd sourced address information – Must collect the items identified in the minimum content guideline App can be used by – Local police and firemen – Real estate agents – Boy Scouts and Girl Scouts – FEMA Corps – Public Resulting address information would be used as “seed” data for local governments with no data and for QC/QA of existing data 26

27. For more information…. On the NAD Summit: https://sites.google.com/a/appgeo.com/nationaladdressdatasummit/home On the NAD Pilot: https://sites.google.com/a/appgeo.com/usdot-national-address-database-pilot- project/home 27

28. Contact Info Steve Lewis Chief Geospatial Information Officer U.S. Department of Transportation (202) 366-9223 steve.lewis@dot.gov 28