The Census Bureau issued a request for information for tools to flag potential changes in addresses, roads and boundaries that might affect the accuracy of the 2020 census.
The Census Bureau is looking for software and other techniques that would help automate the detection of changes in addresses, streets and housing structures that might affect the accuracy of the upcoming 2020 census.
The Bureau’s Geography Division is responsible for updating address data and street centerlines so that population and housing figures for censuses and surveys are correctly allocated.
In the past, this was accomplished by sending out “field listers who walked every street in the nation to verify and, where necessary, update the address list,” according to a request for information recently issued by the agency.
However, in order to cut costs and maintain accuracy, the Bureau is now looking for an automated solution to spotlight locations where changes in addresses, roads and boundaries are occurring. The Bureau described its objective as, “to know ideally prior to when a new house, community, or other residence has been built or at the time it occurs or shortly thereafter.”
Also of interest would be to know where housing and related buildings have been demolished and their addresses retired as well as those that have been abandoned by local government or considered unfit for habitation, according to the Census RFI.
To narrow down the field of potential suppliers of change-detection applications, the Census Bureau offered a list of questions to help qualify the solutions, including:
- Do you have the ability to detect change between two vintages of address postal and location data?
- Do the techniques identify change for a specific building or address?
- Can you detect changes in structures as well as associated attributes, including mailing address and the number of units within a structure?
- Can you identify clusters of common housing unit such as mobile homes, trailer parks, multi-unit or row houses and town houses?
- Do you have the ability to identify areas where growth is not probable/possible?
- How frequently do you acquire new data and perform change detection on that data?
- Do you use crowdsourcing as a data acquisition source? If so, describe the process you use to validate the quality of information obtained through this process.