The Counting US app is boosting the time and accuracy of point-in-time counts of people experiencing homelessness.
The Augusta, Ga., Housing and Community Development Department (HCD) cut the time it takes to process its annual census of residents experiencing homelessness from between 30 and 90 days down to minutes by replacing a traditionally manual task with technology.
Like many municipalities, Augusta had relied for years on volunteers with paper and pens to conduct these so-called point in time counts, which take place nationwide at the end of January. The U.S. Department of Housing and Urban Development (HUD) requires continuums of care -- federally funded regional or local organizations that oversee homeless services -- to conduct PIT counts at least biennially to determine funding levels.
For the 2019 count, Augusta switched from paper to digital tools from Simtech Solutions. Now, volunteers use the Counting US app -- usually on their own mobile devices -- to enter data on the number of people experiencing homelessness they see during the count period and to surveys some of them, asking questions about age, health and veteran status, for example, to give the local and federal government a better picture of trends in homelessness.
“The transition was disruptive. It was game-changing from 2018 to 2019,” said Daniel Evans, community development manager at HCD.
Volunteers download the app from the Apple App or Google Play stores, register for an account and enter a region-specific setup key to have their account associated with their region’s data store. After entering in this unique key, the Counting Us app is automatically configured with the survey questions and geographic boundaries for the region. The app uses conditional logic to control the questions displayed, which helps enumerators stay on track and avoid errors.
“We can prevent bad data from even existing,” said Matt Simmonds, the company’s president. “You can’t be 17 years old and be a veteran. You can’t be 8 years old and be a parent.”
Data is stored locally only when the mobile device is not connected to the internet. When it is connected, the data flows from the app to Simtech’s PIT regional command center, where it’s incorporated into dashboards, reports and maps. Administrators can click on map layers to see counts by city, county, ZIP code or census tract. They can also find out how many surveys were conducted within a specific region and run a HUD PIT report for the selected locations.
“You don’t know what you don’t know, so back in the paper survey days, when someone from an agency or a group of surveyors brought me a stack of surveys, and they said, ‘It’s all good,’ I said, ‘OK, great,” Evans said. “With the command center functionality from Simtech, we’re actually able to see the surveys from the users as they come in, so I can look in there and say, ‘Hey, Ms. Hartman, I’m seeing that you submitted 14 surveys, but you had 17 people enrolled in your program that night. What happened?’”
What’s more, each survey is geotagged so that officials can see which areas have been covered and where volunteers linger. All data can be exported in CSV format, and reports can be published as PDFs.
“Each survey is geotagged and date and time stamped and includes the name and contact information for the volunteer who collected the information,” Evans said. “This enables count administrators to monitor the count in real time and reach out to volunteers if they leave their designated coverage area or if there are any questions about the information that was submitted.”
PIT count reports are due to HUD in April, but with the technology, “the day that we close the account, we’re able to give an unofficial estimate,” he added. “I knew it was 560 people the day it closed because the results were right there in a nice, pretty printout. It comes out in a PDF with charts on it and broken down by subpopulations.”
Other communities are applying technology to their PIT counts. King County, Wash., used an app for the first time this year, as did the Southwest Arkansas Partnership. The San Diego Regional Task Force on the Homeless teamed this year with the San Diego County Sheriff’s Department and Chula Vista Police Department to fly drones over large areas to find homeless encampments before volunteers set out on their counts.
The Ohio Balance of State Continuum of Care was another of the 44 communities that used Simtech’s tools for the 2020 count. It also uses Tableau Software to create maps of the data, including color coding counties to show higher concentrations of homeless populations and the number of homeless households. Hovering a mouse over each county pulls up details such as how many individuals have mental illness or a substance abuse disorder.
Los Angeles County uses technology to do more than submit data to HUD. Last December, it stood up an online, interactive map that shows hot spots among its 88 cities with high rates of people experiencing homelessness and pins representing shelters, safe parking lots for people who live in their vehicles, supportive housing, winter shelters, family hotels used for crisis housing and interim housing in development or under construction. It went public Dec. 10, 2019, and had gotten more than 25,600 views as of Feb. 19.
To build the Los Angeles County Homelessness & Housing Map, Steven Steinberg, geographic information officer for the county, and his team used PIT data plus about a dozen other datasets from at least five departments. The first step was getting it all into the same format -- most of the data arrived by email in a Microsoft Excel spreadsheet. Now, a data work group is creating standards and formats so that when departments share their data, its coding and structure are consistent, he said, adding that the ultimate goal is to automate the process so that as data updates, so does the map.
“I think the real value that we were able to bring to the table came with the visualization of the data,” Steinberg said. “It’s one thing to look at a list of numbers of people experiencing homelessness, and addresses and census areas that they’re located in, but that doesn’t really give you a picture of where there’s an alignment or misalignment of those data with other resources that are serving that community.”