Geospatial and Data Science Team Pilots ‘Data Science for the Public Good’ Project in Marshalltown

Map showing the network of bus stops in Marshalltown relevant to where children and older adults reside in town.More people are riding the bus in Marshalltown. In fact, ridership increased by 11% this past year, the highest percentage increase in ridership of rural transit providers in Iowa.

Region 7 of the Federal Transit Administration took notice and recognized the Marshalltown Municipal Transit team with an Award of Excellence at a recent conference. This award came after the Iowa Public Transit Association recognized Marshalltown transit administrator Kevin Pigors and Marshalltown Municipal Transit for the same achievement earlier this year.

This significant increase in ridership can be attributed to data-driven decisions made by Pigors, who joined the Marshalltown Municipal Transit team in January of this year.

Data are valuable tools, if one knows where to find them and how to use them. Unfortunately, small, rural communities don’t always have the resources and expertise necessary to use pertinent data effectively in the planning and decision-making process.

In an effort to assist these communities, Iowa State University initiated a “Data Science for the Public Good” pilot project in spring 2018 in Marshalltown, with the goal of engaging researchers with local leaders to help them solve problems using their own data and knowledge. The project results gave Pigors a framework within which to make positive changes to the community’s public transportation system.

“Kevin…instantly went into action using that data and some other things that he knew generally about our operation to figure out how we can ensure that we are getting more riders onto our system,” said Jessica Kinser, Marshalltown city administrator.

Map showing the network of bus stops in Marshalltown relevant to population density.“Specifically, there’s an area where we have a pretty heavy multifamily complex that was an area that really was not being served in any sort of way. He made route adjustments to get that area on a bus route…it’s a low-income housing tax credit project as well, so there’s a big need to try to serve that area,” she said. “And that was one of the areas that came up through the mapping and data analysis as not being served where you had a heavy presence of children.”

The Marshalltown pilot project was a collaboration of several colleges and departments at ISU, with the ISU Extension and Outreach Community and Economic Development Geospatial and Data Science staff playing an integral role in the project.

Data Science for the Public Good (DSPG) is a program developed at Virginia Tech that connects data science students with communities that can benefit from data-driven research. According to Christopher Seeger, ISU professor and extension landscape architect and a key collaborator on the project, the goal of DSPG is to employ a process called Community Learning through Data Driven Discovery (CL3D).

CL3D involves “community-based research where the community participates in asking and answering questions that drive information gathering and provide insights relevant to program or policy decisions.”

“Virginia Tech developed this CL3D process and they’ve been using it with success for a while now,” said Seeger. “But how does it work in other localities where you might not have as much data readily available in a format that we can use?”

The first step of a DSPG project is the data-discovery process, during which researchers determine what data sources they should use to solve a community problem. The pilot project in Marshalltown was focused specifically on the community’s transit system to determine whether or not it was serving the target audience.

“CL3D is a collection of approaches for engaging with stakeholders as well as technologies that support this process—different tools that are good at helping you do data cleaning, helping you do data maintenance and providing meta data,” said Seeger.

“Part of [the process] was to discover on our own what we could do. Interestingly enough, our methods are similar to what [Virginia Tech is] already doing, but the tools are slightly different,” he added.

One of the challenges the project team faced in Marshalltown was the lack of certain types of data, as well as existing data needing to be cleaned up and consolidated in order to be useful.

Because there were minimal digital data on pedestrian infrastructure, the team utilized OpenStreetMap, aerial imagery, and other tools to create much of the pedestrian network data—primarily sidewalks, trails, bus stops, and bus-stop accessibility.

Part of this process involved conducting an inventory of the locations of all the bus stops. A student rode the bus and mapped the bus stops using GPS. When the bus stopped, the student recorded it, as well as where there were bus-stop signs and the bus didn’t stop.

By juxtaposing the GPS route information collected with existing demographic and economic data, the project team was able generate maps that illustrated how well the Marshalltown public transportation system was serving different types of residents.

For example, the map on page 6 illustrates how well the system serves children and older adults. The map on this page shows the city’s bus stops in relation to population density.

A second challenge came in the form of a tornado that struck Marshalltown on July 19, 2018, causing a great deal of damage and changing the focus of the community, the actual infrastructure, and the availability of city staff to help with the project.

Despite these issues, Seeger, along with GIS systems analyst Bailey Hanson and systems analyst Rakesh Shah, was able to work closely with Kinser, Pigors, and other city staff, who shared data with the project team, reviewed several iterations of the maps, and offered feedback.

The Marshalltown pilot project is the forerunner to a three-state initiative that includes teams from Iowa, Virginia, and Oregon. The Iowa team, including project lead Cassandra Dorius, assistant professor in the College of Human Sciences; Shawn Dorius, associate professor in the College of Agriculture and Life Sciences; and Seeger; have partnered with ISU Extension and Outreach to infuse data science into agriculture research, workforce development, and extension activities, with a focus on rural communities.

A goal of the project is to lead to advances in the applications of big data across problems in health and nutrition, food and agriculture, youth development, and community resource development.

“Rural communities collect data every day on issues of transportation, housing, emergency management, parks, potholes—issues that effect the whole gamut of community life. Our team works with communities to collaboratively solve the problems they find most pressing with data they already have, but may not be using,” said Cassandra Dorius.

“By leveraging the strengths of CES, CED, and university faculty, as well as community expertise and data, we can create a sustainable approach for future evidence-informed decision making, like what you see in Marshalltown,” she said.

The project will provide training on DSPG to undergraduate and graduate students, faculty, and extension professionals engaged or interested in community development, food, health, nutrition, youth development, forestry, agriculture, and natural-resource management in rural communities. The students and faculty will form community-focused research teams and complete collaborative data-science projects in partnership with local rural community stakeholders.

Students interested in learning more about DSPG can email questions to dspg2020@iastate.edu.

The Marshalltown pilot project also led to the Iowa League of Cities, in partnership with ISU Extension and Outreach CED, obtaining an AARP Community Challenge Grant.