- July 29, 2014
- Public Safety
Data science and big data are hot topics in today’s business and academic environments. Corporations in a variety of industries are building teams of data scientists. Universities can barely keep up with student demand for courses. The hope is that new analytic methods, combined with more data and computational power, will uncover insights that would otherwise remain undiscovered. In the private sector, these new insights lead to new revenue opportunities and more targeted investments.
Recently, these methods have also shown promise within the public and social sectors, where large complex datasets have long existed. However, few governments or non-profit agencies can afford to hire the talent necessary to take advantage of these new methods.
To address this need, in 2013 The University of Chicago, in partnership with the Eric and Wendy Schmidt Foundation, started a fellowship program, Data Science for Social Good (DSSG). The first year of the program was a success with over 550 applicants from around the world and 36 fellows selected to work on 13 projects with non-profits, local government agencies, and federal agencies.
After seeing the success of DSSG in Chicago, Raj Bandyopadhyay, an Atlanta Data Scientist and Founder of Data Science ATL, decided to partner with the team in Chicago and Georgia Tech to start DSSG-Atlanta. This summer, the Atlanta fellows, from different universities and countries, have been working on five projects with non-profit and government agencies around the city. One project has been in partnership with the City of Atlanta and focuses on data from the City’s 911 Emergency Center.
The project’s initial concept was formed during a brainstorming session with several members of the broader Atlanta community and finalized between the Focus of Results Atlanta (FOR Atlanta) team in the Mayor’s office and the Atlanta 911 Center leadership. The overall goals of the project were to:
- Identify opportunities to reduce dispatch time
- Identify opportunities to more efficiently manage human capital
- Create a decision-making tool for Atlanta 911 Center leadership
So what type of person and team did DSSG-Atlanta attract to tackle these challenging social and public problems? The team of fellows, all from Georgia Tech, who worked on this project included:
- Kevin Johnson, PhD Student, School of Industrial and Systems Engineering
- Umashanthi Pavalanathan, PhD Student, School of Interactive Computing
- Alex Godwin, PhD Student, School of Interactive Computing
In addition, the team was guided by a mentor, Dr. Bistra Dilkina, who is an Assistant Professor at the Georgia Tech School of Computational Science and Engineering. She assisted in identifying the appropriate analytical approaches and developing the project scope and plan.
The fellows began the project by conducting site visits to the Atlanta 911 Center, exploring the data via descriptive statistics, and outlining variations across different geographically set areas pertinent to the Atlanta Police Department. They conducted analysis to explore the workload of dispatchers and how volumes affect dispatch time. Throughout the analysis they used a variety of tools, including R, R’s GIS capabilities, and the ggplot2 visualization library.
Their primary challenge was to find something meaningful within the roughly one million police dispatches that the Atlanta 911 center records each year. In all, they analyzed five years of data or approximately five million dispatches. These dispatches included more than just citizen calls and captured many self-initiated services, such as traffic stops, community appearances, or following a reckless driver. Each dispatch included several data points, including time of day, type of dispatch, unit dispatched, and geo-location.
It quickly became clear to the fellows that the traditional notion of workload (dispatch volume) did not capture the complexity of the work observed during the site visits. To weight dispatches more appropriately, a simple survey was developed by the team and then completed by 30 random dispatchers. Weights were then applied to dispatch types using a distribution from the survey results, effectively turning the notion of workload into an index. This new approach to defining workload provided a variety of new insights. Coupling this with several other predictors, the team was able to develop a model to test different scenarios. For example, one scenario that has gained traction as a result of the analysis is the movement of administrative dispatches (e.g. extra job check in and check out) to a single dispatcher, which creates greater availability for other dispatchers to focus on priority dispatches.
The fellows are leaving behind two products for the City. The first is an in-depth analytical report showcasing findings, modeled scenarios, and recommendations. The second will be a customized web-application, built using the D3.js Java Script library, which provides ongoing access to a dynamic tool for viewing the data and analysis. This tool provides the Atlanta 911 Center leadership the ability to effectively make decisions using real-time data and results.
The success of this project stems primarily from having an amazing team of dedicated fellows. Their complimentary skill set and hard work has provided the City of Atlanta with valuable insights that will ultimately result in better outcomes for residents. In addition, having a dedicated project manager within the City that planned and organized data for the project, guided the analytical directions, and facilitated interactions between the fellows and the Atlanta 911 Center was critical. This position ensured the analytical scope was practical and usable, the data was accessible, and also leveraged existing relationships and operational knowledge to move the project forward.
The prospects for continued use of data science methods within the City of Atlanta are blossoming rapidly. With each new undertaking and successful outcome on projects like DSSG-Atlanta, public and social sector leaders further understand the potential. Already, future projects, using data from the City’s soon to be launched 311-platform, are in the works. Over time, the value of this type of analysis will fully bloom and hopefully all public and social sector leaders will gladly expand investment in these methods and this type of talent.