By integrating diverse data streams, our research team is able to provide policymakers with a detailed view of how emergency services resources can be deployed to maximize safety and efficiency.

fire truck being dispatched to the scene of a fire

Problem

During an emergency, accurate information about the types of equipment and personnel needed on-scene can save lives. Policymakers rely on past incident data to inform critical decisions on resource allocation, but the extreme variability of these data sets often forces them to base their judgments on an incomplete picture of overall trends.

Emergency dispatch centers provide one valuable data stream

Arlington County and SDAL have forged a relationship to liberate and repurpose existing municipal and external data sources to support the community through data-informed policy development.  By utilizing local administrative data combined with state and federal data, blogs and social media, and commercial data, a community can translate their values into action and solve problems of importance. Our first project to realize Arlington County's vision of public safety began with an analysis of fire and EMS data.

Methods

To re-create emergency incidents out of time- and location-specific information, several heterogeneous data streams must be linked. Leveraging the full range of available data required the SDAL team to develop analytical methods that could account for significant variations in source, accessibility, structure, and informational bias. In exploring the viability of external data sources such as social media, we are developing a process for working with local communities.

By linking these data and mapping them geographically, policymakers can increase the situational awareness of firefighters and EMS personnel. The rhythm of the county emerges as these data are a surrogate for identifying the flow of people through the county over the course of an average day, month, or season.

Impact

A comprehensive data analysis revealed that some of the largest demands on public safety resources are coming from vulnerable regions of Arlington County, where concentrated populations of individuals rely on 911 calls to address a wide variety of non-emergency concerns.

Map of Arlington County displaying regional trends in 911 calls

This finding allowed officials to hire a nurse practitioner to work with these vulnerable populations, teaching them to access a full range of municipal services and allowing emergency response resources to be directed to areas where they are needed most.

Our findings have also laid the foundation for a broader collaboration with fire departments throughout Virginia and California to develop a new data science portal that will improve the health and safety of first responders.

Publications

  1. Keller, S. and Shipp, S. 2016 (forthcoming) “Building Resilient Cities: Harnessing the Power Of Urban Analytics.” In J. Bohland, P. Knox (Eds) The Resilience Challenge: Looking at Resilience through Multiple Lenses, Charles C Thomas Ltd Publishers.
     
  2. S. Keller, S. Shipp, M. Orr, D. Higdon, G. Korkmaz, A. Schroeder, E. Molfino, B. Pires, K. Ziemer, and D. Weinberg. (2016). Leveraging External Data Sources to Enhance Official Statistics and Products.  Report prepared for the U.S. Census Bureau. Social and Decision Analytics Laboratory (SDAL), Biocomplexity Institute of Virginia Tech.
     
  3. S. Keller, S. Shipp, A. Schroeder, 2016. Big Data Change and the Privacy Landscape, Annual Review of Statistics and Its Application (in-press).

Back to top