I work at the intersection of data science, spatial analysis, and public sector problem-solving. My work focuses on building tools, dashboards, and analytical workflows that turn complex data into clear operational insights for agencies, researchers, and decision-makers.
Broadband Data Stroll v3: Transparency in Federal Broadband Investments
I am proud to share my work as a Lead Data Scientist on the Data Analytics team at the U.S. Department of Commerce, Office of Inspector General on the launch of the Broadband Data Stroll Version 3. This effort was developed in close partnership with the US Department of Agriculture OIG and the FCC Office of Inspector General to bring greater transparency to federal broadband investments and to support oversight work across agencies.
A major part of my contribution was developing the Broadband Serviceable Location Mapper (BSLAM) using ArcGIS Experience Builder. This third release expands on earlier tools by performing extensive analytics using the FCC’s Broadband Serviceable Location Fabric and Broadband Availability data to help users better understand where broadband investments are being made and where service gaps remain.
Key updates in this release include:
- Expansive BSL mapping across the country for terrestrial-based broadband access
- Large-scale analysis of hundreds of millions of data points and more than $111 billion in federal investments
- Cross-agency comparisons of broadband awards from four agencies within the same geographic area
- Enhanced tracking of broadband availability over time and identification of areas without terrestrial-based service
Building the spatial data workflows and oversight dashboards behind this work has been a long-term effort, and it is exciting to see the transparency and decision support this platform is now enabling.
You can explore the BSLAM project here: Broadband Serviceable Location Mapper
Developing the FQM Mapper: Enhancing Data Quality at the U.S. Census Bureau
During my time at the U.S. Census Bureau as a Data Scientist, I spearheaded the development of the Field Quality Monitoring (FQM) Mapper, an ArcGIS Online-based mapping and analysis dashboard. This tool was designed to enhance the FQM program by providing near-real-time identification and corrective action of potential data issues in survey field data collection.
The FQM Mapper offers a range of functionalities, including:
- Visualizing the distribution of field data across different regions.
- Identifying statistical outliers and potential anomalies.
- Monitoring key performance indicators.
- Facilitating data-driven decision-making for improved data quality.
This project involved collaborating with personnel focused on data engineering to develop schemas and tables for metrics calculations. The Census Bureau’s Unified Tracking System’s (UTS) data warehouse was used to extract, transform, and load tables for metrics calculations.
I led the development of coding solutions for outlier and anomaly detection for Field Representatives and Field Supervisor Areas using AHS INR data. My research and utilization of novel coding methodologies using Python and R were integrated within the U.S. Census Bureau survey database infrastructure.
Extensive documentation of outlier/anomaly detection and metric development in Jupyter Notebooks provided a clear and robust work trail from data extraction, transformation, analysis, and visualization.
The development and implementation of the FQM Mapper significantly improved the efficiency and effectiveness of the FQM program, enabling better monitoring, analysis, and reporting capabilities. It also supported informed decision-making and process improvements in data collection operations.
One Health, GPAWSS and Machine Learning Integration
While I was at the US Army Public Health Center, working as a Epidemiologist, I gave two presentations detailing the data processing and limitations for companion animal surveillance within the Department of Defense. GPAWSS is a surveillance platform designed to provide surveillance data to inform commanders and VCOs of the distribution, frequency, and incidence of various companion animal diseases. The platform uses multiple heterogeneous data streams including: Remote Online Veternary Record (ROVR) EHR data, laboratory data, and data from a civilian corporate companion animal practice. Data in GPAWSS is managed by the One-Health Division and is displayed on an interactive, web-based platform (Tableau and R-Shiny) tracking disease frequency and incidence globally. GPAWSS also has outbreak detection capabilities.
One presentation detailed the pitfalls of utilizing Master Problem Lists as the main effort of companion animal disease surveillance and what can be used to “fill in the blanks” or “drain the sea of uncertainty”. https://waughr.us/images/GPAWSS_Data_PRES.pdf
The second presentation dealt the Integration of high-performance computing and machine learning within the Army Veterinary Service in order to improve Surveillance of Companion Animal Disease within the Department of Defense. This presentation was part of the Association of Veterinary Infromatics Talbot Symposium 2020. https://waughr.us/images/20200817_ML_GPAWSS.pdf
My work at the University of Florida
My PhD Dissertation Presentation investigating the integration of spatial data within the transmission of zoonotic diseases within two distinct populations. https://www.dropbox.com/s/759iry2pvfpyrp0/Dissertation_Defense_031518_V11.pptx?dl=0
I created some maps for a economic article in investigating the use of Urgent Care Centers And Hospital Emergency Rooms in Florida https://www.bebr.ufl.edu/survey/website-article/use-urgent-care-centers-and-hospital-emergency-rooms-cross-sectional-survey
My American Public Health Association Annual Meeting Award winning presentation analyzing trends in Uber data and Alcohol related arrests in a college town https://www.slideshare.net/mobile/waughsh/aphapresentationasof102016edit-1
My GIS Analysis in 2010, investigating the risk of Oil Spill hazards to the US Gulf Coastline. https://www.slideshare.net/mobile/waughsh/final-project-11370910