Presentation by Dr. Robert Stewart, Senior Scientist, GeoAI Group, Oak Ridge National Laboratory.
Understanding how humans occupy the built environment is critical to a wide array of applications, including urban resiliency, natural hazards loss analytics, energy efficiency, transportation, and population distribution modeling. Globally, efforts to understand building occupancy continue to rely on a patchwork of disparate resources that vary widely in availability and sophistication. To use these effectively still requires expert judgment to harmonize the data and produce meaningful occupancy estimates. Despite the extraordinary societal impacts of big data, including IoT, social media, cell tracking, webcams, imagery exploitation, and a wide array of open-source data, the challenge of estimating population is still not a “solved problem”. The problem largely remains an open interdisciplinary challenge requiring the qualitative and quantitative contributions of social scientists, population scientists, remote sensing practitioners, computer scientists, and data scientists. Against this backdrop, there is a unique opportunity to leverage the qualities of Bayesian reasoning to explicitly harmonize disparate data, capture process uncertainty, and intuitively convey occupancy estimates to a wide range of consumers. In this talk, I present a Bayesian model that: 1) provides an explicit, systematic approach to engaging data, data harmonization techniques, and expert judgment in the production of occupancy estimation, 2) probabilistically evolves and refines estimates over time as new data and expertise emerge, and 3) retains and characterizes uncertainty emerging from expert judgment, data, and inference. I also provide insight for working and interning at Oak Ridge National Laboratory (DOE's largest science and energy laboratory).
Robert is a senior scientist in the GeoAI group at the Oak Ridge National Laboratory (ORNL) and adjunct associate professor of Geography at the University of Tennessee. Roberts leads projects engaged in a wide array of R&D including machine learning, spatio-temporal analytics, data mining, big data workflows, simulation, visualization, and tool development. His work is informed by and applied to a wide range of use cases emerging from population dynamics, maritime safety, geomatics, urban dynamics, security, energy-water nexus, health, environmental risk and many others. His own research is focused on applied mathematical, statistical, and computational methods in the areas of spatio-temporal analytics, probability modeling, and uncertainty quantification with an emphasis on risk and decision support. As a faculty member at UT, Robert engages graduate students in geography, mathematics, and the Bredesen Center Data Science Ph.D. program. He regularly serves on thesis committees, advising, and facilitating internships at ORNL.
To view the presentation, see the CGA calendar for Zoom details.