Using Advanced Machine Learning For Air Force COVID-19 Hotspot Prediction, Crisis Response, and Downstream Assessments

Fraym uses AI/ML to produce local information about populations in austere geographies. We aim to produce predictive models in support of “COVID-19 Focus Area 1: Decision Support in Combating the Virus.” We will collaborate with AFWIC to provide the Air Force with unprecedented insight for COVID-19 force protection, resource planning, and operational practice changes in Africa. Specifically: Identifying virus ‘hotspots’ and anticipating virus spread — Fraym will identify local ‘hotspots’ of populations that exhibit the highest risk for contracting and transmitting the virus, based on population density, health characteristics, age, healthcare access, and more. This will illustrate where U.S. forces operate in high-risk areas, and provide critical situational awareness on the most severe likely outbreaks.   Identifying humanitarian risks — We will inform USAF response planning by modeling the most at-risk populations for food shortages, health infrastructure failings, and economic shocks.  We will use this to enable efficient aid and resource allocation for US and partner forces. Anticipating instability and emerging threats— Finally, Fraym will model risk for political unrest, extremist activity, and other destabilizing events as a result of the pandemic. This will inform U.S. strategy to maintain power projection, guide partner collaboration, improve operational planning, and allocate resources to mitigate destabilizing events. Fraym is confident that this three-stage effort to provide ML-powered predictive models will enable improved force protection, resource allocation and operational effectiveness in Africa.