Using data science to end Veteran homelessness
Abstract:Homelessness among Vietnam-era male combat veterans over the age of 65 with co-occurring substance use and mental health disorders is a major concern, especially in high-risk areas such as Maricopa County, Arizona, where extreme heat exacerbates health risks. Current federal and state laws restrict access to shelters for veterans who use illegal drugs, forcing them to live on the streets. This Capstone Project introduces VetLogistics as an artificial intelligence-enabled smartphone application that uses real-time geolocation sharing, actionable intelligence, automation, and optimization to bridge service gaps by connecting homeless veterans with critical resources like water, food, clothing, and tarps. VetLogistics is founded on Psychological Resilience Theory and encourages trust through trauma-informed, non-intrusive engagement. This initiative is consistent with the Grand Challenges for Social Work, which seeks to end homelessness and harness technology for social good. A participatory research approach was used, which included conducting interviews with veterans and analyzing homelessness patterns. The findings revealed that, in addition to transportation issues, systemic barriers associated with veterans exist. VetLogistics addresses these challenges by encouraging direct collaboration between service providers and veterans. VetLogistics uses an artificial intelligence-powered platform to improve resource distribution by aggregating real-time data to optimize resource allocation, predict homelessness trends, and inform policy interventions, with pilot testing carried out in collaboration with organizations such as US Vets Phoenix. The findings suggest that the VetLogistics application will play a significant role in reducing heat-related mortality among homeless combat veterans.