Leveraging shared autonomous vehicles for vulnerable populations during pre-disaster evacuation

More and more severe weather events necessitate better evacuation strategies, particularly for vulnerable populations without access to private vehicles. This study investigates opportunities for shared autonomous vehicles (SAVs) in faster evacuation of the Houston, Texas coastline, with a focus on the crucial first-mile connection to bus stations. Microsimulations of various scenarios reveal performance of SAV systems with different seating capacities, fleet sizes, shared-ride acceptance levels, and coordination with bus departures to shelters. The Houston evacuations appear most effective with smaller, 5-seater SAVs, operated at a ratio of 1 SAV per 14 evacuees. The results underscore the importance of SAV-to-bus coordination, demonstrating that strategic ride-sharing and scheduling can significantly reduce evacuee wait times and fleet operating costs while maintaining timely arrivals at distant shelters. This work illuminates the practical implications of integrating SAV technology with existing transit systems to improve emergency management for carless populations.