Optimize EMS Responses During Extreme Events

Overview 

Optimizing ambulance allocation and routing is one of the most efficient ways for EMS to save more lives at virtually no cost. However, current EMS software was developed under models that assume normal demands. They are unable to adapt to disasters such as the COVID-19 pandemic, where traffic patterns change, case clusters emerge, and hospitals rapidly reach capacity. Decisions which are optimized for normal times can suddenly become very inefficient, leading to significant delay in care. Ideally, emergency responders would like to synthesize real-time information on case clusters, hospital capacity, wait times, and traffic situations to coordinate responses between all ambulances.  

Findings 

First, the project team conducted a retrospective study quantifying the impact of Covid-19 on the temporal distribution of EMS demand in Travis County, Austin, Texas and proposed a robust model to forecast Covid-19 EMS incidents.  

Second, the team reduced response time by providing optimal ambulance stationing and routing decisions by solving two stage stochastic and robust linear programs. To this day, there is little open-source code and consistency in ambulance dispatch simulations. They began to bridge this gap by publishing OpenEMS, an end-to-end pipeline to optimize ambulance strategic decisions.  

Finally, the team provided a detailed case study on the city of Austin, Texas. The team found that optimal stationing would increase response time by 88.02 seconds. Further, the team designed optimal strategies in case Austin EMS were to permanently add or remove one ambulance from their fleet.