Optimal placement of ambulance stations using data-driven direct and surrogate search methods

Area of Science:

• Operations Research
• Public Health
• Emergency Medical Services

Background:

• Ambulance station placement significantly impacts emergency response times.
• Optimizing ambulance locations is crucial for improving emergency medical services (EMS) efficiency.
• Current ambulance station allocation in Umeå may not be optimal for minimizing response times.

Purpose of the Study:

• To determine the optimal placement of ambulance stations in Umeå municipality.
• To minimize the median response time for priority 1 alarms (MRT1).
• To evaluate and validate different black-box optimization approaches for EMS resource allocation.

Main Methods:

• Utilized data-driven approaches on Västerbotten county ambulance data (2018).
• Employed black-box optimization methods: GPS (smooth-free) and surrogate (smooth-based).
• Compared one-by-one vs. all-together approaches and local vs. global data strategies.

Main Results:

• Relocating stations can reduce MRT1 by 80-100 seconds.
• Global data yielded better solutions but required more computation.
• The one-by-one approach was more effective and less time-consuming than the all-together approach.

Conclusions:

• Relocation of ambulance stations significantly decreases MRT1.
• Optimization can partially compensate for the loss of ambulance resources.
• The one-by-one approach is recommended to reduce optimization cost and dimensionality.