Population Protection and Emergency Response

Optimizing emergency response is fundamental when time and resources are limited. Evolving complexities and uncertainties of on-the-ground situations offer significant challenges. We describe modeling and computational advances and real-time adaptive decision-support tools that optimize operations concerning treating the injured, providing temporary shelters, distributing essential supplies, and dispensing prophylactic medication.

Working with the Centers for Disease Control and Prevention, the work advances:

• Methods for solving large-scale NP-hard facility location problems.  The model takes in geographic and census data, socio-economic status, real-time twitter feeds, crowd-sourcing information, and on-the-ground population moves, and returns a set of facilities for temporary shelters and distributing essential supplies. Large instances involving millions of people can be solved within seconds,
• A unified framework for simulation and optimization that captures human cognitive and behavioral elements (both workers and affected population); and rapidly optimizes overall resource allocation and operations for best performance.
• 6-stage novel disease propagation model that tracks intra-facility disease spread and adjusts operations for casualty mitigation.

The work enhances national health security and response capability.  It prevents illness and saves lives as it improves the ability to reach out to affected populations.  It improves efficiency and reduces labor requirements and operating costs. It facilitates quality assurance and training of emergency personnel.  It empowers public health leaders in decision and policy making, helping them with training and planning without requiring high-cost and time-consuming exercises. It allows planners to rapidly test plans, and learn and observe system characteristics and trade-offs. It facilitates experimentation with and implementation of new ideas, and encourages collaborative competition across regions to achieve best performance.

Users have employed the system to plan and operate H1N1 mass dispensing sites; setup distribution nodes for earthquakes; shelter, screen and decontaminate affected populations in radiological incidents; and  establish service sites for people displaced by hurricanes and floods.  

By improving emergency response capability and efficiency under limited resources and stressed environments, our work facilitates rapid and effective interventions for at-risk populations, and lessens the subsequent disease cost, care burden, mortality, and economic impact.