CAREER: Extreme Weather Events and Emergency Medical Services: A Discrete 0ptimization Modeling Framework

该学院早期职业发展（Career）项目奖旨在研究如何通过考虑其社会背景下的基本决策问题，重新制定和重构具有相当社会相关性的重要服务系统模型。本职业规划的动机是紧急医疗服务（EMS）系统中重要的、及时的资源分配问题，即如何在极端天气事件中提供协调的EMS对医疗紧急情况的响应，从而整合迄今为止文献中分别解决的两类危害缓解问题。特别是，本研究探讨了如何将服务器（医疗单位）最佳地调度到地理上分散的客户（患者），以及在正常和极端天气事件期间调度策略如何变化。紧急医疗调度协议通常是为在正常天气条件下运行的系统设计的。一般来说，对于在极端天气条件下运行的系统如何改变调度协议，几乎没有指导。研究计划的核心挑战是重新制定和重新构建新的硬离散优化问题类别，这些问题捕获了围绕服务系统的社会环境，并通过探索新的算法和启发式以及通过表征模型的结构特性来解决离散优化模型。在这个项目中开发的离散优化模型提供了新的公式，通过调查EMS系统的特殊需求来重新制定和重新构建新的问题类别。这些新模型和算法可用于为EMS系统的设计和操作提供基本见解，以应对极端天气事件期间出现的医疗紧急情况。具有挑战性的扩展研究了如何同时定位和调度医疗单位，并利用委托代理问题模型研究了紧急医疗调度的博弈论方面。教育部分的核心挑战是创建一系列教学和推广活动，通过推广教育公共安全领导人；创建一系列关于将先进的分析工具应用于风险和危害的播客；将研究作为利用社交网络工具（博客、FaceBook和Twitter）进行推广的工具；开发国土安全和应急管理数学模型课程；以及指导学生。通过将研究模型纳入外展活动和其他教育活动，并利用外展和教育组件更好地为研究模型提供信息，可以实现研究和教育组件之间的整合。这一项目的更广泛影响是将研究与重要风险和危害问题的教学结合起来，并将其应用于紧急医疗服务。研究部分导致了实时分配稀缺EMS资源的新策略，这有可能改变在正常和极端天气事件下EMS资源的分配方式。与EMS社区合作伙伴的协作安排为该项目提供了丰富的实际数据，该社区合作伙伴将是该研究的直接受益者。教育部分创建了一系列针对公共安全领导人并使用在线社交网络工具的外展活动。教育和推广活动综合并反馈到研究活动中。这个项目的一个积极影响是将知识传递给公共安全领导者和研究者领域以外的学生。最后，研究成果、教育材料和外展活动计划将广泛传播给研究人员、从业者和整个社会。

This Faculty Early Career Development (CAREER) Program award investigates how to reformulate and reframe important service system models that have considerable social relevance by considering the fundamental decision issues within their social context. This career plan is motivated by important, timely resource allocation problems in emergency medical service (EMS) systems, namely, how to provide a coordinated EMS response to medical emergencies during extreme weather events, thus integrating two types of hazard mitigation problems that have been addressed separately in the literature thus far. In particular, this research investigates how to optimally dispatch servers (medical units) to geographically dispersed customers (patients) as well as how dispatching policies change during normal and extreme weather events. Emergency medical dispatching protocols are typically designed for systems operating under normal weather conditions. In general, little guidance exists for how dispatching protocols may change for systems operating under extreme weather conditions. The central challenges of the research program are to reformulate and reframe new classes of hard discrete optimization problems that capture the social context surrounding service systems and to solve the discrete optimization models by exploring new algorithms and heuristics as well as by characterizing the structural properties of the models. The discrete optimization models developed in this project provide novel formulations that reformulate and reframe new classes of problems by investigating the particular demands of EMS systems. These new models and algorithms can be used to provide fundamental insights into the design and operation of EMS systems in response to medical emergencies that arise during extreme weather events. Challenging extensions investigate how to simultaneously locate and dispatch medical units and investigate game-theoretic aspects of emergency medical dispatch using principal agent problem models. The central challenges of the educational component are to create a portfolio of teaching and outreach activities that educates public safety leaders through outreach, to create a series of podcasts about applying advanced analytical tools to risk and hazard applications, to use the research as a vehicle for outreach using social networking tools (blogs, FaceBook, and Twitter), to develop a course on mathematical models for homeland security and emergency management, and to mentor students. Integration between the research and educational components will be achieved by including the research models in the outreach and other educational activities and by using the outreach and educational component to better inform the research models.The broader impacts of this project are aimed at integrating research with teaching and learning on important risk and hazard problems with application to emergency medical services. The research component leads to new policies for allocating scarce EMS resources in real-time, which have the potential to transform how EMS resources are allocated during normal and extreme weather events. A collaborative arrangement with an EMS community partner provides a wealth of real-world data for this project, and this community partner will be the direct beneficiary of this research. The educational component creates a portfolio of outreach activities that are targeted at public safety leaders and that use online social networking tools. The educational and outreach activities synthesize and feed back into the research activities. A positive impact of this project is the transmission of knowledge to public safety leaders and to students outside of the investigator's field. Finally, research results, educational materials, and outreach activities programs will be broadly disseminated to researchers, practitioners, and to society at large.