CRISP Type 2/Collaborative Research: Multi-Agent Sustainable Water Decision Theory (MUST): Nexus of Water, Road, and Hierarchic Social Contractual Systems

提供可靠的清洁水对社区的健康和繁荣至关重要。然而，我们的社会经常面临着在没有考虑其社会经济背景或关键基础设施系统和服务的相互依赖性的情况下做出的决定的灾难性后果。例子包括2014年夏天俄亥俄州托莱多市供水系统中有毒藻类的事件，以及最近密歇根州弗林特市的铅污染水，导致联邦政府宣布进入紧急状态。一种普遍的看法是，主要基于短期成本考虑的决定，例如许多市政和城市办事处在面临资源限制时所作的决定，可能使水问题恶化而不是改善。这项跨学科研究开发了一个水利基础设施投资决策支持（WIIDS）模型，并使用丰富的数据库进行校准。WIIDS将可持续供水服务置于广泛的社会经济背景下，考虑系统之间内在的相互依存关系，以及与社区必不可少的合同、物理和服务基础设施之间的相互作用。本研究建立了一个先进的决策模型，该模型建立在一个大型数据库上，以支持不同层次的与水有关的决策。特别是，该模型描述了水系统、道路系统（提供人员和货物的流动性，同时通过机械和环境过程对水管施加负荷）和合同系统（影响影响水管耐久性的社区服务决策）之间的相互依赖性。该模型由两个不同的部分组成。一种是工程子模型，它使用历史数据和预测条件来预测供水基础设施选定部分的故障概率和相关成本。另一个是社会经济子模型，它利用来自工程子模型的输入，以及当前和预测的业务影响、交通影响、社区人口统计和关键基础设施（例如，医院、学校），提供全面的维修/更换/改道建议。该模型除了考虑直接解决方案成本外，还考虑社会成本、间接经济成本和公共卫生影响，允许对备选决策（例如，反应性维护与预测性维护）进行比较。

Providing reliable clean water is essential for the health and prosperity of communities. Our society, however, frequently faces the disastrous consequences of decisions made without considering their socio-economic context or the interdependency of critical infrastructure systems and services. Examples include incidents of poisonous algae in the water supply in Toledo, OH in summer 2014 and more recently, the lead contaminated water in Flint, MI which led to a declaration of a State of Emergency by the federal government. A common observation is that decisions that are primarily based on short-term cost considerations, such as those made by many municipal and city offices when facing resource constraints, can exacerbate water problems rather than improve them. This transdisciplinary research develops a water infrastructure investment decision support (WIIDS) model, calibrated with a rich repository of data. WIIDS places sustainable water services in a broad socioeconomic context that considers the inherent interdependencies of systems and interactions with the contractual, physical, and service infrastructures essential to communities. This research formulates an advanced decision model that builds on a large repository of data to support water-related decisions at different levels. Particularly, the model describes the interdependencies of water systems, road systems (that provide mobility of people and goods while applying loads on the water pipes via mechanical and environmental processes), and the contractual systems (that affect community service decisions that influence the durability of water pipes). The model consists of two distinct parts. One is an engineering sub-model that uses historical data and forecast conditions to predict the probability of failure and the associated costs for the selected sections of water supply infrastructure. Another is a socioeconomic sub-model that utilizes input from the engineering sub-model along with current and forecast business impacts, traffic impact, community demographics, and critical infrastructures (e.g., hospitals, schools) to provide holistic repair/replace/reroute recommendations. The model allows comparisons among alternative decisions (e.g. reactive vs predictive maintenance) considering social costs, indirect economic costs, and public health implications besides the direct solution costs.