"Modeling Complexity in Disaster Environments"
Designing an information system to support decision processes in emergency environments presents an extraordinarily complex set of challenges for both technical and organizational researchers. Emergency managers need to understand the performance of the response system at several levels of operation simultaneously. While they need to know the detailed requirements for performance at a local site, they also need to recognize the consequences of a failure at any one site for its neighboring sites in the system. Further, managers need to recognize the consequences of cumulative failure at one level of operation that may lead to potential failure at other levels. Without the capacity to understand the interdependencies of a complex technical system, managers cannot anticipate the destructive consequences of the aggregation of apparently minor failures at single sites and take adequate decisions to prevent cumulative damage to the system. Innovative information and simulation technologies to support this decision process are essential. Providing real-time, decision quality information to practicing managers in timely, graphic form that displays the operation of a complex infrastructure system at multiple scalar levels would enable managers at their respective positions within the system to monitor interactions among the components and adjust performance reciprocally to reduce risk. This capacity would generate a self-organizing approach to the management of risk that would use local information to achieve a global goal. It introduces a second dynamic, the system’s informed adaptation to changed conditions that serves to interrupt the spread of dysfunction from the disaster event throughout the entire system. It is the interaction between these two dynamics – spreading dysfunction and cascading adaptation – that measures the system’s performance – or fragility - under threat. This task can most effectively be performed computationally. Currently under development at the University of Pittsburgh, the Executive Dashboard for Crisis Operations (EDCO) will enable practicing managers at different levels of responsibility and in different locations to monitor the status of critical operating conditions and systems on a wide area basis. The Dashboard will enhance the existing Interactive, Intelligent, Spatial Information System (IISIS) prototype that is being tested at the University of Pittsburgh. The IISIS prototype, designed in accordance with national standards for the Incident Management System and data representation in visual and graphic display, provides managers with real-time decision support to mitigate, prepare for, respond to, and recover from, extreme events. The function of the Dashboard will be to integrate real-time information from multiple sources in the disaster environment into a more advanced metric for assessing the level of risk to the whole community. The need for such a DSS is increasing with growing exposure to risk in metropolitan regions, critical interdependencies among lifeline systems, and greater uncertainty regarding threats from natural, technical, and deliberate disasters. This research is supported by National Science Foundation grant, CNS ITR#0325353, Secure CITI: A Secure Critical Information Infrastructure for Disaster Management.