The decision support software developed through this project represents a new tool to aid decision makers in making informed, efficient decisions. The software tool builds on earlier research that explored interactions among technical systems, organizational processes, and physical and social conditions that affect information flow in managing risk and uncertainty. The design approach develops practical decision models using Bayesian networks and influence diagrams to assess uncertain conditions, based on systematic identification of interdependencies among the component technical, organizational, and knowledge systems that characterize urgent operating environments. It integrates technical skills in computer programming and simulation design, grasp of business dynamics and marketing, and understanding of context, policies, and constraints of emergency management. This decision support module identifies options available for action, given actual constraints and near-real time information from multiple sources, and calculates the probability of success of each option, based on the collective judgment of experienced emergency managers. This decision support tool addresses problems of scalability and simultaneity in information flow processes that have hindered inter-organizational decision making in large-scale, regional disasters. Modeling potential outcomes can systematically enable managers to compare a broader range of options.
If successfully developed, this dynamic decision support tool may have a transformative effect on how communities manage risk. As the number, type, and severity of disasters increase in a global society that depends increasingly on large-scale systems in transportation, power generation, communication, and gas, water and wastewater distribution, the cost and consequences of failure in these socio-technical systems escalate exponentially. Managers may need improved tools to monitor these interdependent operating systems simultaneously, and adjust and adapt the balance between demand and resources available to manage sudden surges in demand from extreme events. This technology has the potential to benefit communities through helping local governments, nonprofit organizations, and small businesses increase their capacity to manage their continuing exposure to risk, but reduce losses by more informed, effective decision making.
Louise K Comfort, Principal Investigator, University of Pittsburgh Project/Grant Period: 10/01/2012 - 03/31/2013 Major goals: Two major goals were stated for this project: 1) to identify the market for a prototype Dynamic Decision Support System (DDSS) to assist emergency managers in considering a wider range of options for action in dynamic, uncertain environments; and 2) to determine the feasibility of initiating a start-up company to bring the DDSS to market. We defined four basic tasks to undertake within six months, October 1, 2012 - March 31, 2013: 1) develop a business plan for a start-up company suitable for review by third-party investors; 2) develop a SBIR proposal to support product development for the proposed start-up company; 3) engage students in market research necessary to extend the proposed prototype in a relevant policy domain, and 4) demonstrate our working options module to an invited audience of potential clientele, investors, and interested decision makers in a half-day workshop. Major Activities: We undertook four major types of activities: 1) intensive interviewing of emergency managers at local, state, and federal levels; 2) participation in local public safety fairs and regional meetings of emergency managers; 3) field observation and interviews with key emergency management officials following Superstorm Sandy; and 4) demonstrations of our Dynamic Decision Support System to key officials in interested agencies. Specific Objectives: Our objectives in conducting the interviews included: 1) determining specific needs of emergency managers in making decisions before, during, and after extreme events; 2) assessing the current state of information management systems used by the respective emergency service agencies; 3) identifying the 'added value' of our DDSS for emergency managers; and 4) determining where and when our DDSS could most easily and effectively be introduced into the large, complex, interactive emergency response system. Significant Results: The Center for Disaster Management received a Community Resilience Innovation award from FEMA in late May, 2013 to conduct a pilot project for building community resilience to disaster risk among public, private, and nonprofit organizations in Allegheny County, PA. Importantly, the award represents FEMA’s recognition of the merit of the DDSS for communities seeking to improve decision making in risk environments, using a 'whole community' approach to risk reduction. Products Two academic papers were presented at international conferences, six invited presentations were made to interested policy makers in federal agencies, and 63 interviews were conducted with local, state, and federal managers with emergency management responsibilities. The papers reported our development of a prototype dynamic decision support system (DDSS), using Bayesian modeling to identify options for action in complex, changing environments. The DDSS allows users to model interdependencies among physical conditions, resource constraints, and time constraints. Based on known constraints, the DDSS rapidly calculates the probability of effectiveness for different options that could be taken to reduce risk, protect lives, and secure property, based on recognized criteria. We developed protocols for knowledge elicitation and templates that assist practicing managers in clarifying decision rules. Figure 1 below illustrates the Bayesian reasoning process. A detailed description is available at : www.cdm.pitt.edu/CDMProjects/DynamicIncidentManagement/tabid/388/Default.aspx Participants Two other participants engaged in this research: Mark Voortman, Postdoctoral Fellow, CDM, and Paul Petrovich, Mentor, Center for Enterprise Development, University of Pittsburgh. Four agencies/organizations serve as partners to this project: Allegheny County Emergency Management Services, Pittsburgh, PA; Federal Emergency Management Agency, Washington, DC; National Association of Workforce Boards, Washington, DC; and the US Army Corps of Engineers, Pittsburgh and Mobile, Alabama Offices. Impact The prototype dynamic decision support (DDS) module offers an innovative approach to linking the traditional process of situation assessment with a more explicit calculation of available options for decision in complex, dynamic environments. The DDS module enables rapid assessment of possible strategies of action under varying conditions of knowledge, resources, skills, and time. If widely available and used by personnel in public, private, and nonprofit organizations with responsibilities for managing risk, the DDS will contribute to building a more knowledgeable, capable, informed society that is increasingly able to manage its own risk, reduce losses, and enhance the sustainability of the natural and built environments. Changes in approach and reason for change Our initial approach was to target local emergency managers as potential users of the DDS who represent a large potential market. Yet, response from public agencies at different levels of responsibility indicated that developing a carefully designed demonstration project for the DDS would be the most effective strategy to gain interest and adoption by emergency services agencies. Our Community Resilience Innovation grant, funded by FEMA, enables us to design and conduct a pilot project in collaboration with Allegheny County Emergency Services, PA. This pilot project will create the basis for further extension of the DDSS model to other counties and possibly to other states and the nation.
