Multilayer graph networks for modeling and analyzing exercise scenarios in civil protection and disaster response

Abstract

Based on a reflection of the emergency sector as critical infrastructure and with reference to the concepts of preparedness and prevention as well as criticality, the dissertation examines to what extent software applications can be used to support the planning and evaluation of practical exercises in civil protection and disaster response. A network-theoretical approach is chosen which allows to model scenario-based exercises as dynamic networks. In particular, it is thereby achieved that actions and structures from the exercise can be represented and analyzed in their temporal context through different types of relationships or interactions. First of all, the unique dual role of the emergency sector in the system of infrastructures is addressed and the importance of the practical exercise as a training method and for simulating real operations for the sector is explained. Based on a literature review and interviews with practice partners from different organizations and authorities of civil protection and disaster response in Germany, the exercise will be analyzed with regard to its objectives and the methods used. In this context, especially practical approaches and problems will be discussed and compared. The analysis shows that each exercise is very different in terms of its underlying objectives and approaches and depends strongly on the needs of the respective organizations and authorities. Furthermore, it becomes apparent that especially the planning of such exercises is very complex and therefore there is a need for support tools for the development of scenarios and the definition of exercise boundaries. Also in connection with the evaluation of the exercises, there is a need for a software solution to support the processes. It is apparent that, despite the existing awareness of the importance of evaluations, hardly any systematic procedures are available for this purpose, and that the evaluation as a whole is often not carried out with the necessary consistency, partly due to time constraints. It is particularly noticeable that communication between the participants is often identified as a potential for errors during the exercise, but that the actual communication interactions taking place during the exercise are not yet systematically recorded and analyzed. In addition to the analysis of the practical exercise, the dissertation also evaluates the relevant research literature from the fields of organizational research, emergency management and risk and criticality research. The potential of social network analysis for the aspects of planning and evaluation of exercises is identified and decisively worked out, which, supported by statements from practice, allows to define requirements for a software system. In order to meet the specific requirements of the exercise, a concept for a network-based support software is designed and implemented in a tablet-based demonstrator application called ScenarioBuilder BOS. The application supports the user in modeling and developing exercise scenarios as well as in analyzing and evaluating them in various ways. For example, simulations of cascading effects can be carried out or the centrality of the various actors in the network can be compared. The aim of the application is to enable the user to develop interpretation approaches and to question actions and relationship structures by presenting the scenario in different perspectives. In order to evaluate the application and the benefit of social network analysis as a methodology to support the planning and evaluation of exercises, the thesis describes four use cases for the ScenarioBuilder BOS, three of which are evaluations of real civil protection exercises in different organizations and authorities. The fourth case describes the use of the application to develop the scenario of a fictitious exercise and serves to validate the simulation and other functions of ScenarioBuilder BOS. The use cases show that the application and the associated methodology of social network analysis has a great potential especially for the evaluation of exercises. It enables a systematic recording and evaluation of communication relationships in particular and can therefore make a valuable contribution, for example, to assessing the workload of actors, analyzing compliance with command structures or explaining dynamics in teams.