Decision Support System for Large-Scale Evacuation Logistics – DSS_Evac_Logistique

In this project, we consider large-scale evacuations of mid-size towns in situations, where evacuees have to change their center of life for a time lasting from several days to many months with the eventual goal to return to their respective homes. We consider the complete chain of events, including strategies to evaluate the necessity of evacuations, logistic questions of scheduling and transportation, the location of emergency units and shelters, and the provision of necessary supplies of various commodities to evacuees during their absence from home. Notice that we locate our project in the context of crisis management more than a context of risk prevention.

In the project, it will be investigated which of the industrial logistic methods can be carried over or adapted to the evacuation environment, and which evacuation situations require the development of new logistic methods. The joint work of the Franco-German consortium will be centered about and driven by the necessities of practice defined and controlled by the participating end-users. Noteworthy, they will provide data, rule the definition of crisis scenarios and validate the results outcomed by this project (evacuation plan, location of emergency facilities, ...).

The methodological approach is based on optimization and simulation with a specific emphasis on multicriteria optimization, in which several – in general conflicting – objectives have to be considered. This is typically the case in evacuation scenarios. We also insist on the need for introducing estimates of human behaviors in case of crisis notably to validate by simulation the potential feasibility of the calculated evacuation plans.
The goal of our research is a user-friendly decision support system (DSS) in which the decision maker can ultimately search in a data base of Pareto optimal planning concepts, i.e. a planning representing a good trade-off between the criteria used to evaluate the notion of a "good planning". It is important to emphasize the “supporting” character of the DSS, since the system provides evaluating parameters (e.g. cost, time, effect, etc.) associated to decisions and the trade-off between these parameters, but final decision have to be left to decision makers, i.e. human beings.

Such analysis of emergency situation and evacuation scenarios associated to these situations requires knowledge from several areas of scientific expertise: urban planning, disaster modeling, security, location, transportation, constrained path computation, simulation, etc. Therefore, one aim of this project is to provide a DSS which exploits the multidisciplinarity skills of the scientists and industrials who are participating and who are representative of these areas of expertise.