| Uncertainty in Situation Awareness – Dempster-Shafer |
|
Organization responsible: TRT-FR People involved:
Project description: Decision making is the process of sufficiently reducing uncertainty and doubt about alternatives to allow a reasonable choice to be made from among them”. This definition stresses the information gathering function of decision making. It should be noted here that uncertainty is reduced rather than eliminated. Very few decisions are made with absolute certainty because complete knowledge about all the alternatives is seldom possible. Moreover, in the context of emergency, decisions are usually taken place under uncertain and complex situations. Such situations are characterized by three elements: threat, urgency, and uncertainty. Then, an agent (decision-maker, stakeholder, decision support system…) risks losing something of great value. Most agents are faced with difficult decisions. It is a simple fact that decision represents trade-offs between all dimensions of risks which are difficult to be observed and evaluated. Moreover, agents must take action quickly because without intervention, this undesirable situation will be realized in a limited amount of time. Different actions will generally incur different costs and have different outcomes under different circumstances. The agent's decision-making process is further complicated by various sources of uncertainty (noisy information, exogenous events, actions with non-deterministic outcomes, etc). To make a decision understandable, a successful situation understanding support system is one that assists the human decision maker in order to enhance the performance of decision makers in high density and/or ambiguous situations. Furthermore, a methodic way of building a current situation analysis picture is required, otherwise the consequence would be a chaotic and ineffective management of the emergency situation. The major challenge is uncertainty, that is a property resulting from a lack of information about the world for deciding if the statement is true or false; it is a property of the relation between the information and the agent’s knowledge about the world. In other words, uncertain information occurs when the concerned agents may interpret the available data differently. In that context, situation understanding picture needs some techniques for combining and aggregating such uncertain information in order to establish a consistent state of global knowledge. The models proposed to represent quantified uncertainty are usually based on probability, possibility or belief functions. The project objective is to transform the raw, imprecise, conflicting and often paradoxical information received from the different sources into statements understandable by both human and machines. The ability to adapt efficiently to changing environmental conditions, learning how to anticipate possible changes in its environment is inescapable. A good understanding of the situation implies associated risk forecasting. The main goal is to define an information perception and comprehension model to allow the user to browse in the Situation Understanding Module (SUM) through a risk assessment prism. Publications: |