Modelling a Fighter Pilot’s Intuition in Decision Making on the Basis of Damasio’s Somatic Marker Hypothesis

摘要

Damasio's Somatic Marker Hypothesis” (Damasio, 1994) provides a neuro-cognitive account on how humans manage to solve complex decision problems in a relatively efficient way. Damasio proposes that specific centers in the brain, which relate to feeling certain body states resulting from emotional responses on a situation, generate so-called ‘Somatic Markers' that help to make a selection of viable options for action. This process involuntarily guides the experts' decision and provides him with the sense that a complex decision was taken care of by gut feeling, or rather intuition. Fighter pilots very much depend on the type of aggregated experiences that constitute their intuition. It enables them to stay separated from their enemies and attack them from the right angle, at the same time coordinating with their wingmen in the formation, and all of this in a few seconds. Their considerations of the tactical situation, the possible maneuvers and actions and the desired outcome of the situation do not proceed rationally. These decisions and evaluations are generated subconsciously and intuitively. Fighter pilots obtain this competency gradually through training, in a 'live' or a simulated environment. In the latter type of environment, the opponent aircraft and their operators can be represented by Computer Generated Forces (CGFs). In so far these CGFs adapt themselves to the tactical environment, they can be considered virtual intelligent agents. The idea is that such agents behave human-like. They observe the environment, orientate themselves in this environment, decide on what to do next and perform actions in a way that humans would do, including opportunistic, erroneous, untraceable and sometimes surprisingly brilliant behavior that is so characteristically human. To discover the essence of realistic role behavior, expert knowledge of fighter pilots is used. In addition, available knowledge from Cognitive Science is translated into specific models that underlie these virtual agents. Such models may include elements of intuitiveness, creativity, opportunism, autonomy, and situation awareness, but also elements of exhaustion, task saturation, and psychological and physical stress. Present models of decision making are predominantly based on the assumption that the basis for decisions is merely rational. Kahneman and Tversky (1976) already pointed out that even for fairly simple decisions humans do not operate on a purely rational basis. Hence, a model on the basis of Damasio's Somatic Marker Hypothesis has been designed to account for the role of intuitiveness, emotions and feelings in fighter pilots' decision making. This computational model has been modelled in the declarative hybrid modelling language LEADSTO which integrates temporal logical and numerical capabilities. The model focuses on key aspects of the Somatic Marker Hypothesis. For example, a central feature of the model is Damasio's claim that somatic marking helps to reduce the number of options in order to make a cost-benefit analysis more manageable. The validity of the model is demonstrated by means of simple tactical situations taken from the air-to-air fighter domain that are used as scenarios for the LEADSTO software environment in which simulations with the model have been performed.