Decision-Making and Risk-Taking Behavior in Lunar Landing

Abstract

Model-based simulation and human subject experiments can be used to develop quantitative methods for analyzing the human-automation task allocation of a system early in the design process. An integral part of the human-system model is a representation of human decision-making and risk-taking behavior. These behaviors were investigated in a lunar landing human subject experiment. Subjects were asked to select a landing aim point that was near both a point of interest and hazardous region. It was expected that the placement of the landing aim point would vary with the probability of manual versus automatic flight and whether estimated touchdown dispersions were remembered by the subjects from earlier in the experiment or presented graphically on scatter plots. The experiment found that subjects did systematically modify the placement of landing aim points. Further, presenting landing dispersions graphically allowed subjects to compensate for touchdown deviations in both risk-critical and non-risk critical directions.