Section: New Results

Use of Bayesian networks for situation awareness risks prediction

In all domains involving complex human systems interactions, such as the robotic domain, human errors may have dramatic impacts. These errors are often linked to situation awareness issues. We recently proposed a new method to predict situation awareness errors in training simulations [2]. It is based on Endsley’s model and the 8 “situation awareness demons” that she described. The predictions are determined thanks to a Bayesian network and Noisy-Or nodes. A maturity model is introduced to come up with the initialization problem. The NASA behavioral competency model is also used to take individual differences into account.