Modeling driver behaviour with different degrees of automation : a hierarchical decision framework of interacting mental models

An integrated driver model is presented in which cognitive constructs such as driver needs are causally connected to the orchesfration of skill based driving tasks. The proposed three layer hierarchical structure is composed of satisficing decision makers who communicate with intermediate layers of dynamic mental models. The decision makers direct the information flow and decide which mental model is consulted and/or activated and when. Mental models on the other hand provide information at different levels of abstraction that guides the decision making process. With the aid of this hierarchical driver model, predictions can be made about how automation of particular driving subtask may influence the overall driving behavior and to what degree these may be driver dependent. We particularly focus on the effects of introducing an Adaptive Cruise Control (ACC) system by exploring the dynamics of the driver's ACC mental model as a function of experience with the ACC. K e y w o r d s : Driver model, mental models, adaptive cruise control, satisficing decision making; humanmachine interaction.