Experience-based analysis and modeling for cognitive vehicle data

Abstract

Herein, we discuss an experience-based approach to cognitive vehicle research. We prepared six subjects with individual experiences and skills. We made them drive their cars on the same test course and collected data on their mental status and operations, as driving data, using a simple brain wave sensor, and the acceleration sensor in a smartphone. We analyzed the data and constructed models using our previously developed qualitative cognitive analysis and modeling tool. We then experimentally compared the variation between the analysis and the modeling results. Consequently, we could find a parameter that indicates the degree of experience and skill of each driver. We could also derive rules from models, which explained the experience and skill of each driver. Our experience-based approach yields an understanding of unconscious operation and situated cognition based on a driver's experiences, and how it allows novice or experienced drivers to retain and re-apply driving experiences and skills.