Off-road fitness-to-drive assessment using driving simulator (Effect of experience on fitness-to-drive)

• Main Author: Chan, Peggy Myat Kay Khine
• Other Authors: Wong Yiik Diew
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering::Civil engineering::Transportation
• Online Access: http://hdl.handle.net/10356/68036
• Institution: Nanyang Technological University

Factors that have been found to correlate with driving crash risk are driving experience and fatigue. The purpose of this project was to investigate how driving experience affects the fitness-to-drive performance of a driver in a driving simulator. A driving simulator was used in this experiment for its advantages over on-road experiments such as less risk involved to life and resources involved. A total of 43 participants were categorised into three groups of subjects: active drivers, inactive drivers and no licence subjects. They completed a test scenario with five different hazard scenarios first under non-fatigue (fresh) conditions and then repeated under fatigue (non-fresh) conditions. Four different indicators were used to analyse the driving performance, namely, average speed (km/h), offset of steering wheel (measured in terms of average variance), average number of collisions and lastly, average number of navigational message mistakes made over the course of the whole driving route. Numerous hypotheses were also made based on the performance indicators and these hypotheses helpe to create and support the two main hypotheses made for this project. The first hypothesis is that active drivers (with most experience) perform the best and no licence subjects (with least experience) give the worst performance. The second hypothesis is that inactive drivers will give a performance somewhere between that of the active drivers and no licence subjects. However, the exact position in the performance level of the inactive drivers is unknown between the two wide spectrums and it is worthwhile to find out which spectrum this group of drivers is closer to. Using various statistical tools such as t-test and 95% confidence interval graphs, the findings suggest that active drivers give the best driving performance and the no licence subjects give the worst performance, as expected. However, an important finding is that inactive drivers perform at almost the same level as no licence subjects. These findings suggest that if exposed to actual on-road driving which is riskier than simulated driving, inactive drivers would be equally prone to crash risk as no licence subjects with no driving or training experience. Measures need to be taken to specifically address inactive drivers who wish to return to active driving. Some recommendations are also provided in this study.