Senior Thesis: A study on trust and perceived risk in automated vehicles
As automated driving becoming more prevalent, it is important to understand drivers’ attitudes towards automated driving, which is a hybrid of complex, dynamic, and safety-critical situations. In order to investigate the trust formation in a simulated Level 3 automated driving task, two experimental studies were conducted about how people assess risk in different driving environments and the preliminary information about automation reliability. For Study 1, nine different driving scenarios were designed, which differed by driving speed, the traffic mode, and car behaviors. A methodology for categorizing risk in the automated driving context was validated. Based on results from the first study, a 2×2 repeated measures mixed design was conducted. The second study evaluated perceived risk for driving and prior information about automation reliability to test differences in drivers’ attitudes. Results showed that participants reported the highest level of trust when presented with high-reliability information and driving in a low situational risk drive. An interaction effect of reliability information and perceived risk of drives was found for drivers’ perceived automation reliability. This indicates that prior information and situational risk influence drivers’ trust levels towards automation and relationships with automated systems. Based on my findings, driver training materials and in-vehicle interfaces should be designed that encourage appropriate trust in automated driving. This could increase the adoption rate and enhance driving safety.
Keywords: trust in automation, acceptance, perceived risk, automated driving, car-driver interaction.
Please contact me if you are interested in reading my full paper.