Decision making in a transportation mode experiment

The rise of rapidly improving Artificial Intelligence technology has resulted in various governmental institutions and organisations’ heightened interest in the integration of Autonomous Vehicles (AVs) into the transportation network. As such, this paper seeks to provide insights on individuals’ dec...

全面介紹

Saved in:
書目詳細資料
Main Authors: Kwa, Wei Jun, Loong, Pin Sheng, Lua, Shu Chyi
其他作者: Jonathan Tan
格式: Final Year Project
語言:English
出版: Nanyang Technological University 2021
主題:
在線閱讀:https://hdl.handle.net/10356/147564
標簽: 添加標簽
沒有標簽, 成為第一個標記此記錄!
機構: Nanyang Technological University
語言: English
實物特徵
總結:The rise of rapidly improving Artificial Intelligence technology has resulted in various governmental institutions and organisations’ heightened interest in the integration of Autonomous Vehicles (AVs) into the transportation network. As such, this paper seeks to provide insights on individuals’ decision making of transportation mode choices by utilising a Driving Simulation (DS) experiment as a platform for decision-theoretic analysis, comparing between driving themselves (Self), delegating to other drivers (NAV), or AVs. Effects of individualistic characteristics such as gender, self-performance, driver license and private vehicle ownership, and behavioural biases like driver’s overconfidence, risk preference, and trust in technology on decision-making in a transport context were also analysed using Discrete Choice Model (DCM) to postulate probability and preference of transport choices. For this research, a wider variance to AV’s performance was specially allocated to replicate uncertainty of AV’s current technology. Effects of individual’s learning and experience on transportation mode choice were also studied using Adaptive Learning Models where short-term recency and long-term memorisation effects were validated at 1% significance level. Results indicated individuals’ reliance on recent mode choice and private benefits obtained, and memorisation of choices across a period. However, they lacked the abilities to remember the benefits received for each individual trip. Results from the DCM indicated previous trip value, trust, private vehicle ownership, and self-performance to be statistically significant, but there was a lack of evidence to suggest the same for driver license ownership, risk preference, and gender. The results from previous trip value and trust also provided a ranked preference of AV > NAV > Self while private vehicle ownership and self-performance only indicated Self > AV and NAV. These suggest over-reliance and biasness towards AVs and thus, provides policymakers with insights regarding factors and policies to consider for mass adoption of AVs.