Reliability of Public Transportation Systems with Autonomous First-/Last-Mile Services in a Stated Mode Choice Experiment

To achieve a modal shift away from private car use, public transportation must become a more attractive alternative, particularly with regard to reliability. While travel time and cost have been extensively studied as key determinants of mode choice, the impact of travel time reliability has received increasing attention only in recent years. However, quantitative surveys addressing this issue remain scarce, and most existing evidence is based on stated choice studies with highly heterogeneous results. One reason for this inconsistency is the complexity of representing reliability, as it can only be described through a distribution. Many forms of representation appear to overwhelm respondents. This study therefore seeks to bridge this gap by conducting a stated choice experiment with an improved visualization of reliability, presenting travel time and five equally likely delay times as bar charts and incorporating realistic choice situations that account for both trip purpose and transported luggage. The survey, conducted with a net sample of 859 respondents, reveals that travel time reliability is a decisive factor influencing mode choice. Delays in public transportation are particularly disfavored, with the mean delay rated 3.7 times more negatively than travel time itself. This valuation significantly exceeds the ratios applied in existing cost-benefit analyses, such as those used in German transport infrastructure planning. These findings suggest that the perceived importance of reliability is substantially higher than previously assumed. For on-demand services, the study indicates that displaying slightly longer expected travel times with early arrivals is preferable to promising shorter times that result in delays. These insights underscore the need for policymakers and transport planners to prioritize reliability improvements in public transportation planning and investments.