Long-term is no term: A systematic review of learning effects and the understanding of “long-term” in the context of driver-vehicle interaction

Abstract

The appropriate usage of automated driving systems must be learned, with the learning process supposedly tied to the continuous engagement of the driver with the system. Despite extensive research on the short-term effects of automated driving systems, there is limited understanding of the long-term usage and the adaptation processes involved. This work addresses this gap through a systematic literature review focused on the effects of repeated or prolonged exposure to automated driving systems. In particular, the review aims to clarify definitions and theoretical constructs related to the long-term usage of driving systems, to provide recommendations for the methodological design of studies investigating learning and adaptation processes, and to synthesize existing findings on the effects of repeated exposure. A comprehensive literature research adhering to the PRISMA guidelines resulted in the review of 96 articles. The review emphasizes that other influencing factors, such as the sequence and quality of experienced events, play a crucial role in driver adaptation rather than the mere duration of exposure. Finally, a conceptual model of driver adaptation is developed, distinguishing between learning processes and temporary state adjustments and serving as a basis for exploring and comprehending the mechanisms and dynamics of such processes. This model highlights the need for future studies to adopt a more nuanced approach to “long-term” exposure, considering intermission periods and the influence of system capabilities, limits, and failures on the development of mental models. The paper concludes by offering recommendations for future research, stressing the importance of both behavioral and attitudinal adaptation.