Behavioral insights on individual and social adaptation to automated mobility using a serious game

Introduction: Serious games generate a virtual environment where players are immersed in simulated conflicts, guided by distinct rules. They replicate intricate systems, like transportation, encouraging social engagement via reinforced learning. Unlike preference-based studies, these games offer enhanced real-time feedback on players' actions. Thus, they reveal how user experience and social interaction influence decision-making over time. We use a serious game to study the willingness of travelers to adopt automated mobility, specifically shared modes of transport, an important step toward alleviating congestion, enhancing the quality of urban living, and improving people's health and well-being. Methods: For each scenario, 100 participants were randomly divided into ten groups of ten interacting players. They chose independently out of three automated transportation modes- shared ride, shared car, and automated transit-over 50 simulated days how to commute to work. They aimed to maximize their overall score by arriving punctually, which was influenced by their mode and departure time and the choices of fellow players. Cross-nested logit kernel choice models were estimated based on the game data. Results: In the recurring congestion scenario players learned to adopt the shared ride at the expense of transit; in the nonrecurring congestion scenario, random incidents increased the use of transit and shared car (ride alone). Conclusions: Congested traffic motivated a shift to ridesharing at the expense of private rides and public transport; however, the latter was highly demanded when traffic became unsmooth and travel times more uncertain. The implications can be translated to health promoting polices to encourage sustainable travel behaviors while also improving transportation efficiency.