Abstract
Although bike-sharing has gained great popularity as a sustainable commuting mode around the world, it has been performed poorly regarding its low mode share in some countries, including Singapore. As a response, this study proposes a novel research framework that incorporates a mixed logit model and customized agent-based travel simulations to investigate the underlying reasons behind the unpopularity of bike-sharing. A mixed logit model is calibrated according to the responses from a travel survey to understand commuters’ perceptions towards bike-sharing in Singapore. The commuters’ perceptions obtained are then incorporated in tailored agent-based simulations to test the effects of promotion methods on the resulting mode share of bike-sharing as a commuting mode. The numerical experiment conducted in Singapore shows that, even with a large amount of subsidies on bike-sharing and timely relocation operations, commuters in Singapore are still very reluctant to take it as a daily commute mode.