Offering shared mobility options at transit stops can potentially increase the service area of a stop and consequently, possible detours in transit lines can be eliminated to decrease in-vehicle travel times for through-passengers and reduce operational costs. However, current research mostly focusses on shared mobility options and expected behaviour only, whilst not looking at this integrated transit network design problem. Additionally, most focus of current studies is on the integration of shared mobility in urban areas and/or around train stations, leaving a gap on suburban areas and transit lines with lower demand. In order to answer our research question “what the effects are of increased route directness for low-demand transit lines in conjunction with offering shared mobility at transit stops”, we developed a mesoscopic model extension for the aggregated four-step transport model to model changes in travel behaviour as a result of straightened transit lines and the simultaneous integration of shared modes. Discrete choice models are used to accurately model first and last mile preferences of people, based on the access and egress distance, demographics and available (shared) modes. Finally, the probability of passengers cancelling their complete trip as a result of increased first and last mile distances is also explored. This model framework was applied to nine case studies in the Netherlands. The synthesis of the case studies resulted in key factors contributing to a promising redesign of the transit network. The main factor is that through-passengers should significantly outnumber local passengers, by at least 75%-25%. Additionally, the increase in access and egress times should not be significantly larger than in-vehicle time savings of through-passengers. Moreover, it is found that the mode share of micromobility in the first and last mile is approximately 15% across the different cases, whereby the highest usage can be seen for people under the age of 25 and for distances greater than 1 km. Finally, it is concluded that the additional costs of shared mobility are on average only 10% of the savings in operational costs.

Shared transport modes can potentially contribute to first and last mile connections of public transport (PT) trips but this remains quite underexplored in the literature. Our study explores the user preferences for shared modes as first and last mile option to connect activity locations. We have focussed on local public transport in the Utrecht province, The Netherlands, which includes bus and tram lines. Its diversity in land use and PT network density, the overall high bicycle usage, as well as the increased proliferation of shared mobility concepts yield promising information which can be a harbinger for future PT integration worldwide. For both the urban and suburban areas of the province, we have designed and conducted a stated choice experiment. Respondents were able to choose from shared bicycles, e-bikes, e-scooters, and e-mopeds to reach their urban destination from a PT stop. For suburban destinations, we also included light-electric vehicles (LEVs), e-cars, and demand-responsive taxi services. Such a complete list of possibilities to travel by shared modes allows comparing the different options and producing trade-offs not available yet in the literature. A sample of 499 respondents (285 urban and 214 suburban PT travellers) considered their first and last mile mode choice of a recent PT trip in light of the new options. Results show that shared (electric-)bicycles and e-scooters are generally preferred over other shared mobility options. The latter specifically targets younger people (<26 years) and travellers towards suburban destinations. Still, a majority of PT users prefers not to use shared modes in the first and last mile. We found that age, current cycling behaviour and weekday/weekend travelling are the most important factors which determine these preferences. We argue that shared bicycles and e-bikes are the most capable modes in providing benefits to PT travellers in this context and, given the relatively low travel time sensitivity, can best be distributed around the most important PT stops.