Choice modeling for policy decision support towards attractive, reliable, and efficient public transport systems
For urban areas to remain or become attractive and liveable areas, public transport needs to constantly ensure that it caters for users’ changing needs and travel patterns and provide them at the same time with an attractive, reliable, and efficient service. The motivation for this project is to develop a policy decision support tool that can help authorities, policymakers, planners, and other involved actors investigate the effects of their decisions on public transport users and service performance. Examples of such decisions could be related to fare changes, ticket pricing, as well as off-peak discounts. However, potential applications of the tool go beyond this and include disruption management or studying the effects of changes in supply. Shortly, the contribution of this project is to enable the evaluation of policies in early stages of the decision-making process and difficult-to-evaluate effects on public transport users and usage.
The covid-19 pandemic has had various impacts on how people use public transport. Traveling during off-peak periods is becoming more practical and realistic than prior to the pandemic, as people’s flexibility in travel patterns and commuting has
increased. Encouraging more travelers to switch to travel during the off-peak periods can help reduce crowding, improve on-time performance and thus save public resources and increase level of service and thus public transport attractivity. It is also expected that seasonal tickets (e.g., monthly subscriptions) will become less attractive, making it necessary to design new fare products that better cater for the prevailing travel patterns and passenger preferences in the post-corona phase. We identify the need for easily adaptable policy decision support at the early stages of planning strategies, focusing on the attractivity and effectiveness of public transport in the changing world. We also expect that the framework proposed in this project will be adaptable to changing needs in the future.
The policy decision support tool that considers choices and socio-economic attributes needs to address three aspects of public transport: supply, demand, and travelers’ choices. With this project, the main focus is on passenger choice modeling and generation of a representative synthetic population enriched by socio-economic background. In order to compose a synthetic population and various travelers’ choices, we use a large data set of anonymized smart card data travel diaries. The usual practice is to use travel surveys with a small and often specific group of respondents, which may not accurately represent the entire population of travelers. In contrast, in this project we plan to explore a modern and innovative data-driven approach for choice modeling that considers empirical observations that include almost all the public transport users with imputed socio-economic and activity attributes.
This is the project led by KTH, and collaborated with VTI and SL in Sweden.