Unraveling the sustainability of mode choice A neighborhood sustainable mobility index for the Netherlands

This research develops a Neighborhood Sustainable Mobility index (NSMI) for the
Netherlands. It does so by using the Dutch mobility survey, ODiN. This mobility dataset is
enriched with open data to include neighborhood characteristics. These neighborhood
characteristics are gathered by using available datasets and performing GIS analysis. A random
forest classification model is used as a tool for the development of a mode choice prediction
model. Via the creation of a synthetic population per neighborhood, the random forest
classification model can be used to generate a modal split per neighborhood for short,
medium, and long-distance trips. Multiplying the modal split prediction of the different trip
lengths with the pollution score of the respective transportation mode allows the creation of
an NSMI score. This research has found that the built environment characteristics that were
included in this research do not influence mode choice behavior. Personal characteristics and
trip characteristics are found to be the main determinants of mode choice behavior.
Furthermore, the NSMI has indicated few differences in sustainable mobility scores for
neighborhoods in the Netherlands because no influence of built-environment characteristics
on mode choice was found.