Boosting Winter Green Travel: Prioritizing Built Environment Enhancements for Shared Bike Users Accessing Public Transit in the First/Last Mile Using Machine Learning and Grounded Theory

Abstract

Shared bikes are widely used in Chinese cities as a green and healthy solution to address the First/Last Mile issue in public transit access. However, usage declines in cold regions during winter due to harsh weather conditions. While climate factors cannot be changed, enhancing the built environment can promote green travel even in winter. This study uses data from Shenyang, China, to investigate how built environment attributes impact the travel satisfaction of shared bike users who utilize bikes as a First/Last Mile solution to access public transit in winter cities. By employing machine learning algorithms combined with Asymmetric Impact-Performance Analysis (AIPA) and grounded theory, we systematically identify the key attributes and rank them based on their asymmetric impact and urgency of improvement. The analysis revealed 19 key attributes, 17 of which are related to the built environment, underscoring its profound influence on travel satisfaction. Notably, factors such as the profile design of cycling paths and safety facilities along routes were identified as high priorities for improvement due to their significant potential to enhance satisfaction. Meanwhile, features like barrier-free access along paths and street greenery offer substantial opportunities for improvement with more modest efforts. Our research provides critical insights into the nuanced relationship between built environment features and travel satisfaction for First/Last Mile shared bike users. By highlighting priority improvements, we offer urban planners and policymakers a framework for creating livable, sustainable environments that support green travel even in harsh winter conditions.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 3 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.