Interrelationships Between Urban Travel and Other Infrastructure

This dissertation advances the science of how transportation interacts with other urban systems by demonstrating how alternative data sources can reveal mobility patterns and human activity in near real time. Traditional transportation research depends on household surveys and administrative records, which are costly and slow to capture change. In response, this work examines social media and electricity consumption as proxy indicators of travel demand. First, it shows that sentiment derived from geo-tagged Twitter posts can be structured into covariates that significantly improve forecasts of telecommuting prevalence, particularly during disruptive events such as the COVID-19 pandemic. Second, it finds that residential electricity usage profiles in the Chicago region correlate with commuting characteristics, revealing spatially coherent patterns of mobility behavior. Third, it demonstrates that incorporating electricity usage data improves public transit demand forecasting using advanced machine learning models, especially under system shocks when historical data alone are inadequate. Together, these studies provide a framework for integrating unconventional and conventional data to support adaptive transportation planning. The dissertation also discusses the limitations of these proxy signals, such as demographic biases, data aggregation, and correlational relationships, and outlines future directions for causal inference, expanded data sources, and agent-based simulations to enhance scenario evaluation and policy testing.

Author dc:creator

• Juan G. Acosta Sequeda (23292016)

• Smart Cities and Urban Systems
• Travel Behavior Analysis
• Machine Learning Applications in Transportation

Statement dc:rights

• In Copyright
• Open Access after 2028-01-01

DOI dc:identifier

https://doi.org/10.25417/uic.31451644.v1

OAI identifier oai:identifier

oai:figshare.com:article/31451644