Modeling Long-distance Intercity Passenger Travel in the Global Multi-modal System

点击翻译按钮获取中文摘要

COVID-19 has highlighted the critical role of out-of-town travel in our lives and our economy. Despite its importance, modeling long-distance travel has received limited attention by transportation engineers whose efforts have instead focused on daily, more routine, local trips often to address the need for capacity expansion. Given that long-distance travel by plane, train, bus, and automobile has been estimated to account for 30-45% of miles traveled, this gap in strategic engineering tools compromises our ability to optimize access as well as to minimize greenhouse gas emissions, congestion and system-wide costs. The overarching goal of this program is to develop the data methods needed to facilitate accurate long-distance travel demand forecasting models. Our unique tactical strategy will depart from the data, concepts and approaches appropriate for use in daily travel models. We will embrace alternative ways of using a) semi-passive mobile device traces, b) spatial databases, and c) network accessibility measures that integrate surface modes with air travel. Simulating annual overnight travel on a nation-wide scale with global destinations cannot be agent-based, the state-of-the art for local travel modeling. Instead, quasi-disaggregate flows defined by new measures of accessibility across large spatial scales and networks will be used. Specifically, we will: 1.Design and validate a hybrid method for "Big Travel Data" that includes the person- and trip-based attributes essential for demand modeling. 2.Create data-driven conceptual frameworks of long-distance travel and indicators of accessibility and social network distribution to inform model training. 3.Develop bundled destination/mode choice models utilizing novel spatial measures. The methods developed in this program will allow alternative systems of long-distance travel, comprised of electric vehicles, automated vehicles, and new mobility services, to be assessed for their environmental impacts, social equity and costs. The Canadian transportation system is too important to evolve unchecked, and this program will facilitate better measurement of the current system, to better design the future system. Training graduate students in the context of transportation systems is ripe with advantage. First, large complex relational databases require programming skills with attention to accuracy, spatial correlation and privacy/ethics concerns. Second, data collection requires mixed methods from mobile device tracking, to internet-based data and surveys/interviews. Third, the transdisciplinary subject matter is attractive to recruiting diverse people and disciplines to engineering. Fourth, inequity in travel opportunity requires explicit coverage of inclusion in models. Finally, communication skills can be honed so that the next generation of engineers can build the collaborative partnerships needed in both public and private sectors to effect more sustainable transportation systems.