Are We Planning Tomorrow with Yesterday's Travel? Cynthia Chen and Ram Pendyala
Cynthia Chen and Ram Pendyala: are our models and data are up to the task?
People are making fewer trips. In 2001, Americans made 4.1 trips per person per day; in 2017, 3.4 (18% fewer). In England, people made 1,086 trips per person per year in the mid-1990s; in 2024, 922 (15% fewer). A growing share of people do not leave home at all on a given day. Yet our streets are not emptier: meals, groceries and packages now come to us, carried by someone who drives for a living.
Picture one delivery driver making 30 stops for 30 households. Phone data sees one busy driver. A travel survey sees 30 households that made no trip. Which one is measuring demand? So is travel demand falling, or is it moving to other people, other vehicles and other purposes?
The answer matters for every transportation plan. Transportation is undergoing profound changes that challenge longstanding assumptions about how, why, and how much people travel.
The gig economy, delivery services, and ridehailing are blurring the boundaries between personal travel, commercial activity, and driving for a living
Automation, autonomy, and artificial intelligence (A3) could further transform activity participation, time use, and mobility choices.
These shifts are unfolding alongside evolving public preferences, perceptions, and priorities (P3) for transportation investment.
But our models were built on household travel surveys and on yesterday's travel patterns. Meanwhile, location data from mobile phones has given agencies more data than ever. But more data is not the same as a clearer picture. The people whose travel is changing fastest (older adults, low-income and zero-vehicle households) may be the people phone data sees least well.
At this Friday's Mobility Forum, I will host two leading scholars of travel behavior:
Ram Pendyala, Professor and Director of the School of Sustainable Engineering and the Built Environment at Arizona State University, where he also directs TOMNET, a U.S. DOT University Transportation Center on travel behavior and forecasting.
Cynthia Chen, Professor of Civil & Environmental Engineering and of Industrial & Systems Engineering at the University of Washington, where she directs THINKLAB.
We will explore:
What is changing in how, why and how much people travel, aging, new household forms, falling trip rates, working from home, outsourcing errands, the gig economy, and automation and AI, and can models built around yesterday's patterns guide tomorrow's infrastructure decisions?
What location-based service (LBS) data can and cannot tell a planner: how it is generated differently from a travel survey, which people, places and hours it under-samples, and how it can strengthen household travel surveys rather than replace them.
Where should agencies invest next: better surveys, new data products, more behaviorally realistic models, or shared data and modeling infrastructure? And where can AI help?
This session builds on our Forum with Michel Bierlaire on next-generation travel demand models.
I look forward to seeing you there!
– Jinhua
Professor of Cities and Transportation, MIT
Host, Mobility Forum