TDM Tidbits: How to leverage AI in TDM
AI is quickly reshaping how TDM professionals understand and influence travel behavior. From anticipating shifting commute patterns to refining program design with greater precision, AI offers new tools to strengthen impact and decision-making. At the same time, thoughtful implementation is required to address challenges around data use, equity, and transparency. In this TDM Tidbit, members share their insights on how AI can elevate the field while being applied in responsible and meaningful ways.
Hear from Jossimar Fuentes, TDM-CP (Alta Planning + Design), Peter Williamson, TDM-CP (Santa Barbara County Association of Governments), and Eliza Yu-Dietz, TDM-CP (MTA Shared Mobility (Operations)) in our April TDM Tidbit!
Jossimar Fuentes, TDM-CP
The near-term value of AI in TDM is not that it gives us a crystal ball. It is that it can help us see patterns earlier and in ways that are actionable. For prediction, AI can combine travel survey data, transit performance, weather, events, school schedules, parking conditions, and historical network behavior to identify when, where, and for whom travel choices are most likely to shift. In practice, that means better detection of recurring patterns. Effective TDM is about intervening at the right moment and not just telling people what options exist but reaching them at the decision stage. AI can help identify those moments.
Peter Williamson, TDM-CP
The Santa Barbara, California region has an AI Bike Map project that uses machine learning to categorize every street and road by levels of comfort (High, Medium, Low, Non-Conforming). Using Replica data to measure average speeds and vehicle volumes, this project uses AI to juggle user preference, recent traffic trends, and infrastructure tags in OpenStreetMaps to do what took weeks in a matter of days.
Eliza Yu-Dietz, TDM-CP
AI can help predict and respond to travel behavior patterns through information collected from map apps, traffic data and cars that have the ability to sense its surroundings and communicate with other vehicles. This information can help make traveling more efficient where if a traffic jam or construction is along your route, AI can reroute you to avoid sitting in traffic and maximize your travel time.
Jossimar Fuentes, TDM-CP
On design, AI can help TDM teams move faster from broad assumptions to granular segmentation. We can identify distinct groups based on barriers and likely triggers. That segmentation fits what we see in the real world, where different barriers require different interventions. Structural barriers and behavioral barriers are not the same things, and we must understand the difference before asking AI to do it for us. AI can help distinguish outputs from outcomes. Programs may have high engagement but low sustained mode shift. Another may have good participation but strong repeat behavior among users with higher potential for VMT reduction. TDM evaluation is better when it can measure adoption and retention.
Peter Williamson, TDM-CP
For users, AI can demystify "alternative transportation" as what's best for the user's needs. While platforms are catching up, there's a practical ask Gemini or CoPilot what your best options are hack in the meantime. On the backend, TDM practitioners can demystify open feedback about barriers to adoption. I've had AI summarize thousands of open-ended comments that I don't have time to read.
Eliza Yu-Dietz, TDM-CP
AI-powered tools can improve the efficiency and accuracy of TDM program design and evaluation through various data collection that can help inform where people are traveling to and from the most, what the issues tend to be (ie. a transit desert or crash data), which can then be used to make informed decisions around redesigning a road with safety measures or making the case for a new bus route, etc.
Jossimar Fuentes, TDM-CP
Like any new technology, there are questions about ethical uses. AI supplies information at breakneck speed and is reviewing points of view and biases of the loudest groups or individuals. That underlying base knowledge may underrepresent people, groups, and ideas. We risk losing the very thing that makes TDM professionals good at what they do, putting ourselves in the journey of the user. We learn so much from empathy and connecting with people based on our knowledge of their current behavior. ACT has an opportunity to take a lead role and discuss AI in our work, specifically, holding ourselves to a higher standard for our use of LLMs.
Peter Williamson, TDM-CP
I've heard concerns about environmental impact and privacy. If you can use AI to save yourself months of personnel time, your efficiency could outweigh the negatives. If one person switches from driving alone for a day, you earn yourself a guilt-prompt!
Eliza Yu-Dietz, TDM-CP
Some potential risks and challenges of applying AI in TDM is the monetization of this information and how private companies might use this to leverage business or drive up pricing for consumers. Some ways to address this responsibly is by making sure contracts and agreements stay up to date, allow for data to be publicly accessible and ensure this information would not become cost prohibitive to public agencies.
Jossimar Fuentes, TDM-CP
For equity, AI can potentially start seeing where the system breaks down when we are just planning for an average commuter. We cannot know every barrier someone faces while making the decision on their travel, but AI can be used to expand our knowledge of the demographics. We can drill down to a small detail that prevents someone from a behavior change moment. We must then in real practice find out if those assumptions are true. It brings us back to the things we do best and there is no replacement for genuine curiosity and desire to support communities through their real battles and barriers.
Peter Williamson, TDM-CP
I think AI allows the everyday person to become impactful advocates. A resident can prompt AI to explain their jurisdiction's circulation element like they are 5 years old, and then have AI draft a planner-speak response.
Eliza Yu-Dietz, TDM-CP
AI-driven insights could open new opportunities for advancing equity, accessibility, and sustainability by finding and understanding where there’s a need and for what populations, such as a specialized facility or a stadium shuttles people back and forth predictably between the same places on a recurring basis and there’s a shortage of shuttles. AI-driven insights could fill this gap and can offer solutions such as offering employer/business discounts for a van rideshare program or provide pop up micro-mobility services to reduce traffic congestion when there is limited thoroughfares at a busy, popular attraction.
Interested in contributing to a future article? Two upcoming topics include The role of TDM in public health and Entrepreneurship in TDM. Visit the form below to share your responses and be featured in an upcoming TDM Tidbit!