Suzanne Childress
20 years translating transportation data into decisions. Translating messy transportation data into analysis that planning agencies can actually use: working collaboratively to drill down to the essentials, rigorous, interpretable, and connected to the decisions being made. Available for consulting, research collaboration, and peer review.
Household travel survey design and successful application in planning, equity analytics, discrete choice model estimation, forecast evaluation, uncertainty quantification, and travel demand modeling (activity-based and trip-based). Two decades estimating, validating, and explaining models whose outputs directly shaped transit investment, long-range plans, and state legislation.
Data warehouse design, reproducible analytical workflows, and Python/R/SQL pipelines that give agencies lasting analytical independence. Experienced transitioning consultant-built systems into documented, maintainable in-house products and building open-source tools that make regional data accessible to planners.
Program design and management for large-scale research initiatives — from sampling strategy through public data dissemination — including leading multi-year, multi-million-dollar survey programs and building analytical teams. New technique development, AV/AI evaluation methodology, and translating findings into investment recommendations and policy-facing documents. Available for advisory, peer review, and strategic leadership roles.
What it's like to live long enough to revisit your own assumptions. Looking back at AV scenario work from 2014 — what aged well, what didn't, and why returning to old forecasts is one of the ways judgment actually gets built.
So much of planning and modeling work happens inside contradictions. The challenge isn't to eliminate complexity — it's to be honest about it, and keep showing up with curiosity and humility anyway.
C. Thi Nguyên's book on gamification made me think about how planning metrics quietly reshape the world — how congestion relief, VMT reduction, and mode share each capture something real while missing something essential.