Observe urban environments
Build valid, transparent measures from maps, imagery, sensors, administrative records, and human experience.
Urban Science × Artificial Intelligence
Studying cities as dynamic human systems.
We develop spatial, computational, and AI methods to understand urban environments, and translate evidence into more equitable, resilient decisions.
Across the network
Research agenda
Methods grounded in urban questions.
Our work links urban theory, geospatial data, sensing, simulation, and responsible AI. The goal is not AI for its own sake, but stronger evidence about how cities change and whom those changes serve.
01Urban representation & multimodal learningHow cities are encoded from imagery, maps, mobility, and text.
02Human-centered sensing & perceptionHow people experience urban environments, measured at scale.
03Mobility, accessibility & inequalityWho can reach what, at what cost, and why it differs.
04Climate resilience & environmental justiceHow exposure to heat, air, and water risk is distributed.
05Urban agents, simulation & decision supportWhen computational agents can inform planning, and how to test them.
Approach
Measure, explain, translate.
Identify urban mechanisms
Combine computational models with careful research design to move beyond prediction toward explanation.
Support responsible action
Create open evidence and tools that researchers, communities, and public institutions can interrogate and use.
Platforms
Where the work is released.
The lab publishes through three open platforms, each with a single role. This site holds the agenda; the platforms hold the artifacts and the community.
URBANPROBE
Data, indicators, maps & methods
Open urban research infrastructure with documented provenance, validation, and reuse pathways.
XURBANAIModels, agents & benchmarks
Registries of urban AI releases, each documented with its task, evaluation, and failure modes.
CITY COMPUTING COLLECTIVECommunity & collaboration
An open community for finding collaborators, sharing data and tools, and working in public.
Work with us
Where to start depends on what you need.
Each route goes straight to the place where the conversation, the form, or the person is.
StudentsIntroduce yourself in the CollectiveSay what you study and which questions interest you. Introductions are read by the lab and by others in the community.Join the Collective →
Researchers and practitionersPropose a collaborationA structured post with the question, place, data, and the expertise you are looking for.Post a collaboration request ↗
Data and indicatorsContribute or request dataPropose a documented dataset, indicator, or method for UrbanProbe, or ask for data you need.Open UrbanProbe forms ↗
Models and benchmarksSubmit a model, agent, or resultPropose an entry or a leaderboard result for XUrbanAI, evaluated against the published standard.Open XUrbanAI forms ↗
Everything elseContact the lab directlyPress, partnerships, teaching, or anything that does not fit above.Contact →