Walk-Free

Walk-Free shows college students how hot every street on campus will feel at the hour they walk it, and finds the shadiest way there. A problem we know firsthand: we’re students, and every summer we walk between classes in the worst heat of the day, arriving drained, sweaty, and late, because every map only knows the fastest way: straight through the sun. That isn’t just uncomfortable. Heat now kills about 546,000 people a year worldwide, roughly one every minute, and it follows students to class: hotter school years measurably cut how much they learn (Goodman et al., American Economic Journal, 2020). Shade-routing apps already exist, but they answer one question: which way is shadier. We built Walk-Free around a student’s whole day. Instead of checking the weather, guessing which side of the street has shade and planning around the heat, you just ask out loud, in any of 25 languages. “What’s the shadiest way back to my dorm right now?” Walk-Free plans the route on a live 3D map of campus and reads the answer back to you. It knows campus building codes like PCL and GDC, cuts through air-conditioned libraries and student unions, tells you the coolest time to leave for class, switches to well-lit paths after a late night at the library, and shows how much hotter the sun will feel than the shade, using live weather, humidity, air quality, and heat alerts. On UT Austin’s campus (our campus), the shade route keeps a student out of full sun for about 9 hours over a semester. It brings Meta’s models into our world: Muse Voice Transcribe turns a spoken question into text, and Muse Spark acts on the map rather than just chatting, calling Walk-Free’s own tools to plan routes, rank the hottest streets and plant trees, then answering in the student’s language. Meta’s DINOv3 canopy map finds the trees nobody mapped: 62,748 tree crowns on our six campuses, against 4,375 in OpenStreetMap. Muse Wonder reasons through harder questions, like planning a whole day of walks. It makes something possible that wasn’t before: a student can check the map against reality by snapping a photo of a street, and SAM 3.1 picks out the trees, pavement, and sky to compare with what we predicted. Then they can do something about a hot street. Walk-Free finds where new trees would help the most students, plants them in 3D, uses Muse Image to show the greener street as a realistic photo, and writes a proposal to the campus or city with a link to see it planted. A sweaty walk to class becomes a case for change. Six campuses and Midtown Manhattan open instantly, and any US city can be built from OpenStreetMap. Try it with no sign-up at https://walk-free.vercel.app/demo.