SafeWay: 2nd Place at the AI.WOMEN Hackathon

2nd place overall at the AI.WOMEN Hackathon in Hamburg, 12-13 September 2026, out of 300 participants. The team was invited to the AI.SUMMIT 2026 in November.

Built with Hannah Kalker, Janne Achenbach, Dominika Oliinyk and Megha Darda, SafeWay is a pedestrian routing app that finds the safest walk home, not just the fastest one. It scores a few route options against the traveller's own preferences (lighting, street type etc.), so each person gets a route personalised to them.

The technical challenge: Hamburg doesn't have an official open streetlight dataset. SafeWay pulls street data from OpenStreetMap via the Overpass API, classifies each segment as unlit or lit and by type (residential, main road etc.), and turns that into an avoid layer. This is fed into OpenRouteService's routing engine as avoid-polygons, which generates the route alternatives live and scores them against the user's preferences.

The QR code at the end of the demo video only works within Hamburg's coverage area, as it was built for the city we hacked in, in the time we had.