The V Team

OpenStreetLightMapper is an open-source, AI-assisted tool that turns street-level imagery into an evidence-backed public-lighting inventory. Users select a geographic area, explore Mapillary image sequences, detect poles and luminaires, review observations, merge repeated sightings, and save their inventory in a local workspace. The project addresses a practical challenge from my work designing LED roadway-lighting optics and installations: understanding existing lighting arrangements often requires collecting fragmented information from individual municipalities. My longer-term goal is to characterize regional patterns—such as pole spacing, mounting height, and arrangement—to support more efficient optical and lighting design. Built during one available day of the hackathon, the prototype demonstrates a working detection, review, and persistence workflow. Automatic geolocation and reliable dimensional measurements remain areas for further development.