The map said there was nowhere to sit in Kolkata, so I asked Gemma to look
This is a submission for the Hacktoberfest Open-Source AI Challenge: Touch Grass . What I Built My father has knee pain. He likes walking, but he stops going out when he can't be sure there is somewhere to sit before his knee gives up. So the question for every walk is not "how far" but "where can I rest". Bench to Bench plans a short walk around wherever you are, where a place to sit is never…
This is a story about an app called "Bench to Bench" that helps people with knee pain find places to sit during walks. The app plans a short loop with a place to sit at least 150 meters away from the previous seat, without stairs, gravel, or rough ground. In Mainz, Germany, the app can find 431 designated benches using OpenStreetMap data, planning loops up to 1,294 meters long.
However, in Kolkata, India, OpenStreetMap knows of no places to sit, so the app relies on Gemma, an AI model, to find real benches from street photos.
Gemma analyzes street photos from Mapillary and describes any potential seating areas, such as low ledges, platforms, or chairs outside shops. The app then checks these suggestions against a fixed format and counts any that fail the criteria. In the tests conducted, Gemma found 7 places to sit in 165 photos, compared to the map's inability to provide any suggestions. The app successfully planned a 254-meter loop in Kolkata, ensuring the user never had to walk more than 69 meters from a seating option.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
