Organised Editing/Activities/footpath.ai
About footpath.ai (footpath.ai)
footpath.ai specializes in pedestrian infrastructure mapping, leveraging advanced AI and computer vision technologies to create detailed maps of urban pedestrian networks including sidewalks, crosswalks, trees, benches, bike racks, and more. footpath.ai is a spinout of the University of New South Wales (UNSW) based in Sydney, Australia.
OSM Contributions
The footpath.ai team continually contributes to the OpenStreetMap (OSM), elevating the quality and quantity of the pedestrian infrastructure data and walkability-related features that can be used for navigation and wayfinding for humans and robots. The footpath.ai team is dedicated to improving OSM by working on:
- Adding and editing pedestrian infrastructure data (sidewalks, footways, crossings, and kerb cuts)
- Adding and editing sidewalk amenities like benches, bicycle parking, garbage bins, etc.
- Adding and editing trees
The "Current Mapping Project" section documents all active mapping tasks.
footpath.ai Data Team
The footpath.ai team improves OpenStreetMap by fixing errors and adding data using proprietary data made open to the OSM community, government open data sources, and customer feedback. Active team members are 
Meead Saberi (on osm) and Tanapon Lilasathapornkit (on osm)
Information about all other team members can be found here: [1]
Current mapping projects
List of Projects we are working on:
| Activity | Geography | Description | Tag |
| Adding and editing sidewalks and amenities | Sydney, NSW
Australia |
This continuing project aims to edit and add sidewalk data to OSM using footpath.ai proprietary data (with an open license for the OSM community) and existing government open data sources. Each area is first surveyed by a group of walking mappers who collect street-level imagery captured on the sidewalks only. Images are then processed through automated and manual workflows to map sidewalks and amenities (street furniture) on the sidewalks. Once sidewalk and amenities data are generated, we upload point data (both individually and in bulk) link benches, bins, bicycle parking, and trees after human validation, and conflation with existing OSM data to remove duplicates and minimize false positives using JOSM and Rapid Editor. Given the complexity of the sidewalk data, line data representing sidewalks and crossings are only uploaded manually using Rapid Editor, with no bulk upload. | #footpathai |
Tools & Data Sources
Main tools that footpath.ai team uses for mapping are JOSM and Rapid Editor.
footpath.ai uses several sources to enhance OSM:
- footpath.ai street-level imagery
- GPS traces
- Satellite imagery listed in the OSM Rapid Editor
- Open-source external satellite and aerial imagery
- Feedback from OSM contributors and users
- Local knowledge of the footpath.ai mappers in their locality
We’re happy to share any internal sources that we’re using if requested by the OSM community.
Rights for OpenStreetMap
You are permitted to use footpath.ai images to gather metadata for contributions to OpenStreetMap, in compliance with the Open Database License (ODbL). This metadata may include details such as sidewalks, stairs, curbs, trees, signs, POIs, benches, bicycle parking, and more. You may publish the extracted metadata directly to OpenStreetMap, in accordance with the OSM License/Contributor Terms. We recommend using the tag source=footpath.ai or providing a link to footpath.ai.
This license is exclusively for contributions made to OpenStreetMap. Any other use of footpath.ai images and data outside OpenStreetMap must comply with the footpath.ai general Terms of Use.
Other Contributions
- State of the Map US 2023 | Mapping Sidewalks and Street Furniture with AI and Computer Vision
- FOSS4G Thailand and State of the Map Asia 2023 | Enhancing Equity and Sustainability in Urban Planning: The Use of Artificial Intelligence and OSM to Identify Heat-Vulnerable Bus Stops
- State of the Map Asia X Pista ng Mapa 2022 | Moving from car-centric to people-centric mapping: using artificial intelligence and computer vision to map sidewalks