AR: The role of open map data on the next computing platform
summary
Argues AR will be the next major computing platform and maps must become richer for it; demonstrates concrete work: machine learning on street‑level imagery to extract sidewalks and buildings, conflating hundreds of millions of buildings into OSM via rapid editor and human review, generating 3D building data from open LiDAR, and using VPS (visual positioning) for sub‑GPS localization. Shows these pipelines run in cities (Boston, Washington DC), are already usable for pedestrian routing and AR overlays, and require combined open datasets plus validation.
AI-written summary · about
chapters (10)
- 0:00Introduction to AR and open maps
- 2:27History of AR applications
- 5:15Modern AR use cases
- 7:45Why meta uses open map data
- 10:55Open data tools and initiatives
- 13:25AR as next computing platform
- 16:53Improving pedestrian navigation data
- 18:18Visual positioning systems
- 20:06Future AR experiences
- 20:52Q&A with audience
related talks
15mPutting the AR puzzle together with OpenStreetMap
48mEmpowering AI Innovation Through Open Geospatial Ecosystems
18m4 Keys to Augmented Reality’s Future
39mBig AR Talk: Three Innovators Discuss Unlocking AR's Potential
38mIntro to ARCore
16m"Ubiquitous GeoContext"
38mDeveloping the First AR Experience for Google Maps
14mOverture and OpenStreetMap: Expanding Open Map Data