ARCore and indoor navigation, lessons learned
summary
ARCore plus Sceneform and Firebase cloud anchors are used to build shareable indoor routes; cloud anchors host sparse point maps on Google servers and resolve in a few seconds but expire after one day and face quotas of 30 write / 300 read requests per minute. Long routes require multiple anchors to avoid drifting, which increases latency and server cost. Demo creates tappable 3D tube routes and an Android UI, and a private toolchain called Maggelana is built to produce and reuse AR routes across devices.
AI-written summary · about
chapters (10)
- 0:00Introduction to AR
- 4:27Indoor navigation challenges
- 7:23AR indoor demo teaser
- 8:29Android ARCore framework
- 14:00Cloud anchors explained
- 14:23Use cases for ARCore
- 20:24Building ARCore apps code
- 29:44Technical lessons learned
- 33:19Maggelana tool chain
- 35:53Q&A session
related talks
18mDeep-dive on ARCore | Fundamentals of designing AR apps with machine learning models
16mBuild location-based augmented reality with ARCore geospatial API
41mFusing AI with AR: Custom Markers
37mA Real use case of practical AR-AI couple - Size my luggage
39mA Real use case of practical AR-AI couple: Size my luggage
38mDesigning Augmented Reality Systems for Speed
17mLabelAR: A Spatial Guidance Interface for Fast Computer Vision Image Collection
39mNow smile (also in 3D)! Exploring AR, ML and Camera related APIs on Android