Indoor Navigation with Augmented Reality in a hospital

We implemented a prototype of an indoor navigation app that uses Augmented Reality for localization and showing the way. The position and orientation of the device is determined by the view of the camera only. For localisation we use multiple points clouds, "Area Targets" from the Vuforia SDK. An Area Target is a set of data (so called "point cloud") which contains information of unique points in the environment. A navigation can be started from any position of the supported area, which is on two floors and in total 690 m² big. The app can be extended to support much larger areas by adding more Area Targets. We conducted two usability tests with total 20 people. Most of the people agreed that with such an app indoor navigation is easier than a 2D map or spoken information.

Paper in English (Short): https://ebooks.iospress.nl/doi/10.323...
Paper in German on ResearchGate: https://www.researchgate.net/publicat...
3rd place award at Medical Informatics Congress (DMEA) in Germany 2021: https://www.bfh.ch/de/aktuell/news/20...

Tutorial Playlist:    • AR Indoor Navigation Tutorials  

Useful links:
How to create & use Area Targets: https://library.vuforia.com/features/...
The fastest way to a destination is calculated by using Unity's NavMesh (actually from NavMeshComponents) and NavMeshAgent:
https://docs.unity3d.com/Manual/nav-N...
NavMeshComponents: https://github.com/Unity-Technologies...
Vuforia Fusion: https://library.vuforia.com/articles/...

Chapters:
0:00 Difficulty of way finding in hospitals
0:16 Start app and select destination
0:26 Searching initial position
0:35 Navigation
1:26 Making of
2:00 Conclusions

The app was created with Unity and is available for Android and iOS (only as Beta-Version). Because it only makes sense to use the app in the supported area the store links are only provided there. The app was translated in German, French and English by using the package "Localization" of Unity.
https://docs.unity3d.com/Packages/com...

This project is the Medical Informatics bachelor thesis of Michael Däppen and Joshua Drewlow at the Bern University of Applied Sciences. The prototype was implemented in the college building of the Medical Informatics study course due to the Covid-19 pandemic. For more information just leave a comment below.