Indoor Positioning Techniques Applied to Collaborative AR

Augmented Reality applications use positioning of devices to render digital objects according to the user’s view. For indoor locations applications usually exploit QR-code, camera stream or markers, which introduces limitations in some scenarios. This paper presents a system that makes use of Bluetooth Low Energy (BLE) devices for indoor positioning of users to extend augmented reality functionalities, also enabling a collaborative remote interaction with 3D objects. Exploiting RSSI fingerprinting and machine learning algorithms, the proposed solution achieves a good localisation accuracy. The AR module, developed using A-Frame and AR.js, overlays 3D models in real time, offering immersive and interactive content without requiring dedicated mobile applications. Experimental activities and a web demonstrator are described.