Augmented Reality Indoor Positioning and Navigation
Conducted as part of a group with Anna Horwath and Christian Neurohr
Relevant GitHub Repos:
Abstract
Indoor navigation is a topic undergoing research within robotics, but can also be applied in commerce, manufacturing or industrial applications. We investigate how AR Head-Mounted Display Magic Leap 1 can be integrated into an indoor navigation framework aimed for commerce or asset management. Magic Leap 1 supports spatial mapping of real-life environment as well as tracking users in six degrees-of-freedom, and provides the ability of recognizing and estimating the orientation in space of fiducial markers. We propose an end-to-end framework that is able of mapping out an environment, digitally processing it by using the A* algorithm for shortest path, and providing a framework for connecting client devices for issuing navigation requests.
Our system proves to function end-to-end although fails to improve task completion time when tested in a mock-up environment. This can be attributed to over-simplified test task, low resolution of guidance elements and general inexperience with AR devices by the test subjects. Our study concludes that the first iteration design should have an improved user interaction and it should be tested with a different methodology. It also identifies future work and features such as IoT communication, computer vision integration and a revamped user interface based around human-computer interaction research.