Model based tracking is a principle that uses augmented reality to project 3D objects on a physical object without the use of QR codes or trackers, but by using the outline of a CAD model as a reference for image recognition. Advantage of this workflow is that a environment or physical object does not need to be adjusted in any way to project the AR, and you can show digital information as an extra layer with great precision. Out prototype (made within the sivar project) shows the potential of model based tracking as a tool for virtual training on real objects: we visually guide a user step by step to showcase assembly steps.

In a second prototype we used model based tracking principles in a different scenario than training. This specific application of the technology could be used as a visual  communication tool between different departments in a production process environment. Image a assembly piece falling out by a quality check and it needs to be checked by the repair engineers. Now operators need to manually check and input information about the default and repair actions in a tree structure or material list. With AR model based tracking this process can be accelerated by adding and reading this information visually on the 3D model.

More information on: www.sivar.be