Augmented Reality only shines when it can show information relevant to the environment and relevant to the user. Indeed, if you want to augment a machine, you need to get relevant information to that machine at the right time. For example, the user may need to see the maintenance status of the machine along with some information on how to operate it.
To this end, the holy grail of AR is that the system is able to interpret objects (what and where it is), know the layout of the environment (can place objects with correct shadows and occlusion), remain persistent (if you look somewhere else, the 3D object should stay) and have ubiquitous networking availability (5G to download and show objects in real time). This is what one calls « deep AR« , versus « shallow AR« .
This is all promised in technology based on SLAM (Simultaneous Localisation and Mapping) which allows users to understand the physical world through point clouds, GPS and other inputs. This makes it possible for AR applications to recognise 3D Objects & Scenes, as well to Instantly Track the world and to overlay digital interactive augmentations.
Aside from being able to place objects, it is also important to get the correct objects to show. Through integration with a company’s back-end systems (ERP, SAP, SalesForce, …), the AR app will be able to discern which objects to show and when. This way, relevant information can be extracted at any time from the cloud and shown to the Augmented Reality user.