Augmented Reality solutions often depend on a 3D model of the surroundings so they (1) can associate what the user is looking at with specific physical equipment (e.g., for diagnostic purposes) and (2) can generate an augmented image that can be accurately overlaid over the physical image the user sees.
In some cases there is an existing CAD model that can be used (see Augmented Assembly Use Case) but in many others a model of the physical environment (a “mapping”) has to be first created. A SLAM system uses image processing on what the user is seeing to build an abstract model of their physical reality (a 3D model), that can then be used to locate the user, and as the basis for visual augmentation (hence Simultaneous Location and Mapping). A “cloud” SLAM is where the SLAM processing is provided from cloud services.
1000 Realities is a Polish software company focussed on simplifying and accelerating the creation of industrial AR solutions. They have adapted their AR Assistant Software to run in the mobile edge.
Business Need: Accelerate the ability to incorporate AR applications (for the value they deliver).
Consumer AR headsets are creating large product volumes that drive display and electronics development, very significantly lowering the cost of headsets that can be used for commercial and industrial purposes, in turn enabling solutions such as these.
Edge Need: SLAM processing typically requires a GPU accelerated server. Using a cloud slam eliminates the need to deploy and operate that server locally (e.g. in a factory) but requires a low latency, high-bandwidth connection to provide an adequate augmented display. Hence, edge cloud services seem like an ideal solution.
Ease of Incorporation: Manufacturing processes are often controlled by complex, legacy, 3rd party software systems, that need to be significantly modified and adapted to incorporate solutions such as these.