An Efficient Approach to Optimizing Assets for Mixed Reality

Author by Sam Figie

Companies are beginning to really adopt mixed reality technology solutions to solve real business problems.  Mixed reality or MR for short, being the concept of melding the virtual with the physical world through technology device platforms capable of bridging this space between worlds: 3D and virtual concepts existing in and interacting with physical space.



Why Invest in Mixed Reality?

MR solutions are already being used by some companies to aid in rapid design workflows, customer experiences, or even remote maintenance scenarios. In fact, Gartner predicts that by 2019, 20% of large-enterprises will have evaluated and adopted Augmented Reality, Virtual Reality or Mixed Reality technology.  It is often the case that companies which can benefit the most from MR technology are companies that also have an existing CAD engineering 3D asset workflow.  This means there is often a library available of potentially thousands of 3D assets that could be incorporated into a Mixed Reality solution.  The only challenge present is these assets need to be optimized and converted so they can be rendered in a real-time 3D application engine such as the widely popular Unity 3d.


Optimizing Assets via an Automated Approach
Optimization is key when it comes to incorporating CAD assets into a MR experience.  It is often the case that most CAD drawings are extremely detailed, and the geometry can contain millions upon millions of polygons and vertices.  Such highly detailed assets are naturally not suited for real time rendering in an application running on an augmented reality device such as Microsoft’s HoloLens.  This being the case due to the hardware profile of the HoloLens is closer to a top tier mobile phone as opposed to the engineering graphical workstation which was used to create the asset initially.

The process of reducing complexity of CAD assets can be a manual process, weather a technical artist reduces the level of detail or even goes as far to create a low fidelity version of the original asset.  This approach is not ideal and increases both time and cost in implementing a Mixed Reality solution.  Luckily, there are more efficient solutions available, one being a software solution which Microsoft acquired in early 2017 called Simplygon.  Simplygon is a solution tool-set that is designed to optimize asset geometry and reduce the level of detail in a way that still maintains a high visual fidelity.  For this very reason it has been widely adapted in the video game industry and is a perfect candidate for optimizing high detailed assets.  These assets can then be incorporated into a VR/AR experiences allowing them to be rendered in a performant manner, at a stable framerate.


For larger enterprises with hundreds or thousands of CAD assets that could be potentially brought into a mixed reality application, Microsoft has created a cloud-based solution that can be incorporated into an asset pipeline called Simplygon Cloud.  By combining the power of Simplygon with the Microsoft Azure cloud platform, asset optimization and processing can be offloaded and incorporated in a larger scale pipeline with ease.  The Simplygon cloud service can be called via its C# SDK or REST API interface.  The asset flow is straightforward, first assets are uploaded into Azure BLOB storage along with an SPL file which defines what Simplygon operations will be performed on the assets, then Simplygon cloud service is invoked and specialized compute resources begin to process the assets.  When the asset job is complete, the output files can then be downloaded from Azure BLOB storage where they are in an optimized format ready to be used in a mixed reality experience.

Tags in this Article