At VJ Yourself Ltd we developed an online management platform for our 3D and VR applications. Our goal is to speed up the development cycle, and make complex data and application-code dynamically editable.
 

Extending Unity3D's game-engine with web-based cloud & CMS helped us to keep the felxibility of the web and couple it with the rendering and VR capabilities of a high-end gaming platform.  

We use a standard document model to represent any data or code stored in the cloud. These documents (serialised data + meta info) are stored in MongoDB and managed by a Node JS based service and available for the clients via the Rest API.

The web based client and the Unity application are using the same Rest API to communicate with the cloud. On the web we can use any standard frameworks (ReactJS, Angular), but for the standalone app we had to develop our own client-cloud library. It gives seamless integration for the rest of the C# code.

In the cloud we store various data types: 3D models, textures, serialised objects, numerical data sets ... Their actual structure and validation are described by JSON Schema, stored in the DB as well.

Each object is extended with standard meta information: name, thumbnail, tags, etc.. Through these meta data, any content management system can easily query, visualise and manipulate them, without parsing and understanding the underlying data.

The next screen shot is demonstrating the web-based admin page, querying 3D models.

Just by manipulating the meta-data we can create new data structures an relationships ( links ) between objects. On the next slide we are building a new ordered sequence of 2D Textures.

The cloud and the VR run-time are ready to handle numeric data-sets. They are managed via the same web-admin introduced previously.

3D data-set from the stock market, visualised as a 3D heat-map

and as 3D bar chart

Complex networks and graphs are another relevant field for VR. We created a dynamic visualiser for  graphs stored in our cloud. The next slides shows the London Tube map in JSON data, and then images from the interactive VR visualization.