Loading and sharing models converted by Unity Reflect

Greetings,

We are currently developing a protoype VR viewer for BIM models and Unity Reflect seems our best bet for importing models created with BIM modeling software.

Initially we thought of using the default Unity Reflect Develop workflow (broadly described here):

• The client would create a model in the BIM modelling software of their choice (Navisworks, Revit, etc)
• We would upload that model to Unity Reflect.
• Import the model from Unity Reflect into our Unity project (as a prefab + texture/materials/meshes/prefabs folders)
• Customize the final result and release.

Problem is, our client wants to edit the model down the line, and see those changes when running our software right away, without having to change the VR Viewer executable. This breaks that workflow since we will have to import and load converted models at runtime, which would be fine (Unity Reflect Viewer is proof that it is possible) except there is one BIG caveat: The viewer software cannot connect to the internet, just to a local network. At least not when on duty.

We have thought of several scenarios to circumvent this issue, but we need confirmation on some points, to check if they are feasible:

Scenario A:

• Client uploads the changes to unity reflect.
• Starts our VR viewer on a device connected to a network with internet access to update the model from Unity Reflect.
• Use the VR viewer connected to the local network enviroment with no internet access.

- Can Unity Reflect downloads be cached/stored in the VR viewer to be used offline at a later time?

Scenario B:

• Client uploads the changes to unity reflect.
• Client uses Unity Reflect Develop to import the model into a Unity project on their side and pack the assets (prefabs + folders) into an addressable.
• The VR Viewer would check for addressables with converted models and show options to pick a version.

Are those valid scenarios?

Also, do Unity Reflect perform some kind of optimization when converting models to unity prefabs? As I'm sure you know, VR is a demanding platform in terms of performance and framerates below 75 fps start to become uncomfortable for users. If we could follow the standard workflow we could apply some optimizations but loading models at runtime rules the optimization process out.

Furthermore, is it planned a 100% offline version of Unity Reflect, where the conversion process can be done without having to connect to Unity servers and/or convert at runtime? I know this may be a stretch but it would be a silver bullet for this type of situation.

Kind regards

 

At this moment we are targeting desktop platforms. The offline requirement is due to client's internal security policy. In fact we are discussing if it's allowed to sync models to Unity Reflect's servers, which if I'm right, is an imperative step.

Thanks for the fast response.