The Technical Cornerstones of Professional VR

In the first blog in this series we talked about the need to consider (even if briefly) your organisation’s ambition around VR.  

We looked at how there’s a huge spectrum of options available, each complementary to the other, but no one that’s the outright winner in all scenarios. We finished by picking up on the idea that, whatever the option, a successful VR set-up is about marrying a number of technological dimensions together in a way that allows your intended users to work naturally in a manner that best fits their context, immediate requirements and expectations. In order to explore that in a little more depth, this blog will look at the technology behind every successful VR deployment.

In essence, a VR system is made up of four components, the engine/software to drive the graphics, the tracking to know where you are in the scene, the hardware that is allowing you to see the 3D scene and most importantly your 3D content or data.

This means that if you’ve just bought yourself a HMD off the shelf, you’ve got your hands on two of those four components. If I may use my favourite car analogy, it’s like having the body work and chassis/wheels without the engine and with the risk that you’ve not picked the right combination to take you up some unpaved roads in winter which is a key requirement you happen to have. This is back to the roadmap topic of the first blog again.  

In fact, it goes farther than that. It’s not just about making sure you’ve chosen the right body shape, sufficient suspension travel, transmission ratios etc., it’s also about the experience of driving and the ability of that car to take you places. The same is true of an effective VR system and while it becomes more important the larger the system you’re considering, you still need to bear it in mind if you’re considering a basic PC/HMD type set up. In this case, you’re more able to compensate and adjust if you get it wrong, but better again to have thought about the user experience and objectives up front.

Taking each component in-turn, let me explain the role they play and what questions you should be asking yourself when selecting them.

3D VR Software

The Software/Engine, drives the graphics, takes input from the input devices and tracking, and also provides the platform or environment for you to develop your 3D scene.

When choosing the software, you should ask yourself:

• Does it support large complex 3D models and can I easily change the level of density of the mesh? You don’t want to find that you’ve locked yourself into models of a certain scale and so ruled out particular use cases.
• Does it support animated data, so I can show how my 3D model moves? The world is dynamic, your visualisations will need to reflect this    
• Do I need to code every element of my visualisation or can I make updates via a what-you-see-is-what-you-get interface? The world is dynamic and requirements evolve: if every change requires code changes and recompilation, you may find that this becomes an unsustainable cost to supporting your VR operations
• Can I ingest data from a range of sources? Visualisation is not just about geometry.  In many use cases the true value derives from being able to unpick the meta-data that sits underneath this.  The sources for which may come from a variety of non-CAD sources.
• How quickly can I get to a viewable and accurate visualisation once I have read in my data? The value of a visualisation is obviously undermined if the importer you’re using is removing layers of important detail or is otherwise reducing the fidelity of the output.  The usefulness is reduced if updating your visualisation is such a laborious process that you can’t rely on having the latest information to hand within a reasonable time-scale.
• Does it support a variety of VR hardware and viewing options, so I am covered for future hardware developments? As the roadmap piece noted, there isn’t one viewer option that suits every user, all the time.  You need a solution that covers all the bases and will evolve to support emerging technologies
• Does it support collaboration, so I can go online with my colleagues and work together on a 3D model? And if it does allow collaboration and sharing with others, does it do so securely?

Tracking System

The tracking system determines your position in the 3D world, and usually uses a tracking sensor camera that records your movement. With HMDs, such as the Vive or the Oculus, these tracking sensors are supplied as part of the headset bundle, but that is not the case with all headsets.  If you are using a 3D screen, Activewall or “Cave”, like our own ActiveCube, then you need to have a method of tracking the user’s position as part of the system.

When choosing the tracking, you should ask yourself:

• Will I just need to track my head & hands positions?
• Will I need to track my whole body and all my limbs?
• Will I need to track multiple people within the same tracked space?
• Will there be multiple devices to track, such as a controller, a tool, or a weapon?

Viewing System

The viewing System is the hardware that allows you to see the 3D scene in stereo. HMDs are obviously one option. Larger scale facilities such as ActiveWalls, 3D Screens and “CAVEs”, like the Virtalis ActiveCube, all deploy stereo projection and when combined with a user wearing 3D glasses, a 3D image appears in front of you.

When choosing the viewing system, you should ask yourself:

• Will it just be me viewing the VR scene (in which case an HMD is an option) or will it be a group of viewers (in which case an ActiveWall or ActiveCube is the right option)?
• Will it need to be a multi-user experience with networked HMDs, or connected to ActiveWalls?
• Will it be collaborative, with colleagues connecting in from other locations?
• Will I need to be fully immersed, so I experience all the sensations, such as height or noise?
• Will it need to be multi-screen?

3D CAD DATA

The 3D data or 3D scene is what you see and what you interact with, so what is in the scene and what you can pick and interact with is important. If you use the “create and publish” model used by many real-time engines, then you have to remember to code into your application everything you are going to need to do in the 3D scene, much the same way as a game developer codes into the game all the functionality he wants you to have. The interactive approach used by VR software such as Visionary Render, means that you are constantly in a menu-driven, interactive 3D scene, so your options are not limited – much the same way as your CAD software works.

Some other points to consider when you build your 3D scene are:

• Will your scene by controlled by multiple input devices or external inputs, like a live data feed?
• Will it use multiple datasets?
• Will it use multiple data types?
• Will it need to be distributed via a network, or to users where you want them to have limited functionality?
• Will it need to be collaborative?
• Will it use mechanisms?
• Will it use custom GUI?

Like I said, it is a lot to consider.  But it is entirely possible to quickly iterate toward a roadmap which will allow your organisation to plot an effective route forward.  At Virtalis we have been advising organisations of all scales on this very topic for thirty years, which is why we feel confident in offering the following advice, which is essentially the TL;DR of this and the previous blog: VR has proven itself to be an extremely effective industrial tool.  To make sure it works for your organisation now and in the future, just take a moment to think about what you wish to get from it and who you wish to use it.

We, of course, would love to help you plot that course and help you on that journey.

Read our previous blog in this series