Mobile Toggle

Why Virtual Reality will not scale for Enterprise Training!

Why Virtual Reality will not scale for Enterprise Training!

Posted By: John Cunningham, Chief Revenue Officer and President of the VR/AR Association of Central Florida.

Did this capture your attention?

There is clearly a lot of excitement and investment to use Virtual Reality within the Enterprise to improve worker performance and reduce operational costs. Superdata (thanks Stephanie Llamas!) recently issued a report that stated that 71% of the $5.1B spent in 2018 on VR was related to training applications (https://www.superdataresearch.com/xr-training/).

There is no shortage of VR experts and Solutions providers who make the claim that soon all training will incorporate virtual reality.

Sorry folks, I hate to be the bearer of bad news but VR Training will not scale in the Enterprise the way it is being developed today. While there are many very compelling use cases, pilot projects, and even some positive implementations of VR for enterprise use, they have not been designed to scale across the enterprise to train on highly complex and expensive equipment where the real ROI is to be uncovered and where companies are willing to invest. Why is that you might ask?

Today, almost all VR training development is done using a “hand coding” process. Meaning that developers are creating the VR training SOFTWARE by hand in Unity and Unreal and creating really cool “experiences” that get training managers, technical users, and even executives excited. However many people don’t understand what had to happen in the background to create that software and what the limitations and true costs of this approach are when it is time to scale it across the enterprise.

The hand-coding process is fine for creating, one-off training “experiences” and proof of concepts, but imagine creating a virtual training application for operating and maintaining complex machinery such as an aircraft or a piece of power plant equipment. Those systems have thousands of graphical objects and hundreds of procedures that generally come from operations and maintenance manuals. It is fundamentally not possible to manually develop and manage this software without armies of developers…and time! That is one of the reasons that VR is not scaling yet…it takes too long and is too expensive for most enterprise customers. The manual hand coding approach can work with military applications because they have multimillion dollar budgets for development and can wait 2 years for a virtual trainer to come out of the process, but that won’t cut it in the enterprise. We live (or die) by the quarter and need to show incremental progress and ROI!

For the past 15 years the DiSTI Corporation has been developing Virtual Training Solutions for Aerospace, Defense and Industrial customers. Starting off with small Part Task Trainers such as engines or mechanical pumps all the way up to highly complex military aircraft, automobiles and power plants. The engineers are pioneers in the process of developing high fidelity, procedure-based virtual training solutions and have the scars of trying to use hand-jamming techniques. In order to overcome these limitations of scale, the DiSTI engineers developed a methodology and process that utilizes a relational database to store and connect all of the objects, models, procedures, constraints, behaviors and relationships and development can be done using a GUI. This approach allows rapid development and iteration management to support the constant changes that occur with training systems.

This approach also enables development work to be done by non software-engineers which further reduces costs and time. We have data proving that the database-driven approach can reduce initial development time by more than 50% and reduce the cost of development by 50% or more, while supporting constant changes and adaptation of the content to support new technologies such as VR, AR, MR, Tablets, etc.

Below is a cost comparison for a project that involved developing a virtual training solution for a large piece of industrial equipment with approximately 200 individual tasks.

This is 60% difference in the cost of initial development.

Initial Development costs are only one area to consider, there are also the costs of changes to the training system. Anyone who has experience in training knows that the minute the training package is developed there are changes that need to be made. So the costs of changes can be quite significant and over a period of time can be more that the initial cost of development.

Example: One company shared that they spend more than 20 man hours on average whenever they need to make a change to their training system (that was developed in Unity using a non database driven methodology). They have on average 4 revisions per quarter to their training application and they have to use Unity Developers to make the changes to the code. So assume $100 per hour for a Unity Developer.

With a database driven approach (using VE Studio for Unity) that same change takes approximately 4 man hours. With VE Studio you can also use a non-developer resource (training team member, technician) using a non-coding UI so the labor rate is less…$60 per hour.

Non Database Driven = 20 hours x 4 changes = 80 hours X $100 per hour (blended labor rate) = $8,000 per quarter.

Database Driven = 4 hours X 4 changes = 16 hours X $60 per hour = $960 per quarter

So while I still maintain that hand-coded VR training will not scale in the Enterprise, there is hope that by using a database-driven methodology that greatly reduces time and cost that will make it possible to scale VR training across the enterprise

Please feel free to comment or challenge me. We are all pioneers in this emerging technology and can benefit from everyone’s experiences!

If you want information about how to scale your Enterprise VR training please reach out to me at jcunningham@disti.com.