We built our own driving school in VR

december 06, 2017

Ahead of Valtech Day 2017 we built a VR application – a driving school simulator called VRoom. The project proved that VR has a bright future – both from a user perspective and a business perspective.

Developing for VR is more reminiscent of developing a game than of developing a website.

The role that VR may come to play in the future has become a bone of contention. One side argues that the technology is a gaming-driven flash in the pan, while the other maintains that VR is the ultimate empathy platform of the future. We felt it was an exciting and challenging subject to tackle – to try to prove that as a platform, VR offers long-term business prospects. Given the variety of backgrounds and perspectives of those in attendance, we thought that the upcoming Valtech Day would be the perfect forum for presenting our conclusions.

The idea of a virtual driving school came up in a discussion about the high costs of obtaining a driving license. When we started building our VR driving school, which we later named VRoom, it was very clear to us that we did not want to build it just because we could or because it would be really cool. We wanted to build a VR application with a clear purpose that it would be possible to deploy broadly and for an extended period of time. Even though the project is a prototype – so far– we needed to be able to measure and evaluate it to see if it would bring both user and business benefits.

But what are the problems, if any, that characterise current driving school courses?

In 2013, I, Robin, wanted to get a driving license, so I signed up for an intensive course. The instructors strongly recommended that I also practice between classes, but with my parents and their cars elsewhere, there was no way for me to do so. Things worked out anyway. I got my driving license but did not feel completely confident about certain driving manoeuvres. For example, the intensive course was a held during the summer, but I had exactly zero minutes of practical experience with winter driving. I had not attempted driving at night either.

Based on that experience, we were able to identify some of the issues with current driving school courses, which we then wanted to solve with VRoom. We decided that we wanted to build a product with:

• the ability to practice driving without needing access to a car or an instructor.
• the ability to practice situations not present during the time or in the geographic location where the student is taking the course.
• the ability to facilitate the training and increase the students’ confidence on the road even after the course is complete.

Could VR be a good fit?

In 2002, Danish researcher Tone Saugstad concluded that “theoretical knowledge is no substitute for practical knowledge of practical elements. Individuals cannot possess theoretical knowledge and immediately have command of its practical implementation. ”A driving course mainly has to do with practical driving operations and acquiring practical knowledge. Our own experiences with VR showed that it can give you the feeling that you are actually doing something – as if it were the real thing – despite the fact that it is happening virtually.

In light of these experiences, we were quite certain that a VR driving school simulator was the way to go.

We established early on that our experience needed to feel as real as possible, and we chose to incline more towards photorealism than stylised graphics, partly because we had a potentially broader target group in mind.

Although 3D modelling was nothing new to us per se, every project always poses its own challenges. What was crucial in this case was rendering optimisation. Developing for VR always requires a lot of processing power, and striving for realism can test the capabilities of even the most powerful computers. In order to achieve a realistic virtual world, for example, there needed to be a high level of shadow rendering and detail, which posed a challenge in terms of optimising performance.

Developing for VR is more reminiscent of developing a game than of developing a website. Compared to coding for websites, we had to develop in three dimensions with everything that entails, for example in terms of physics, mass and collisions.

Test the prototype

Once our prototype was complete, it was time to show it off and give people who were interested a chance to try it out. We stealth-tested VRoom a couple of times internally, but the trial by fire came during Valtech Day. The tests also yielded very valuable insights about what we need to keep in mind when developing VR applications. They also showcased VRoom’s enormous potential.

It was clear that it would be possible to meet the requirements we had set ourselves in advance. The first two requirements, to give users practical knowledge of, for example, winter driving conditions and the ability to test-drive without access to a physical car, were confirmed then and there. To make drivers feel safer even after they have obtained their driving license we need to perform corroborating user tests, but we have high hopes that we will be able to meet this requirement as well.

VR is more than just hype

To close, we can say that VR is definitely a tool that is more than just hype. VRoom made it clear that it is possible to build an application that yields long-term user benefit and business benefit alike. We would be happy to tell you more about the opportunities that VR holds for your industry and the challenges it poses. Don’t hesitate to get in touch if you want to learn more about VRoom – or about VR and 3D in general.