Project Goal

Duke nursing students travel to campus only a handful of times, so this project shows a future where they can complete a simulated joint injection in their home.

Development

For the first 2 months of this project, I met with Duke School of Nursing on a bi-weekly basis so I could figure out what, exactly, they needed from a VR project, and what was feasible within the remaining time of my fellowship. I wanted to ensure that a VR project would solve a problem for them, or fill a need, and not just be a shiny new toy. Michael Zychowicz, my partner in crime and the head of Duke School of Nursing, whole-heartedly agreed.

We eventually determined that a joint injection task trainer, where students practice injecting a patient in their elbow or knee, would be most easily replicated in VR. I wanted to understand more about the process, so for research I did the same task trainers as nursing students, and learned about the average student’s trip to campus. Based on this, I developed grey box prototypes in Unity for Michael to evaluate.

Once Michael and I had agreed upon a design, development unfortunately stalled due to a lack of funding. During that time I continued to refine our original design doc and focused on other projects. Eventually, Michael was able to procure a powerful laptop and an Oculus Rift, and I was able to develop the game.

Using purchased and free assets, I built Task Trainer VR in the span of a month, with the basic prototype coming together within the first week. Because we had an agreed upon design, most of actual development was spent on refinement, and hitting our goal of making the game feel less sterile than the average medical simulation (hence the fireworks).

Lessons Learned

Feature creep will always happen

Even with our originally small scoped design, I ultimately had to cull several nice to haves including an animated character, and additional mechanics. The time lost from waiting for a computer and VR gear simply couldn’t be replaced, and we had to scale accordingly. This seems obvious now, but in the future I’ll ensure that equipment is available before taking on a project.

Sometimes less is more

This project needed to be a proof of concept – anything more, such as additional modes, features, and elements, would’ve only added to our feature creep, and with the limited amount of time I had, I wanted to make the best version of what was possible.

VR development is a whole ‘nother ball of wax

During my simulation fellowship I made several 360/VR video simulations and naively believed that the jump to full scale VR development would be just as easy. After all, I had made several games at that point.

It turns out that all the warnings about VR that I knew very well – it’s a new medium, it requires new methodologies, etc – are very true during development.

That said…

It’s never been easier to develop in VR

Beyond the initial hardware requirements, it’s laughably easy to get something up and running in VR. Oculus’s plug ins and sample scenes are excellent teaching tools, albeit ones that do not show to replicate them. I’ve felt more confident about tackling VR projects since Task Trainer, and like many others, I’m excited to see where this medium goes.