OntZORGt

OntZORGt trains Dutch care workers to spot early warning signs during home visits, like confusion, missed medication or a client quietly struggling on their own. That's hard to teach from slides. They wanted a VR training that runs in the browser and on a headset, for four different care roles. It also couldn't be a one-off, because they wanted to keep adding scenarios without needing a developer.

I built the framework the project runs on. You visit Mevrouw Chong's home, move between rooms shot as 360 photos and use hotspots to talk, examine and complete tasks. Each scenario is a single script file describing its conversations, close-ups and which hotspot unlocks which, so colleagues could build new scenarios without writing code. In VR a stutter or a blank texture breaks immersion fast, so the loader only keeps the current room in memory and never fades in a room before its image is ready.

The training has four roles, each with its own home visit, from a caregiver's social visit to a practice nurse's clinical check. It runs on desktop and Meta Quest.

The training is live and in use, and colleagues keep adding scenarios on the framework, which was the goal from the start. The VR performance work was new territory for me, and it showed me that a lot of what feels like polish in VR is really about avoiding the small glitches that break immersion.