Engineering for the Systemic: VR, Metabolic Health, and Visual Coordination
I've always been fascinated by how hardware can solve biological constraints. But as a dev, I often fell into the trap of thinking about 'features' rather than 'systems.' That changed when I looked at the new EASD 2024 data from Milan. The study shows that the cancer-diabetes gap has widened over the last 25 years. As someone building health tech, this was a signal: metabolic conditions aren't just about blood sugar; they create systemic comorbidities that our current screening tools completely miss.
I remember sitting at my kitchen table in Bristol, cold coffee beside me, scrolling through the abstract on my phone. The rain tapped the window. The numbers didn't just sit there — they pulsed. Widened gap. Twenty-five years. I thought about my dad, his metformin, the way he rubs his eyes after reading the paper. The way he says he's "just tired" by mid-afternoon. I thought about how easily we separate things: eyes here, pancreas there, brain somewhere else. The data said no. The body doesn't read our org charts.
One of the most overlooked areas is vision coordination. I have a friend — Marcus — who spent years struggling with amblyopia into adulthood. We met in a shared flat in Leeds, 2012, both broke, both staring at screens fourteen hours a day. He described it as a constant 'background noise' in his brain — a cognitive load that made him exhausted by 2 PM every day. Not sleepy. Drained. Like his brain was running a background process it couldn't quit. When we looked into it, years later, we realized that for many adults, especially those managing chronic metabolic issues, binocular dysfunction is a silent tax on their mental energy. Nobody talks about it. Nobody checks.
We wanted to build something that removed the friction of traditional therapy. No one wants to wear a patch while working or lounging. The patch is visible. It's awkward. It says patient. So we built Amblyotube for the Meta Quest.
From a technical perspective, the challenge was to implement dichoptic rendering in a way that felt seamless. We leveraged the Quest's stereoscopic display pipeline to ensure each eye received a different visual stream. Using a custom shader for per-eye texture assignment and the Unity XR subsystem, we created an environment where users can watch YouTube-style content while their brain is pushed to fuse the differing images.
The first prototype made me nauseous. Legitimately. I tested it in my spare room, headset hot against my face, a video of a Japanese woodworker planing a board — quiet, repetitive, perfect. The shader wasn't right. The textures flickered. My brain fought the split. I took the headset off, sweating, and sat on the floor for twenty minutes. This is why nobody does this, I thought. It hurts.
We rewrote the shader three times. Changed how the textures loaded. Added a calibration step so subtle you barely notice it — just look at the dot, tap the controller. The fourth build, I watched the whole twenty-minute video. Didn't realize I'd been fusing the image until I took the headset off and the room felt... sharper. The edges of my monitor. The dust on the shelf. My eyes worked differently for a moment.
We didn't want this to feel like a 'medical app.' We wanted it to feel like a video player that happened to be training your brain. By allowing users to choose their own content — cooking shows, slow TV, lectures, whatever — we solved the adherence problem that plagues most vision rehab. You don't "do your exercises." You watch the thing you wanted to watch anyway. The training rides along.
It's crucial to note that Amblyotube is an accessible training layer, not a medical device. It's meant to assist and complement professional care, not replace it. We're focusing on the educational and recreational side of visual coordination, providing a tool that helps users maintain their attention and coordination without the sterile vibe of a clinic. I say this not as legal cover — though it is — but because I've seen what happens when people think an app replaces a doctor. My dad almost did that with a blood pressure monitor once. Scared us both.
Building this taught me that the most effective health tech isn't always the most 'clinical' — it's the one that fits into the user's actual life. The one they use on a Tuesday evening because they're curious about that woodworking channel, and their brain gets a little stronger while they watch. The systemic view isn't just data. It's the quiet afternoon. The rain on the window. The choice to put on a headset instead of zoning out.
If you're building in this space — assistive VR, metabolic-neurological intersections, or just trying to make something that respects the whole person — I'd love to hear from you.
Check out the experience here: https://www.meta.com/en-gb/experiences/amblyotube/25906906972338493/
Top comments (0)