Turning a physical eye routine into software is mostly an exercise in deciding what to leave out.

Any therapy that lives in a textbook faces the same barrier. Patients hear about it, rarely learn the correct way to practise it, and quietly stop. Eye exercises are a clear case, which is why a team at Universitas Brawijaya looked at whether a phone and a cheap headset could carry the routine instead.
The Bates method contains several distinct techniques. Before writing any code, the authors checked each one against what a mobile headset can physically do, and three of them failed immediately.

Two techniques made the cut. Head swing became a pointer driven by the phone gyroscope, moving through the 3D axes so the head has something to follow. Optical swing became a moving object that the eyes track while the head stays still. Both target the same muscles that control accommodation, and both map cleanly onto hardware that already exists in every phone.
The application guides the user step by step, logs exercise history so progress is visible, and switches into a dedicated VR mode once an exercise begins. Small choices like the history log matter more than they look, because they turn a routine into something a person can actually keep up.

To let users see whether the effort pays off, the app also measures refraction. It uses the front facing camera to capture the face, locates the eye position, and converts the measured distance into a diopter value that is shown on screen. That closes the loop between doing the exercise and knowing what changed.

Across a five week trial with 14 participants, the average change was a 0.13 diopter reduction, and participants said the guided format made the exercises simpler, faster and more enjoyable. They also flagged eye fatigue after long sessions in the headset, a limit the authors connect to earlier findings about prolonged VR use.
Brata, K. C., et al. (2023). Virtual reality eye exercises application based on bates method: a preliminary study. Indonesian Journal of Electrical Engineering and Computer Science.