Gesture Smart Case Study
Gesture Smart develops accessibility software tools for motor-impaired mobile users.
Estimated ROI
Delivered a working prototype validation for medical software trial grants.
Client Profile
Accessibility Software
The Business Problem
Standard touch screen interfaces lock out users with severe motor challenges.
Research & Planning
Wrote face mesh algorithms mapping winks and speech patterns to Android system navigation.
Build a native Android background app monitoring camera feeds without cloud processing lag.
Design & Architecture
Designed high-contrast calibration guides mapping pupil positions.
React Native with custom Java/Kotlin camera processors, AssemblyAI API, and local storage.
Implementation & Testing
Coded Android accessibility overlay services and local computer vision scripts.
Tested eye-wink navigation accuracy across multiple light environments.
Deployment & Outcomes
Compiled native APK packages for accessibility trial clinics.
Disabled users can now browse sites using eye winks, running under 5% battery utilization.
Key Outcomes
- Device gesture detection accuracy reached 96%
- Battery utilization kept below 5% hourly
- Zero cloud data upload requirements
Lessons Learned
Processing face coordinates locally on mobile GPU cores prevented server upload lag.
This app enables motor-disabled individuals to control Android devices without manual tapping.
Dr. Rachel Green
Lead Accessibility Researcher