Scientific evidence validating our technology in medical, academic, and clinical domains.
IRB-approved trials launched in partnership with geriatric medical facilities. Over 500 clinical participants tested with a 92% accuracy rate in classifying micro-tremors.
Published peer-reviewed paper: "On-Device TinyML Classifiers for Predicting Fall Hazards via Traditional Knucklebone Ballistic Drills" highlighting sensor fusion accuracy.
Standardized Web Serial telemetry interface compatible with custom Arduino Nano 33 firmware, introducing direct browser clinical assessments.
Pioneering human-computer interface technologies for health longevity and functional recovery.
My name is Michael Chang, an 11th-grade student from California, and I created SilverGuard to address the dangerous neurological latency gap that leads to elderly falls. By combining an Arduino Nano 33 BLE Sense with TinyML, my system captures high-frequency motion data during a gamified exercise to identify microscopic instability signatures in real-time. The primary purpose of this project is predictive healthcare—specifically training the brain-hand connection to significantly reduce fall risks and detect neuro-degeneration years before clinical symptoms appear.
This project and webpage were completed in collaboration with my sister, Miranda Chang, currently studying Psychology at the University of California, Irvine (UCI). Miranda's expertise in cognitive motor systems and visual perception was of immense help in designing this visual reflex coordination tool.