California teens build AI fall-risk system, win $100,000 scholarship

Two talented teenagers, Ruoqi Li, 16, and Jason Yang, 17, from California, have created an artificial intelligence system called SafeStrides. This innovative program uses a smartphone app along with wearable sensors to detect potential fall risks in older adults. The project earned the duo the prestigious 2026 Davidson Fellows scholarship, awarding them a combined $100,000.

SafeStrides merges the capabilities of a smartphone camera with wearable sensors to analyze a person's walking patterns. During a brief guided test, the application gathers video and sensor data to form a comprehensive assessment of the individual's fall risk. The wearable component of the system employs inertial measurement units (IMUs) and pressure insoles.

The IMUs capture movement-related information, while pressure insoles provide data on how a person distributes their weight on their feet. By combining these distinct data types, the system can perform a more detailed examination of walking and movement.

The primary objective of SafeStrides is to identify potential fall risks months before an actual fall occurs. This allows older adults to proactively work on improving their mobility and overall safety. The system was developed in response to the limitations Li and Yang found in traditional fall-risk assessments. They noted that clinical assessments often require individuals to visit medical facilities, and these assessments typically only happen once a year, making it difficult to detect sudden changes in physical condition.

Building SafeStrides involved multiple stages of engineering work. Initially, the teenagers created a bulky Arduino-based prototype. However, it had limitations and could only analyze data after the testing was completed. Their second-generation system upgraded to IMUs, pressure insoles, an ESP32 microcontroller, and Bluetooth modules. This iteration allowed for real-time collection and transmission of sensor information.

The final version of the wearable system focused on a more compact and lightweight design, ensuring it could be worn comfortably without hindering normal movement. Simultaneously, they developed a cross-platform mobile application using Flutter. This app connects to the wearable hardware via Bluetooth and synchronizes high-frequency sensor streams with live camera footage, creating a multimodal approach that integrates different forms of data for analysis.

SafeStrides is designed to be used both at home and in clinical settings. At home, older adults could complete guided assessments using the smartphone app and wearable sensors without the need to travel to a medical facility. The app would then provide an immediate evaluation and suggestions related to mobility and home safety. For medical professionals, SafeStrides offers a convenient tool for conducting fall-risk evaluations without the need for direct physician involvement in every screening. This approach aims to make assessments more consistent and accessible.

The system is intended to benefit the world's 1.2 billion older adults and could shift the focus of fall prevention from occasional screenings to continuous monitoring. This innovative approach differs from traditional reactive medical alert devices, which are primarily designed to alert someone after a fall has occurred.

Ruoqi Li sees her recognition as a Davidson Fellow as a stepping stone towards her future goals. She intends to study an interdisciplinary field combining artificial intelligence and healthcare, with the ultimate aim of developing technologies that enhance quality of life. Li expressed her gratitude for the recognition and her excitement to join the community of Davidson Fellows.

Jason Yang, on the other hand, plans to pursue a career in electrical or systems engineering, continuing to work on embedded technologies. He shared his gratitude for the award and emphasized the team's commitment to creating practical, human-centered technologies and raising awareness about early fall risk detection in older adults.