Design of AI-Enabled Remote Monitoring Slippers for Elderly Fall Detection: Safety and Privacy via Cloud-Edge Collaboration
Mingyu Luo, Yu Cheng, Maodong Wu, Jianfei Ma, Zhengyu Zhang, Ye Yuan, Hongli Chen, Chong Xie, Awwab Hamam, Haoxuan Li, Yang Zhou, Yanze Wang, Jingyun Bi · Studies in health technology and informatics · 2025
As the aging population grows, fall-related injuries among the elderly require effective monitoring solutions. This study proposes an AI-based smart slipper to address the limitations of wearable devices in comfort and privacy. The slipper integrates an inertial measurement unit (IMU) and temperature-humidity sensors with an Arduino Nano RP2040 for data collection. A fully connected neural network (FCNN) analyzes motion data in real time to detect falls. Using an edge-cloud architecture, routine data is processed locally, and critical information is encrypted and sent to the cloud in emergencies. Tests show a detection accuracy of 96.57% and effective remote monitoring. Future work will optimize comfort, sensor integration, and real-world validation. This smart slipper offers a safe, comfortable, and privacy-friendly fall detection solution for the elderly.