International Journal of Scientific Research and Engineering Development

Assistive mobility devices for the elderly and visually impaired increasingly use IoT sensors for real-time obstacle detection, gait monitoring, and fall prediction, but centralized approaches risk privacy breaches and fail to personalize across non-IID user patterns. This paper examines federated learning (FL) within hybrid edge-cloud IoT architectures to enable privacy-preserving personalization. The proposed framework combines low-latency edge inference on resource-constrained devices with secure cloud aggregation of model updates. A prototype smart walking stick—equipped with ultrasonic sensors, IMU, and ESP32—serves as the case study. Trials demonstrate ~96% fall-prediction accuracy, 86–89% personalization gain in obstacle/gait adaptation after 10 FL rounds, <45 ms edge latency, and ~65–70% reduced data upload versus centralized methods, with full privacy via differential privacy and secure aggregation. This addresses a key gap: while FL-hybrid systems advance in wheelchairs and general IoT-healthcare, dedicated implementations for smart walking sticks remain rare in 2026 literature.

Dhakshitha M K, Mrs. N Vaishnavi,"Privacy-Preserving Federated Learning for Smart Walking Sticks Using IoT" International Journal of Scientific Research and Engineering Development, V9(1): Page(1392-1398) Jan-Feb 2026. ISSN: 2581-7175. www.ijsred.com. Published by Scientific and Academic Research Publishing.