Design and Experimental Scenario Evaluation of a Vision‑Based Walking Assistance System for Older Adults
Keywords:
edge computing, inertial sensing, time of flight, polarisation imaging, haptic feedbackAbstract
Pavement hazards require walking assistance systems that connect environmental sensing with timely and understandable warnings. This paper presents a wearable design using an Orange Pi 5, an ESP32-S3, RGB vision, array time of flight sensing, sequential polarisation and inertial monitoring. Directional vibration, speech and 4G/GNSS reporting support local warnings and post-fall assistance. An experimental evaluation scenario specifies YOLOv8n with Ultralytics 8.2.103, 640 × 640 inputs and 100 training epochs. Its dataset specification contains 3200 images, divided by route into 2240 training, 480 validation and 480 test images. The experimental test matrix contains 280 true positives, 15 false positives, 165 true negatives and 20 false negatives, giving 92.71% accuracy. The experimental scenario yields 25 frames/s pipeline throughput and 135 ms mean warning latency. Original reported stone detection results are distinguished from these prototype‑tested outcomes, with 91.1% object precision and 91.2% recall. The analysis quantifies the accuracy, throughput and persistence trade‑offs under clearly‑defined experimental conditions. These outcomes are experimental measurements obtained from model training and hardware device tests, while slippery‑surface warning and fall assistance remain design capabilities.
References
[1]Ma, W. (2023). 基于人机工学原理的老年人下肢外骨骼助行器设计研究 [Research on the design of a lower limb exoskeleton walking aid for older adults based on ergonomic principles] [Master's thesis, Qingdao University]. [in Chinese]
[2]Liu, Z., & Qiao, Y. (2024). 基于改进技术接受模型的老年智能助行器设计 [Design of an intelligent walking aid for older adults based on a modified technology acceptance model]. Journal of Machine Design, 41(12), 175–180. [in Chinese]
[3]Zhang, Y., Han, D., & Wang, X. (2025). 基于KANO/AHP/QFD的介助老人助行器设计 [Design of a walking aid for older adults requiring assistance based on KANO/AHP/QFD]. Design, (17), 36–40. [in Chinese]
[4]Chen, T., Wang, Y., Xu, T., Meng, Q., & Yu, H. (2026). 融合产品语义学的“隐形”下肢外骨骼设计研究 [Research on the design of an “invisible” lower limb exoskeleton integrating product semantics]. Packaging Engineering, 47(10), 22–31. [in Chinese]
[5]Liu, R., Xi, Y., & Chen, P. (2025). 基于深度学习的老年人摔倒行为识别方法 [Method for identifying falls of the elderly based on deep learning]. Electronic Components and Information Technology, 9(3), 228–230. [in Chinese]
[6]Shi, W., Cao, J., Zhang, Q., Li, Y., & Xu, L. (2016). Edge computing: Vision and challenges. IEEE Internet of Things Journal, 3(5), 637–646. https://doi.org/10.1109/JIOT.2016.2579198
[7]Zhou, Z., Chen, X., Li, E., Zeng, L., Luo, K., & Zhang, J. (2019). Edge intelligence: Paving the last mile of artificial intelligence with edge computing. Proceedings of the IEEE, 107(8), 1738–1762. https://doi.org/10.1109/JPROC.2019.2918951
[8]Wang, J., Sun, Y., Chen, Z., Jin, Y., & Xu, Y. (2023). 可穿戴式老年人跌倒防护装置设计与研究 [Design and research of a wearable fall protection device for older adults]. Chinese Journal of Medical Instrumentation, 47(3), 278–283. https://doi.org/10.3969/j.issn.1671-7104.2023.03.009 [in Chinese]
[9]Ultralytics. (2024). Ultralytics YOLO (Version 8.2.103) [Computer software]. https://pypi.org/project/ultralytics/8.2.103/
[10]Rougier, C., Meunier, J., St-Arnaud, A., & Rousseau, J. (2011). Robust video surveillance for fall detection based on human shape deformation. IEEE Transactions on Circuits and Systems for Video Technology, 21(5), 611–622. https://doi.org/10.1109/TCSVT.2011.2129370
[11]Saadeh, W., Butt, S. A., & Altaf, M. A. B. (2019). A patient-specific single sensor IoT-based wearable fall prediction and detection system. IEEE Transactions on Neural Systems and Rehabilitation Engineering, 27(5), 995–1003. https://doi.org/10.1109/TNSRE.2019.2911602
Downloads
Published
Issue
Section
License
Copyright (c) 2027 LingLong Yang

This work is licensed under a Creative Commons Attribution 4.0 International License.
