Related Experiment Video
Updated: Nov 23, 2025
08:05
Design and Analysis for Fall Detection System Simplification
Published on: April 6, 2020
11.0K
Feasibility of Using Floor Vibration to Detect Human Falls
Yu Shao1,2, Xinyue Wang1,2, Wenjie Song1,2
1School of Architecture, Harbin Institute of Technology, Harbin 150001, China.
Summary
This study uses floor vibrations and machine learning to detect elderly falls and postures. The developed framework accurately identifies falls, distinguishing them from object drops and various fall types.
Area of Science:
- Biomechanics
- Machine Learning
- Public Health
Background:
- Falls in the elderly are a significant public health concern, leading to injuries.
- Existing fall detection methods have limitations.
- There is a need for accurate and reliable fall detection systems.
Purpose of the Study:
- To propose a classification framework using floor vibrations to detect falls in elderly individuals.
- To distinguish between different fall postures and differentiate falls from object drops.
- To develop a machine learning-based pattern recognition system for fall detection.
Main Methods:
- A scaled 3D-printed model simulating human movement was used to generate fall data.
- Floor vibrations were recorded during object drops and human falling tests.
- Machine learning algorithms (K-means, K nearest neighbor) were applied for classification.
Main Results:
- Three classifiers were developed: walking vs. fall (100% accuracy), fall vs. object drop (85% accuracy), and different fall postures (91% accuracy).
- The framework demonstrated high accuracy in detecting and classifying fall events.
- Floor vibration signatures proved effective for fall pattern recognition.
Conclusions:
- A novel framework utilizing floor vibrations and machine learning for fall detection in the elderly was successfully developed.
- The system can accurately detect falls and classify different fall postures.
- This approach offers a promising non-wearable solution for elderly fall monitoring.
