Boot Accelerometer Fall Detection via Slope Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current fall detection systems in shoes, which use accelerometers and gyroscopes, lack reliability in accurately detecting falls, especially for elderly individuals or lone workers, due to limitations in data processing and communication protocols.
Innovation Solution
A shoe equipped with an accelerometer providing data along three axes, a processing unit that compares acceleration data with threshold values, determines slope and horizontal position duration, and generates a fall detection signal, which is transmitted via a radio communication module, including features like capacitive sensors for foot presence detection and geolocation, to improve fall detection reliability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional accelerometer and gyroscope data are used for fall detection, then the system can detect falls, but the reliability of fall detection is insufficient
Solution Approach 1:
The patent changes the parameters used for fall detection from traditional acceleration and gyroscope data to the slope of the acceleration curve along the vertical axis and the duration of horizontal position. This parameter transformation enables more reliable fall detection by capturing the characteristic dynamics of fall events rather than relying on threshold-based acceleration detection alone
Solution Approach 2:
The patent segments the fall detection process into distinct phases: an impact phase identified by acceleration threshold exceedance, followed by a horizontal position phase identified by the slope and duration parameters. This segmentation allows the system to analyze different aspects of the fall event separately, improving overall detection reliability
2Measurement precision
If multiple parameters and time windows are analyzed for fall detection, then detection accuracy improves, but system complexity increases
Solution Approach 1:
The patent defines predetermined time windows (first given duration before impact, second given duration after impact) in advance. This preliminary structuring of the analysis framework allows the system to process data within fixed temporal boundaries, reducing the computational complexity of real-time analysis while maintaining high detection accuracy
Solution Approach 2:
The patent applies different analysis methods to different time periods: before impact (baseline establishment), at impact (threshold detection), and after impact (slope and duration analysis). This localized approach to data processing optimizes computational resources by applying appropriate analysis techniques only where needed, rather than uniformly processing all data
Data Source
Figure 1
Figure 2~3
Figure 4~6
AI summary
The invention relates to a boot (1) or a pair of boots. The or each boot comprises an accelerometer (2) and a data processing unit configured to compare the acceleration measured by the accelerometer with a threshold value and, depending on at least the result of the comparison, to start the determination of the following parameters: the gradient of the curve, preferably filtered, of the acceleration measured along the Z axis over time, and the horizontal position duration for which the acceleration or variation in the acceleration along one axis or several of the axes of the accelerometer remains below a threshold value. A fall is detected at least depending on said gradient and on said horizontal position duration. If a fall is detected, a corresponding fall detection signal is generated. A radio communication module transmits said fall detection signal to a remote receiver. The invention also relates to a corresponding fall detection method.