Inertial Fall Detection Using Impact, Immobility, and Body Orientation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for detecting falls of a person are inefficient and prone to false alarms, particularly in environments like forestry work, where the need for reliable detection of falls and immediate assistance is critical.
Innovation Solution
A method and device using an inertial sensor unit worn on the body, particularly the head, to monitor inertial variables for impact, immobility, and non-vertical position criteria, including acceleration and inclination sensors, to accurately detect falls and trigger alarms only when these criteria are met, thereby reducing false alarms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing fall detection methods are used, then fall detection can be performed, but false alarms occur frequently and reliability is reduced
Solution Approach 1:
The fall detection process is segmented into three distinct criteria that must all be met: impact criterion (detecting the fall event), immobility criterion (detecting lack of movement after fall), and nonvertical position criterion (detecting abnormal body orientation). This segmentation allows the system to verify multiple independent conditions before triggering an alarm, significantly reducing false alarms while maintaining reliable fall detection.
2Reliability
If multiple monitoring criteria are implemented, then false alarms are reduced, but device complexity increases
Solution Approach 1:
The inertial sensor unit performs multiple functions using a single device: it detects impact events, monitors immobility duration, and determines body orientation. By making the sensor unit universal and multi-functional, the system achieves high detection accuracy without proportionally increasing device complexity, as one sensor system handles all three monitoring criteria.
Solution Approach 2:
The patent combines impact detection, immobility monitoring, and position sensing into a unified fall detection algorithm that processes data from the inertial sensor unit. By merging these functions into a single integrated system with centralized evaluation logic, the complexity is managed efficiently while maintaining high reliability through multiple verification criteria.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The solution enables reliable and efficient fall detection with reduced false alarms, allowing for timely assistance to individuals who have suffered physical harm, such as during forestry work, by accurately identifying falls and ensuring prompt notification.
Implementation Method 1
The inertial sensor unit is designed to record at least one inertial variable
Data Source
AI summary
A method detects a fall of a person via at least one inertial sensor unit, which is worn by the person on a body part, in particular their head, and which is designed to record at least one inertial variable. The method includes: monitoring whether at least one first recorded inertial variable or a variable based thereon meets an impact criterion, which is characteristic for an impact of the body part; if the impact criterion is met, monitoring whether at least one second recorded inertial variable, which is recorded chronologically after the first inertial variable, or a variable based thereon meets an immobility criterion, which is characteristic for an immobility of the body part for an immobility duration, in particular of at least 30 s, and a nonvertical position criterion, in particular a horizontal position criterion, which is different from the immobility criterion and is characteristic for a nonvertical, in particular horizontal, position of the body part for a position duration, in particular of at least 30 s; and if the immobility criterion and the nonvertical position criterion are met, outputting and/or transmitting alarm information, in particular about the detected fall.