Low-Power Fall Detector Using Triaxial Accelerometry and Barometric Pressure Sensing

Falls are the number one cause of injuries in the elderly. A wearable fall detector can automatically detect the occurrence of a fall and alert a caregiver or a medical rescue group for immediate assistance, mitigating fall-related injuries. However, most studies on fall detection to date have focus...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on industrial informatics 2016-12, Vol.12 (6), p.2302-2311
Main Authors: Changhong Wang, Wei Lu, Narayanan, Michael R., Chang, David Chan Wei, Lord, Stephen R., Redmond, Stephen J., Lovell, Nigel H.
Format: Article
Language:English
Subjects:
Acceleration
Accelerometers
Atmospheric pressure
barometric pressure
Batteries
Detectors
Energy consumption
energy efficient
Fall detection
Firmware
Hardware
Injuries
low power
Microcontrollers
Power consumption
Power efficiency
Power management
Power measurement
Rechargeable batteries
Sensors
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Falls are the number one cause of injuries in the elderly. A wearable fall detector can automatically detect the occurrence of a fall and alert a caregiver or a medical rescue group for immediate assistance, mitigating fall-related injuries. However, most studies on fall detection to date have focused on the accuracy of detection while neglecting power efficiency and battery life, and hence the developed fall detectors usually cannot operate for a long period (a year or more) without recharging or replacing their batteries. This paper presents a low-power fall detector that utilizes triaxial accelerometry and barometric pressure sensing. This fall detector reduces its power consumption through both hardware- and firmware-based approaches. This study also incorporates several human trials to develop and evaluate the device, including simulated falls and activities of daily living. A benchtop power measurement test is also conducted to estimate the battery life with data from a one-week free-living trial. These experiments show that the fall detector achieves high sensitivity (97.5% and 93.0%) and specificity (93.2% and 87.3%) on training and testing datasets, while providing an estimated battery life of 664.9 days.
ISSN:1551-3203
1941-0050
DOI:10.1109/TII.2016.2587761