In this study, we constructed a sensor system that is inexpensive and does not restrain the care recipient while emphasizing QOL. It is a bed leaving behavior monitoring system that can predict when the care recipient will leave the bed by analyzing the output of each sensor.
1) The judgment rate is improved by the machine learning method.
2) Easy to install and inexpensive.
3) The state of the care recipient is detected in three patterns: “sleeping”, “getting out of bed”, and “completely getting out of bed”. Therefore, it is possible to prevent accidents due to delays in judgment.
Pillow sensor Bolt vibration sensor Leg load sensor
Getting out of bed measurement experiment