Survey system for life styles of the elderly and to identify the consumption of food and drink

Abstract

From the viewpoint of welfare costs, it is necessary to survey whether older people eat, drink and live their daily lives in safety. It is especially desired that an older person drinks green tea and/or coffee routinely and autonomously takes meals, namely breakfast, lunch and supper. In addition, it is better for an older person to eat fruits such as oranges and apples to prevent colds, Alzheimers disease and other illnesses. In this study, a system to recognize the eating of a meal, the drinking of a beverage and the eating of a fruit was constructed using odor sensors. Three metal oxide odor sensors were adopted for this system. The system can identify behavior by analyzing characteristics of odor sensors. As for the results, the system recognizes this behavior successfully and also detects the level of fatigue of an older person. A high correlation coefficient (over 0.85) was obtained between the degree of fatigue due to bodily sensations and the output of the sensor. The sensor was highly sensitivity to hydrogen gas. When humans are fatigued or stressed, their exhaled breath contains hydrogen at high concentrations. Discriminating analysis was used to identify the kind of fruits eaten and teeth brushing behavior from odor sensor outputs.