Artificial Intelligent Based Fall Detection System for Elderly People Using IoT

In general, one of the serious problems faced by elderly people is falling. Sometimes, this fall, not a little can cause death. if a person falls and does not get help within one hour, then the impact that must be borne will be felt up to 6 months later. Therefore, Fall detection for elderly people is a crucial problem which requires the development of modern technology that is easy and practical to use. Besides, the use of the devices do not limit and interfere with the activities of the elderly. This paper proposes a fall detection device which is able to monitor and inform all activities of the elderly people, especially some dominant events that have the impact of falling by utilizing IoT-based technology. It used two sensor to detect falling event including gyroscope and sound. The signals sent by the two sensors are then processed by the microprocessor using the fuzzy PSO algorithm to identify and distinguish between ordinary activities and falling events. PSO is used to optimized the membership functions of the fuzzy in order to improve fuzzy performance in identifying falling event. If the results state that a fall occurs, a notification will be sent to the medical officer at that place via a wi-fi network. To test the reliability of the device that have been made, two performance indices are measured, namely sensitivity and specificity. The experiment results showed that the sensitivity and the specificity of the device were 100% and 100%, respectively when it was located on the chest and the abdomen.