Desenvolvimento de sensor para monitoramento de queda em idosos

Abstract

The growth of the elderly population and the increased longevity, imply the need for studies that are concerned with healthy aging and quality of life for the elderly people. Since falls are the main cause of injury in the elderly people, and understanding that fall detection devices today still need improvement, the aim of this study is to develop a sensor to monitor falls in the elderly people, through the association of two sensors – accelerometer and gyroscope, connected to a microcontroller board, Arduino Nano 33. The fall and non-fall classification will be based on supervised machine learning concepts, through the Edge Impulse platform. For this study, a Systematic Literature Review was carried out to understand existing works in the area and their limitations, and then, a model based on machine learning and neural networks has been developed. Thus, the aim is to promote greater autonomy for the elderly public, since the device will have the objective of identifying possible falls that occur during the period that the elderly person is alone.