• Informações do Trabalho

Titulo An Architecture for Detecting Fall Events in the Home of Elderly People Subtítulo Autor GABRIEL DI IORIO SILVA Orientador VICTOR STROELE DE ANDRADE MENEZES Resumo With the concept of Internet of Things becoming more popular year after year, it's not unusual to identify sensors in a large variety of objects in homes. Technologies to reduce the scale of sensors made it possible to transform a home into a smart home and help people with special needs. In this way, digital data is generated massively and numerous studies and outcomes are possible with these inputs, such as event detection and event prediction. Therefore, this work proposes an architecture based on Fog-Cloud with theories from Lambda paradigm in order to predict events of falling for elderly people. To achieve this goal, experiments were made with data collected from a previous work about falling detection already tested and validated. Thus, a step further is done through the analysis of this information to comprehend what are the bests results possible considering sensitiveness and time-window to act when a prediction of falling is performed. Palavras-Chave Time Series, Event Forecasting, Human Activity, Ambient Assisted Living, Elderly Falling.