Onisciente: um modelo de geração de contexto baseado em RFID e sensores

Abstract

In determined situations is interesting to have the capacity of know in which situation something or someone is. Have this information can bring advantages for whom desires to care or control some situation. In this scope the present work have as main finality to present the Onisciente, a model that uses the information from a group of sensors that are monitoring a same element to determine its context and obtain knowledge about its condition. To accomplish that, the Onisciente makes the identification of those entities using a RFID reader with some sensors. Those sensors acquire data related with the environment where the entities are placed in, or data directly from the entities, making some treatment and processing it, generating the context. In Onisciente, a context represents an image of a moment in the time, in which, when analyzed, allow us to verify facts that were occurring and make some decisions. Decisions made by the Onisciente involves classify the generated contexts and, according with the obtained classification, inform unfavorable situations of the monitored entities. To determine a context and realize the classification of what it means, where used beyond the RFID technology and sensor, ontology, context-aware computing and context modeling. The data collection is made using an Arduino hardware with several sensors, and transmitted, with the support of a mobile Android device, to the Oniscient server. The server applies the received data in an ontology that process then and generate the contexts. Those generated contexts can be visualized using and app for mobile devices or through a web page. A series of experiments were realized to evaluate the functionality of the Onisciente. Firstly, specific tests were applied over the ontology to analyze the quality of its constitution, functionality and quality of the generated contexts. Then, use cases were elaborated to test the Onisciente’s functionality and performance. The acquired results determined that the presented model in this work managed to achieve the proposed objectives by generating accurate contexts in the end of the tests, having as highlights, beyond the capacity of realize the same tasks presented in the studied related works, to present resources to integrates with several technologies through web services, dispose of varied resources for the follow the obtained results and to have the capacity of use several types of RFID and sensors.