Modeling Quantified Things Using a Multi-Agent System

The Internet of Things (IoT) aims at empowering anything to connect to the Internet and to collect data about itself and the environment in which it is situated. Therefore, the IoT has been producing a wide range of data. Our aim is to deliver a way of incorporating the small data generated by each thing into larger datasets to get more meaning out of these data. A thing, equipped with sensors, can be monitored and can record specific features regarding its behavior. Then it can insert data into a cloud database in order to provide individual-and collective-level analyses. We call this process "Quantified Things," and we present it as a new area for the application of "Quantified-Self" and "Quantified-Community" that are approaches of the IoT. In this paper, we provide an overview of the Quantified Things (QT) concept and the key requirements for the creation of QT applications. In addition, we present an agent-based model as a solution to meet such requirements. To illustrate the use of this model, we derived an example from one of the Quantified Things applications: "Quantified Fruit." Using the "Quantified Fruit" concept, some fruit storage use sensors to monitor environmental conditions, such as temperature, relative humidity, lighting and some gases that may affect fruit ripening. In turn, they insert these collected data into a cloud database and consequently enable knowledge sharing. Through these collective experiences, fruit storage may provide an advanced informative perspective on fruit shelf life based on local environmental conditions.