Exploring FIWARE-based Data Model for Data Sharing Platforms

This is the Author Accepted Manuscript (AAM) of a chapter published by Springer. The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-031-96235-6_10 

Abstract. In recent years, the volume of data collected from devices and wearable sensors has significantly increased, with a tenfold growth annually. As a result, data models have become essential tools for organizing data, defining relationships between elements and standardizing data, especially in areas where wearable sensors are increasingly prevalent, like mobility or healthcare. In this study, we present an extensible and robust data reference model, inspired on RAISE pilot application domains of health, environment and mobility. The model was influenced by the data needs and challenges observed within its pilots and it leverages the FIWARE Smart Data Models initiative to ensure interoperability and scalability. Three different models were developed, one for each pilot, designed to accommodate evolving data requirements and adapt to each pilot’s implementation needs. Additionally, Fitbit measurements were incorporated to capture health data collected from Fitbit sensors, expanding the model's capability to handle wearable-generated data. In cases where attributes could not be aligned with FIWARE Smart Data Models, the Open mHealth standard was utilized to ensure comprehensive data representation. Descriptive statistics were also employed to analyze main features of the data and capture metadata across all types of pilot variables. The proposed model provides a shared and common schema capable of representing and handling data from diverse sources, devices or platforms, efficiently describing the collected data to produce meaningful metadata.