Scientists who analyze physical phenomena typically work with data gathered from various sensors placed throughout the environment. Visualizing massive amount of data is crucial to the analysis process, but implementing a reliable and efficient visualization tool can be challenging, especially without the support of a large-scale platform.

We propose to take up this challenge with FaST (an efficient model-driven Framework for visualizing large-Scale Time series).

FaST is a model-driven framework that provides a complete solution for the storage, the querying and the visualization of time series in a big data context. It offers a dedicated language for data scientists to efficiently specify the solution's architecture and the data it has to handle. The deployment process is streamlined through code generation and server-side dockerization. The generated tool itself is optimized for performance through ad hoc optimizations. On the server-side, these optimizations involve pre-computation of views based on the Min-Max principle. On the client-side, they come from the anticipation of queries related to the data navigation abilities of the generated tool.

Our current work focuses on a data collection system that involves a USV (Unmanned Surface Vehicle) equipped with various sensors, which is designed to cater to a group of users seeking to monitor the sensors' data. To achieve this, we propose a generic model-driven framework based on the publish-subscribe pattern that enables efficient transmission of the emitted data to the users.