Description and objectives

The aim of the project, which falls under Theme 3, is to identify and implement a solution for the validation of real-time meteorological and hydrological data using Artificial Intelligence and Machine Learning (AI/ML) techniques.

The interpretation of the data collected by the sensors in question can present various difficulties, due to the disruption of readings by transient or persistent events, which can be traced back to various components in the supply chain.

The aim of the project is to develop and integrate techniques capable of verifying and validating these readings and subsequently implementing a service that provides a degree of confidence in the data. This is in line with the RETURN project’s objective of developing new methodologies and technologies for monitoring and promoting a more efficient and sustainable use of data products and services.

The data evaluation process can be conducted using various types of information: geographical aspects, comparisons with historical series, seasonality and other metrics can constitute a solid knowledge base useful for characterising the data. Another indicator is provided by the operational field of the sensors themselves: knowledge of these values reflects a stringent constraint on readings and allows the validity of a piece of data to be identified in the first instance. AI/ML techniques provide fundamental support in this area, allowing this knowledge to be structured into dedicated models, aimed at accompanying individual readings with a measure of reliability, appropriately derived.