DIGITAL TWIN CHS

The Challenge

The main technological challenges consist of integrating into the platform to be developed the various sources of data and information with which the CHS works (sensor readings and calculations carried out in real time and laboratory analyses related to the state of water bodies and infrastructures and water quality, meteorological information, information from the Water Framework Directive and information from the Hydrological Plans), and the new predictive models to be developed (Meteorological Prediction Model, Hydrological Prediction Model, Agronomic Prediction Model, Probabilistic Prediction Model for System Management and Hydroeconomic Model), and to integrate this platform with the CHS's IT infrastructures, with its user management system, and with the CHS's cybersecurity policies.

The main technical challenge is to apply Artificial Intelligence (AI) techniques to existing data, build new predictive models based on AI and their continuous integration into the platform to be developed, in order to improve the daily management of the basin.

An additional important technical-technological challenge to be highlighted is to develop an integrated platform for water management that can be easily transferred to other basins.

The Proposed Solution

Based on the aforementioned objectives and technical-technological challenges, a platform has been developed mainly based on the use of software development tools, data management and development of open source and free use predictive models. Regarding predictive models, deterministic models have been combined with Machine Learning (ML) models, thus taking full advantage of the benefits of integrating the knowledge implicit in deterministic models with the enormous capacity for discovery and generalisation of ML models. Likewise, the most recent agile software development methodologies have been followed, based on the use of continuous integration and containerisation technologies, which has facilitated the deployment of the solution developed in the CHS's IT infrastructures and will enable its integration in other basins.

The Result

As a result of the project, it can be stated that a Web platform has been developed for the water management of the Segura River Basin, which meets the objectives and challenges set. The platform developed is operational and accessible to CHS users, and evolves as a dynamic system, as it incorporates new data generated by both external data sources and the new predictive models developed every day. Likewise, the AI-based predictive models evolve dynamically as the historical data with which they have been trained are renewed.