Project Characteristics

Digital Twin Platform for Desalination Plants project

The Digital Twin Platform for Desalination Plants project, led by Tedagua in collaboration with Grant Thornton, Universidad Pablo de Olavide, Tecnalia Research and Innovation, and Quantia, with file number 20215/C005/00152806, has been subsidized by Red.es through the 2021 Call for Aid for R&D projects in Artificial Intelligence and other digital technologies and their integration into value chains, co-financed by NextGenerationEU funds under the Recovery, Transformation, and Resilience Plan.

OBJECTIVES

The main objective of the project is to develop a prototype (a standardized set of devices, concepts, methodologies, criteria, and activities) that optimizes the efficiency of reverse osmosis desalination plants, incorporating cybersecurity as a pillar in their operation to enhance resilience.

To achieve this, a technological platform will be designed and implemented to support the capture and processing of real-time information from desalination plants, on which applications that are currently unfeasible with available systems, mainly SCADA or CMMS, will be developed.

These new solutions will focus on the following main areas:

1. Application of Artificial Intelligence: machine learning, deep learning, neural networks, Internet of Things (IoT), massive data and information processing technologies (Open/Linked/Big Data), high-performance computing, cloud computing, Natural Language Processing (NLP), cybersecurity, biometrics and digital identity, and virtual and augmented reality.
2. Detecting improvement points in the operation of Desalination Plants through in-depth analysis of the processes involved, from an environmental sustainability perspective (especially optimizing energy consumption) and their cybersecurity.
3. Developing a philosophy of continuous improvement in processes and operations throughout the lifecycle of desalination plants and the production of desalinated water by detecting non-obvious patterns.

To this end, a Digital Twin platform for the desalination plant will be built from a 3D model of the facilities and their main components, which can be enriched with information obtained from sensors in the plant. This Digital Twin will become the knowledge base on which to design and implement AI algorithms that generate simulation models to achieve optimization and obtain patterns and indicators that improve efficiency in operation and maintenance tasks, both in normal operation and in case of anomalies or incidents.

This entire technological base must be integrated into the processes of the seawater desalination lifecycle, in the context of the electrical efficiency of desalination plants by relevant technical personnel, so mechanisms for modeling and interaction with these processes must be established. Therefore, this platform will become the key axis to ensure that desalination plants are operated efficiently and guarantee the highest level of service within the water network.

RESULTS OBTAINED

The project is divided into 5 work packages, whose results are shown below:

Work Package No. 1: Automated informational system based on reality capture

• Result: The deployed platform incorporates advanced analysis capabilities and artificial intelligence functions, allowing for deep and continuous evaluation of operational data.
• Impact: The integration of AI into operational processes provides valuable insights and accurate predictions, facilitating early identification of potential failures and continuous optimization of processes. This results in more efficient and safer operation of desalination plants.

Work Package No. 2: Automatic development of the Digital Twin model

• Results: Implementation of Models in Azure Machine Learning, Development of a Natural Language Processing Model, Analysis of Variables in the Desalination Process.
• Impact: The analysis and optimization of process variables contribute to improving the operational efficiency and sustainability of desalination plants.

Work Package No. 3: Decentralized, scalable, and resilient data model

• Decentralization: Elimination of single points of failure by distributing data across multiple nodes or servers, improving availability and accessibility.
• Scalability: The model's ability to handle an increasing volume of data without affecting performance, allowing the incorporation of new sources and IoT devices.
• Resilience: Fault tolerance and automatic recovery capability in the event of interruptions, ensuring service continuity without data loss.
• Impact: The implementation of a decentralized, scalable, and resilient data model optimizes operation and predictive maintenance, improves data security and efficiency, reduces costs, and facilitates digital transformation in the water industry.

Work Package No. 4: Pattern detection and definition system for intelligent and unattended simulations on the Digital Twin

• Results and impact: The successful implementation of the platform integrated all developed components, significantly improving the representation and management of physical systems, optimizing the operation of desalination plants.
• Design and Implementation of Interfaces: FIGMA was used to design collaborative and high-quality interfaces, with continuous user feedback, resulting in a notable improvement in the user experience and satisfaction.
• Continuous Integration: The use of GitHub for code management and CI/CD process configuration ensured rapid and error-free implementation on Azure. This improved the efficiency in the development and deployment of the platform, guaranteeing continuous updates and optimal functionality.

Work Package No. 5: Automated process modeling and optimization system

• Results and impact: Optimization criteria were defined, and desalination models were integrated, achieving reduced energy consumption and operational costs. The proposed improvements were validated through experiments and demonstrated their effectiveness in real environments.

A 3D BIM web interface connected to a database was developed, with visualization and functionalities that facilitate real-time data management and analysis.

Finally, action plans and recommendations were generated, improving operational resilience and decision-making in desalination plants.