TETRA – the Toolbox and mEthodology for waTeR based AI projects

TETRA – the Toolbox and mEthodology for waTeR based AI projects

The development of modern and efficient tools for monitoring water resources is crucial for ensuring the sustainable availability of this essential resource, which is of great value to both humanity and the environment. Events like the fish die-off in the Oder River underscore the pressing need for improved river protection. The TETRA project aims to enable and accelerate the use of artificial intelligence (AI) in water management. Additionally, the bilateral collaboration of both German and French companies fosters the development of a shared European ecosystem for AI applications in the water sector.

The project’s goal is to develop and provide tools and methods that enable the successful implementation of AI projects in the field of water management. These will be made publicly accessible to both German and French stakeholders to facilitate and promote collaboration with a common toolkit and approach for AI projects. The evaluation of the tools and methods will be based on two use cases: monitoring the water quality of rivers and river restoration.

The TETRA methodology is based on the already available PAISE process model, which was specifically developed for the integration of AI methods into industrial processes and is being adapted for application in water management. Within the scope of PAISE, a toolbox with specific AI tools will be developed. Several applications will be utilized in this context.
The FROST server, provided by Fraunhofer IOSB, is an open-source implementation of the OGC SensorThingsAPI that manages and stores sensor data needed for analysis by AI algorithms. FROST will be extended to FROST-AI within the project to meet specific AI integration requirements. The developed algorithms will be integrated into PERMA, an open-source software developed by Fraunhofer IOSB that enables the management and parameterization of algorithms.
GLUON, a tool for creating and managing ontologies, enables the integration of expert knowledge into AI algorithms. If facilitates semantic search and knowledge modeling in water management.
Edge AI analysis employs technologies to analyze data directly on sensors (edge computing) to reduce latency and ensure data privacy.
Godot Search is a semantic search module that can be used to understand user queries through ontologies to find relevant information more efficiently, and will be improved throughout the project.
Case-Based Reasoning (CBR) for river restoration utilizes case studies and expert knowledge to improve restoration measures in water management.

The ontology, knowledge base, and data from the FROST server will be made available in collaboration with all partners in the TETRA Showcase via WebGenesis (IOSB) in a web portal for demonstration purposes. Through TETRA, a framework for the integration of the AI algorithms and standardized data storage will be created, forming a starting point for a shared ecosystem dedicated to AI-based water projects.

This research has received funding from the BMBF’s (Bundesministerium für Bildung und Forschung) directive on the funding of Franco-German projects on the topic of artificial intelligence, Federal Gazette of 20^th June 2022.