The second general assembly of the AI-EFFECT project, hosted by DNV on March 19th and 20th in Oslo, focused on workshops and discussions about node use cases as part of WP1: AI-EFFECT Use Cases and Testing Methodologies Development.
Six months into the project, WP1 is defining the TEF use cases and the list of services expected from each node. The TEF Test Use Cases concentrate on testing procedures, involving end-to-end steps for verifying AI applications in a controlled environment. These use cases can include TEF Services to support AI development but extend beyond them, offering additional services and capabilities. AI-EFFECT aims to design use cases suitable for AI exploitation, with a comprehensive list of functionalities and specifications for the TEF being defined. Subsequently, the specifications for use cases demonstrating the key services offered by the AI-EFFECT TEF will be identified. Drafts of the use cases and services for all four nodes have been prepared and presented during the meeting.
In addition to node discussions, the project held three workshops: Explainability AI Tool Workshop, aimed to discuss the definition of Explainable AI and how it applies to the future TEF services and what added values it can offer to the TEF users; the Operating Principles Workshop, aimed to define operating principles to ensure that AI-EFFECT functionalities are applicable to a wide range of applications and to lay the foundations for the Ai-EFFECT replicability; and Verification Workshop, targeting the assessment of trust and risks of AI applications in critical infrastructures, such as the energy networks.
The meeting concluded with an overview of activities performed by each task, their outcomes, and upcoming activities. WP1 has been the most active technical work package, with all tasks initiated within the first six months of the project. Other tasks and work packages will commence at different stages of the project.
