Enhancing Trust in AI Decision-Making through Semantification and the FIDES Ontology
In our latest deliverable ‘Semantic data model for integrated digital twins’, we focus on the semantification of data within the use cases that have been developed by the project’s partners, specifically within the Semantic knowledge graph for integrated digital twins.
In specific, the AI-PROFICIENT team outlines the process of ontologically expressing sensor measurements and equipment, highlighting the challenges and considerations that arose during this process. The data model resulting from this work has been implemented in the data layer for AI-PROFICIENT AI services, and this integration is discussed in the deliverable.
In addition to this technical work, FIDES ontology is introduced, an ontology that aims to improve the understanding and trustworthiness of AI systems when they support decision-making. The AI services deployed by AI-PROFICIENT are designed to support decision-making in the production process. By providing insight into how these AI systems are combined and how the decision-making process occurs, the project aims to increase trust in the decision-making services provided by AI-PROFICIENT to operators.
Overall, this deliverable provides important insights into the semantification of data and the development of AI services for decision-making. By outlining the challenges faced during this process and presenting the FIDES ontology, the document contributes to a more transparent and trustworthy approach to AI decision-making in the industrial context.