SAFEXPLAIN partners spent two days in IKERLAN presenting the work carried out in the first six months of the project. Although project partners have been meeting frequently online to discuss the work in their own spheres, this Face-to-Face (F2F) allowed them to share their initial findings and work with the entire consortium. This exercise facilitated critical group brainstorming and contributed to identifying and addressing outstanding issues to ensure that the project and its partners are in sync from the very beginning of the project.
Working in the realm of Critical Autonomous AI-based Systems found in the project’s three use cases (automotive, railway and space) requires a multi-pronged approach that includes a consideration of deep learning for artificial intelligence, the explainability behind the decision-making in AI and complexity of functional safety requirements. Project partners took advantage of this F2F to obtain a deeper understanding about the different aspects of the project and to reconcile them to continue their work toward providing the ultimate goal of a novel and flexible approach to certification and the adoption of DL.based solutions by CAIS.