Refining the TAI Approach: FRAUNHOFER IIS and ARCADA Collaborate on Z-Inspection® within the MANOLO EU Project

A successful collaboration between FRAUNHOFER IIS and ARCADA University of Applied Sciences within the EU-funded MANOLO project has led to a  refinement of the project’s approach to Trustworthy AI (TAI). Through the application of the Z-Inspection® process, ARCADA supported partners in Work Package 3 (WP3)—including Fraunhofer IIS—giving them practical guidelines and providing support with this process.

A unique challenge in this context was that the AI components of WP3 are developed independently from any specific use case as they should be general purpose AI optimization tools.

Based on the insights gained by applying the process of Z-Inspection® for WP3 general AI Components, a refined approach was defined to perform a Trustworthy AI Co-design of general purpose AI optimization tools which is based on a new concept of a so called “triangulation” which basically implies a bottom up combined with a top down approach. In a nutshell, for each general purpose, AI is considered both for specific MANOLO use cases, as defined in WP6, and in a more general setting.

This work is particularly important as it implies an active collaboration of several MANOLO partners in different WPs (WP3, WP5 and WP6).

This advancement not only strengthens MANOLO’s internal methodology but also contributes to the broader EU AI landscape by promoting transparent, ethical, and context-aware AI development.