s-X-AIPI gathering in Valladolid, Spain

May 30

The third week of May marked an exciting event, the Month 24 Internal Periodic Meeting of the s-X-AIPI consortium in Valladolid, Spain. On May 14-15, s-X-AIPI experts gathered at the CARTIF premises in the sunny city of Valladolid. Over these two days, project partners convened to review activities from the past six months and strategize for the project’s last year. The meeting provided an invaluable opportunity for face-to-face discussions of each partner’s scientific progress and accomplishments. Beyond, professional matters the event was complemented by a guided visit to the historical heart of Valladolid.

Key Highlights and Conclusions

The meeting yielded several significant conclusions. Here are some key takeaways:

AI Maturity Model

Defined Methodology. A comprehensive methodology for assessing AI maturity levels across strategic, tactical, and operational dimensions has been defined.
Development Progress. Significant progress has been accomplished in developing maturity models and the AI architecture.

Self-X AI Models and Techniques

Model Review. Various procedures focusing on models for monitoring and meta-model monitoring have been reviewed. These efforts aim to identify patterns and ensure the optimal performance of the AI pipeline. By aggregating and analyzing metadata from different AI models, meta-model monitoring helps detect broader trends and issues, utilizing historical data to pinpoint anomalies.
Validation Activities. Notable progress has been made in developing and validating the self-X AI models of the AI data pipelines across the different use cases. Detailed validation activities, particularly regarding the Human-in-the-Loop (HITL) components, were showcased.

Autonomic Managers for Data in Motion and Human Support

Initial Developments. The initial version of autonomic managers has been developed, focusing on component interoperability to overcome technical challenges.
Future Activities. Upcoming efforts will concentrate on enhancing modular and scalable solutions for edge deployment and near real-time operations where required. This transition will be seamless for all use cases.
Human Interaction Interfaces. Continued development aims to optimize solutions, ensure effective human interaction interfaces, and finalize the corresponding rule-based engines and their optimization.

The close collaboration among partners will continue online until the next face-to-face meeting scheduled for this October in Athens.