Developing an AI capability maturity model for Transmission system operators Bridging the gap between science and practice
Summary
The growing complexities in the energy sector, particularly for Transmission System Operators (TSOs), demand innovative solutions to address operational challenges. Artificial Intelligence (AI) offers significant potential to reduce reliance on expanding human resources for grid operation and maintenance. However, effectively integrating AI requires organizational adaptation, guided by robust enterprise architecture (EA). Enterprise architects play a critical role in designing frameworks and structures that enable TSOs to harness AI’s potential. Despite its importance, executing EA for AI integration is hindered by a lack of knowledge about the capabilities required for AI. This thesis addresses this gap by developing an AI capability maturity model tailored to critical infrastructure suppliers such as TSOs. The research outputs include an AI capability map, a maturity model, and exemplary action steps. The study employs the Design Science Research (DSR) methodology, emphasizing practical solutions to real-world challenges. Combining a literature review with expert interviews at TenneT, Alliander, ENTSO-e and Microsoft, the research identifies key capabilities and their corresponding maturity levels. Generative AI is utilized to define TSO-specific criteria for each capability. The model’s validity is established through two iterative expert evaluation cycles involving enterprise architects and managers at TenneT. The final model features 22 enterprise capabilities categorized into strategic, supporting, and operational capabilities. To demonstrate its practical value, the maturity model was applied to TenneT, assessing its AI maturity and outlining actionable steps for improvement. Exemplary steps are defining an AI vision, organizing an AI enablement hub and the creation of an AI guild. Additionally, some smaller improvements were identified, such as creating AI partnerships and assigning resources to the improvement of IT operations for AI.