Agentic AI for self-deployable 6G networks in the edge-cloud continuum

Purpose and goal

The project will create an Agentic AI system for autonomous deployment and configuration of 6G network and application software across heterogeneous edge-cloud environments. Combining AI agents, local LLMs, and digital twins, it will adapt to unknown conditions, streamline complexity, and enable scalable, secure, and sustainable deployments with improved efficiency and lower costs.

Expected effects and result

A TRL 6 demonstrator will prove autonomous, repeatable deployment in realistic network and application testbeds. The developed solution will reduce deployment time, errors, and resource waste, cutting operational costs. The project results will enhance the partners’ competitiveness, reinforce Sweden’s telecom leadership, and produce high-impact research publications, demonstrations, and future exploitation opportunities.

Planned approach and implementation

Four work packages will steer the activities related to use case refinement and digital twin creation (WP1); AI agent and knowledge base development (WP2); integration into a TRL 6 proof-of-concept (WP3), and coordination/dissemination (WP4). Activities will follow an iterative process with continuous validation in representative environments and targeted risk mitigation, ensuring the final system is scalable, robust, and industrially relevant.