M3 Room

Artificial Intelligence

M3 supports research on cross-cutting AI topics with significance for a larger number of related research rooms, and it fosters collaboration with other AI initiatives.

Key Questions

Integrating touch and olfaction is a key challenge.

Solutions will be explored to optimize networks and computing infrastructures dynamically.

Focus will be on certification and explainability to build user trust.

Research will focus on combining data-driven and symbolic models to improve output quality and reliability.

Approach

  • Multimodal AI: Advancing multimodal sensing, particularly for touch and olfaction.
  • AI for distributed networks: Optimizing communication protocols and network architectures.
  • Integrating symbolic and data-driven AI models: Enhancing AI outputs with symbolic domain knowledge.
  • Verification and certification: Ensuring the safety and reliability of AI-based decisions.

Expected Results

  • Multimodal models and datasets.
  • Validated foundational models and AI solutions for communication tasks.
  • Methods for AI verification and building trust.