WASP to Establish New AI Research Arena

Advanced, large‑scale AI is essential for addressing the major challenges many industries face today: increasing data volumes, more complex systems, and greater demands on computing resources. To meet these challenges, WASP is establishing a new research arena, AI Training and Inference Compute at Scale (WARA AI‑TRICS). The initiative is driven by Sweden’s largest national research program, the Wallenberg AI, Autonomous Systems and Software Program (WASP).

Within WARA AI‑TRICS, researchers and companies will collaborate on the most pressing AI challenges facing Swedish industry. Three key areas have been identified as particularly important:

• AI foundation models for industry and society
• AI at Scale, applying AI to massive industrial datasets
• New methods, such as physics‑informed AI

Building expertise across the entire AI lifecycle

The arena will bring researchers and industry into close collaboration, focusing on real‑world data, systems, and challenges where both advanced research and powerful computing capacity are required to drive progress. The capacity Swedish universities have built in recent years through major investments in AI‑focused supercomputers provides strong conditions for establishing this arena now.

“By bringing together leading researchers and industrial expertise, we are building competence across the entire AI lifecycle – from data curation and model design to training, validation, and deployment,” says Salla Franzén, who has co-developed the proposal for WARA AI‑TRICS.

Results from the collaborative projects will be shared to inspire continued research by highlighting open scientific problems and by making benchmarking datasets and open‑source code available whenever possible.

Multiple companies and researchers already onboard

Salla Franzén (Navigare Ventures AB and AI4S AB) and Beatriz Grafulla (Ericsson AB) are jointly leading the program proposal process for WARA AI‑TRICS, in collaboration with partner companies and researchers. Ahead of the launch, several companies and researchers from WASP’s partner universities in Sweden have already joined the arena. Other major companies are expected to join shortly.

“Companies contribute access to large‑scale digital infrastructure, such as extensive datasets, and they also bring valuable experience and expertise. We want a broad range of industries engaged in the arena to enable the development of different types of AI models and digital infrastructures,” explains Beatriz Grafulla.

In addition to project collaborations with partner companies and participating researchers, the arena will also collaborate with WASP’s other research arenas.