CADA tracker · source extraction
ANNEX I
GRAND CHALLENGES
1. Grand Challenge 1: Environmental sustainability, performance and security of the
Union’s data centres
Testing and deploying technologies for data centres across the Union to surpass state-of-the-
art energy-efficiency and resource efficiency.
This includes achieving lower Power Usage Effectiveness (PUE) and enabling significantly
higher server utilisation rates. Examples include:
(1) Lowering average Power Usage Effectiveness: improving the environmental
sustainability and performance of the Union's cloud and edge data centres to an
average Power Usage Effectiveness (PUE) of 1.15 across the Union. The main focal
areas include enabling the development of:
(a) advanced data centre energy efficiency technologies such as cooling, waste
heat recovery;
(b) quantum computing technologies for cloud and compute infrastructure
operations;
(c) grid integration and advanced energy management systems;
(d) pilot lines for the validation of next-generation energy-efficient technologies at
operational scale.
(2) Raising average server utilisation rates of data centres: raising average server
utilisation rates across the Union’s data centres towards 50%, by integrating for
example, AI-powered technologies for dynamic server utilisation management,
runtime workload management and scheduling or for balancing utilisation, energy
cost, thermal constraints, and latency requirements.
(3) Enhancing the security and resilience of data centres: enhancing the security and
resilience of data centres’ value chain and supply by integrating semiconductor
technologies and quantum technologies designed and manufactured in the Union, and
by improving their resistance to physical and cybersecurity threats, including
targeted attacks.
2. Grand Challenge 2: Cloud stacks
Building end-to-end hardware and software cloud stacks, including AI tools, infrastructure,
services and management layers to bridge the Union's critical capacity gaps.
This includes building AI servers powered by semiconductors and quantum technologies
designed and manufactured in the Union for distributed and decentralised cloud and edge
computing for AI.
Pilot programmes could help demonstrate the capabilities of the European open cloud stacks
in strategically important sectors.
3. Grand Challenge 3: Frontier AI
Developing the next generation of multimodal frontier AI models and systems and pioneering
novel capabilities.
The focus will be on the architectural design and development of next-generation multimodal
models and systems that push the boundaries of current algorithmic capabilities for achieving
superior performance in advanced reasoning, cross-modal understanding and agentic
capabilities; investigating novel approaches to model efficiency, cognitive modelling, and
alternative computational structures, etc.
The potential applications could include foundational science such as scientific discovery and
complex data interpretation, and the development of world models for improved
reasoning, automated management simulation and planning.
4. Grand Challenge 4: Physical AI
Developing advanced physical AI models and systems that operate autonomously and safely
for delivering robust, manipulation and navigation in unstructured environments.
The focus will be on co-designing software and its underlying hardware architectures and on
combining frontier AI techniques with world models supporting physical reasoning for
delivering robust manipulation, navigation, and interaction capabilities with minimal human
supervision.
The potential applications could include autonomous robots, industrial systems and
drones operating in dynamic real-world environments.
5. Grand Challenge 5: Industrial AI
Accelerate the development and deployment of European industrial AI across the Union’s
strategic sectors.
The focus will be on developing European industrial AI models and systems capable of
serving high-value industrial applications. Such models and systems should be adaptable to
sector-specific use cases and enable secure deployment.
The initiatives launched under this grand challenge should rely on specialised computing
resources and testing facilities necessary to validate AI systems in real-world environments
before supporting their large-scale deployment and uptake, including at regional and local
level.
In the automotive sector, those initiatives may facilitate the development and deployment of
innovative software platforms and AI models for automated driving, while in manufacturing,
they may enable the creation of specialised models that optimise production processes. Other
strategic sectors that could benefit from industrial AI may include healthcare, energy, agri-
food and defence.
6. Grand Challenge 6: Cooperative European Industrial Models
Developing cooperative European industrial AI models and systems for strategic sectors by
enabling collaboration at European industrial scale without exposing commercially sensitive
data between participants.
The focus will be on advanced confidentiality-preserving technologies. Those mechanisms
include federated and distributed training approaches where algorithms are brought to the data
rather than data being transferred centrally; secure execution environments, encryption-based
processing, anonymisation and pseudonymisation techniques, access compartmentalisation,
and protections against the extraction of commercially sensitive information from trained
models.
Strategic sectors that could benefit from cooperative European industrial AI models and
systems may include aerospace, pharmaceutics, cybersecurity, mobility, autonomous vehicles
and drones, energy and defence.
7. Grand Challenge 7: AI Agents Platform
Developing a European AI agent orchestration framework, providing the essential
middleware for the resilient and secure deployment of autonomous agents at scale.
The focus will be on (i) exploring innovative technological paradigms that enable multiple AI
agents to collaborate effectively, surpassing the capabilities of standalone systems while
maintaining rigorous security standards; and (ii) on the creation of resilient, cloud-based open
platforms dedicated to the large-scale management of AI agents.
The potential applications could include healthcare (such as clinical decision support and
research coordination), cybersecurity (such as threat detection and response), as well as
foundational science.
8. Grand Challenge 8: Public Sector AI
Developing AI models and systems, based on high-quality data from the public sector
targeting critical domains (such as healthcare, public administration, law and crisis
management as well as public services)
The focus will be on public service solutions that are expected to have a high positive impact
on the most critical public services and are shared across different levels of public sector
organisations.
One target will be to enable data sharing and frontier model development across national
public services to increase the impact on the overall Union’s public sector, including also in
areas handling sensitive data. Privacy-preserving frameworks, (such as federated learning and
high-fidelity synthetic data generation), that make it possible to train of models without
compromising the confidentiality of underlying datasets, and measures to accelerate the
broad uptake of those modems, including at regional and local level, will also help achieve
this target.