ICE2025 - meet HaDEA-managed projects working on AI enabled services for the Cloud edge continuum

The 2025 International Conference on Engineering, Technology and Innovation (ICE2025) was an inspiring event, where conference participants presented and discussed the latest trends in digital transformation. During the conference, 14 thematic special sessions were organised, where synergies between ongoing projects could be explored and future challenges had also been discussed. One of the special sessions (SS14 - Special Session: Cognitive Computing Continuum) was chaired by HaDEA-managed project ENACT with the participation of six other HaDEA-managed, EU-funded projects, selected under 2023 Horizon Europe call on Cognitive Computing Continuum: Intelligence and automation for more efficient data processing. 

While all these projects are complementary in their research objectives, it is crucial to take a step back time to time and explore overlapping and distinct research directions that can shape the future technological perspectives. Towards this aim, after more than 18 months in their implementation, our projects during the special session have concluded the followings:

• Defining strategies early on, for conceptualising and managing resources within the Cognitive Computing Continuum, is a key for delivering successful solutions for the end users. For example, considering resources as specific autonomous collectives or dynamic swarms have different runtime and deployment implications.
• Certain use cases that may work with sensitive data, like eHealth or manufacturing, pose constraints on where the data and service components need to be stored and deployed. For example, predefined geographical locations will pose various challenges for dynamic resource management.  
• One of the key challenges for developing these technologies is the lack of real-world, open-access datasets that can be used for training AI-based orchestration models.
• For future interoperability and sustainability of the services, using open-source and standard modelling frameworks like TOSCA (Topology and Orchestration Specification for Cloud Applications) is crucial. Without such standard representation, cloud edge resource management across different use cases becomes very challenging. Furthermore, EU projects can play a crucial role in getting involved with the work of technical committees, shaping future standardisation efforts.
• Incorporating support for various open-source data spaces and federated data infrastructures in the service domain is a key for future success.

The projects:

• COGNETS aims to revolutionise the Intelligent Infrastructure Management by introducing a scalable and interoperable distributed Middleware Framework for autonomous IoT-to-Cloud computing, sustainable during and after the project.
• EMPYREAN aims to establish Associations, a hyper-distributed computing paradigm, by leveraging collaborative, heterogeneous IoT devices and federated computing and storage resources that may belong to different users.
• ENACT is developing a Cognitive Computing Continuum that can address the needs for optimal resource management and dynamic scaling. By harnessing the power of AI-powered Graph Neural Networks and Deep Reinforcement Learning agents, the project will suggest optimal deployment configurations for hyper-distributed applications. The outcomes of these research and development activities will enhance European businesses competencies and upgrade IT services.
• HYPERAI works with smart virtual computing entities (nodes) that come from a variety of infrastructures that span all three of the so-called computing continuum's layers: the Cloud, the Edge, and IoT.
• INTEND aims to deliver eleven novel software tools, which integrate into an INTEND toolbox. The approach and the tools will be tested and validated on five vertical domains, to achieve the novel intent-based data operation for video streaming pipelines, machine data platform, 5G data infrastructure, urban data space, and robotic AI applications.
• MYRTUS aims at unlocking the new living dimension of Cyber-Physical systems (CPS), embracing the principles of the TransContinuum Initiative, integrating edge, fog and cloud computing platforms.
• SWARMCHESTRATE works on the self-organisation of so-called swarms of interdependent units in a dynamic orchestration space. The swarms will be managed by decentralised orchestration agents with blockchain-based trusted solutions, state-of-the-art cryptographic algorithms, and privacy-preserving data analytics.

Background

Cloud computing is one of the core components for shaping Europe’s digital future. It is foreseen that the European Commission will propose the Cloud and AI Development Act in 2025, with the potential to triple the data centre capacity of the EU, in order to close the capacity gap compared to US and China. At the same time, it is also forecasted that while 20% of the future computation will be done in data centers, i.e. in the cloud environment, around 80% of the computation will be executed on smart devices closer to the user, i.e. edge computing. Such cloud edge computation will also involve training, fine-tuning, and running various AI models in a heterogenous and dynamic environment. This will require technological solutions, where the system is able to sense its environment, optimise it in real time and learn to adapt dynamically its computational loads, i.e. coined as Cognitive Computing Continuum.

Horizon Europe is the research and innovation programme of the EU for the period 2021-2027.