Distributed Learning, and in particular Federated Learning (FL), is emerging as a leading paradigm within the field of Machine Learning, satisfying these two properties. Federated Learning has grown in parallel with the expansion of the cloud to the edge (CloudEdge), but interestingly, both have developed mostly independently, despite their natural parallelism. In MLEDGE (Machine Learning in the Cloud and at the Edge), IMDEA Networks, with Dr. Nikolaos Laoutaris as principal investigator, will work to reverse this trend by implementing FL as an independent but optimized cross-industry layer on top of CloudEdge, using real-world applications and data to demonstrate that this synergy can produce significant benefits for all.

The data economy is estimated to have an impact of €827 billion for the 27 EU countries by 2025 ⁽¹⁾. Therefore, the objective is to enable a thriving ecosystem of secure and efficient edge FL services capable of facilitating the use of sensitive personal and B2B data to train machine learning models (for individual end users or administratively independent organizations collaborating under different trust assumptions - from total to zero, and any level in between).

Efficiency, sustainability, and security.
As Elisa Cabana, Postdoc Researcher at IMDEA Networks, points out: “The project contributes research in areas such as federated learning services (FLaaS), cloud edge processing, efficient use of FL in hybrid clouds, and protection against attacks, protection of sensitive or confidential data exchanged, management of data portability challenges at the edge, etc.” In this context, the team will design a development framework and components to help popularize these types of services, as well as solutions against poisoning or inference attacks launched from rogue edge servers and/or “honest but curious” aggregation nodes. This includes the creation of a watermark to protect against the redistribution of data or metadata exchanged between edge servers within the FLaaS framework.

Other key aspects, as Cabana summarizes, include: “Creating a layer of economic and business logic that implements a fair distribution of costs and revenues among the parties collaborating on training ML models, and supporting DevOps (a set of practices that combine software development and IT operations, aiming to accelerate the software development lifecycle and provide continuous, high-quality delivery) and the continuous development of cloud-based machine learning services, optimizing costs through monitoring, prediction, and intelligent, energy-efficient allocation of computing jobs.” The research will also contribute to designing, implementing, and publicly deploying demonstrators that work with sensitive personal data and feed useful models in areas of the traditional and digital economies such as FinTech, identity, healthcare, transportation, access control, and more.

Technology Transfer to Society:
The project's innovation will foster favorable market conditions for the use of federated learning in the cloud and in federated data architectures, such as those defined by institutions like IDSA or Gaia-X, within an international context. This will lead to innovations of great interest for addressing significant economic, business, and social challenges related to data silos in the economy. “MLEDGE will make advanced federated learning technology accessible to more organizations and individuals, including SMEs and government agencies, and will promote the creation of sustainable businesses for all actors in the value chain (machine learning experts/providers, cloud and data service providers, traditional and digital industry, the public sector, academia, etc.),” says Nikolaos Laoutaris, Research Professor at IMDEA Networks and Principal Investigator of the project at the institute.
The project will be fundamental for the development of cloud and ML/FL infrastructures in Spain and for promoting national R&D&I. It will contribute to the Sustainable Development Goals set by the United Nations for 2030 and promote the sustainable development of efficient networks and FL solutions through practical work that can substantially and positively impact the environment.

Regarding technological solutions, the following stand out:
1. Traditional economy (construction, finance, healthcare, etc.). The use case will be developed by a company to improve processes or decision-making (for example, in real time) based on data or models from FL.
2. Digital economy. An example could be in the field of digital health, such as leveraging information from mobile devices or wearable technologies. Another would be training digital advertising models.
3. Optimization of CloudEdge infrastructures. A key functionality for MLEDGE, which will employ federated machine learning algorithms.
MLEDGE (Machine Learning in the Cloud and at the Edge) is funded (January 2023-June 2025) by the Ministry of Economic Affairs and Digital Transformation, European Union NextGeneration-EU.

(1) European data strategy for the period 2025-2030