PRESENTING THE RAIDO JANUARY 2025 NEWSLETTER.
Welcome to the RAIDO January newsletter! This edition highlights our project’s recent achievements, publications, milestone deliveries, and engagement activities as we continue to advance energy-efficient AI solutions.
January 2025 marks a significant milestone period for the RAIDO project, with the successful delivery of two critical project milestones (Oct 2024 & Dec 2024) and continued research publication activities. Our team has achieved important progress in platform architecture design and pilot specifications, while maintaining active engagement in the European AI research community through clustering events. The project continues to advance steadily as we move into the second year of operations.
ONLINE MEETINGS
Regular coordination and progress tracking across all work packages is successfully ongoing, and all consortium partners and work package leaders are actively contributing to the project.
WP Leaders and technical teams are continuing to participate in biweekly coordination and progress tracking calls across all work packages, focusing on deliverable and milestone completion and cross-work package collaboration
EVENTS
In November 4th, 2024, RAIDO participated in the clustering activities of Achieving efficient trustworthy AI in the cloud-edge continuum (ADRF) [link]. This collaborative workshop brought together five Horizon Europe projects to address critical questions surrounding AI systems efficiency, trustworthiness, and data optimization in the cloud-edge continuum (CEC). The event specifically focused on exploring how AI systems can be improved across multiple dimensions while considering the implications of the emerging AI Act.
Key Takeaways: The clustering event provided invaluable opportunities for knowledge exchange and collaboration among leading European projects working on AI optimization challenges. Through panel discussions featuring key experts from academia and industry, RAIDO gained practical insights into implementing efficient trustworthy AI solutions in real-world scenarios. The workshop successfully raised awareness about practical restrictions and major enablers for efficient trustworthy AI in the cloud-edge continuum, directly aligning with RAIDO’s core objectives. Our participation allowed the team to explore synergies with complementary projects and discuss shared approaches to addressing data quality, environmental sustainability, and regulatory compliance challenges.
Impact on RAIDO: This collaborative engagement strengthened our understanding of the broader European AI research landscape and provided valuable perspectives on implementing trustworthy AI solutions within the regulatory framework of the AI Act. The insights gained will inform our ongoing work on energy-efficient AI architectures and ethical AI frameworks.
PUBLICATIONS
We are proud to announce the publication and presentation of an important paper during the SIKDD Conference in Ljubljana, Slovenia, October 2024. [conference]
Predicting poverty using regression [paper link]. This paper evaluates the use of various regression models to predict poverty rates using a comprehensive dataset of 111 variables from sources such as the UN and the World Bank. This study highlights the importance of addressing multicollinearity and incorporating a wide range of features to improve the generalizability of poverty prediction models. Future research should explore more complex methods, such as neural networks, and refine model hyperparameters for enhanced performance. Additonally, this research contribution demonstrates RAIDO’s commitment to developing AI solutions that address real-world societal challenges, showcasing practical applications of our trustworthy AI methodologies in socioeconomic prediction tasks.
PROJECT MILESTONES & DELIVERABLES
The project has achieved two major milestones that represent critical foundations for our technical architecture and implementation strategy. Milestone 1 involved the completion of D2.1, focused on the availability of pilot specifications, performance evaluation frameworks, enabling technologies, and requirements.
Building upon this foundation, Milestone 2 encompassed the delivery of D2.3, covering the overall RAIDO Platform Architectural Design. Under the leadership of UBITECH-LIM and completed in Month 12, this architectural framework represents a critical advancement in creating the integrated RAIDO ecosystem for energy-efficient AI development and deployment.
Looking toward future deliverables, the team is preparing for Milestone 3, which will focus on D4.1 covering legislation analysis for privacy, intellectual property rights, and real or synthetic data/metadata considerations. Scheduled for Month 18, this upcoming milestone will address the crucial regulatory and legal framework within which RAIDO solutions must operate.
