𝐁𝐞𝐲𝐨𝐧𝐝 𝐃𝐚𝐭𝐚 𝐒𝐡𝐚𝐫𝐢𝐧𝐠: 𝐄𝐧𝐚𝐛𝐥𝐢𝐧𝐠 𝐅𝐞𝐝𝐞𝐫𝐚𝐭𝐞𝐝 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐨𝐧 𝐃𝐚𝐭𝐚 𝐒𝐩𝐚𝐜𝐞𝐬
🚀 Excited to share that Tharindu Ranathunga had the opportunity to speak at Open Data Hub 2025 in Bolzano, Italy, a fantastic event hosted by NOI Techpark and the Open Data Hub
His talk, “𝐁𝐞𝐲𝐨𝐧𝐝 𝐃𝐚𝐭𝐚 𝐒𝐡𝐚𝐫𝐢𝐧𝐠: 𝐄𝐧𝐚𝐛𝐥𝐢𝐧𝐠 𝐅𝐞𝐝𝐞𝐫𝐚𝐭𝐞𝐝 𝐋𝐞𝐚𝐫𝐧𝐢𝐧𝐠 𝐨𝐧 𝐃𝐚𝐭𝐚 𝐒𝐩𝐚𝐜𝐞𝐬,” focused on a key innovation from our ongoing work at the Nimbus Research Centre through the 𝐂𝐎𝐑𝐃𝐒 project, funded by the Sustainable Energy Authority of Ireland (SEAI).
While traditional data spaces emphasise secure and policy-compliant data sharing, our work takes it a step further, enabling Federated Learning (FL) as a service within these ecosystems.
💡 With FL, organisations can collaborate to train models without exposing their raw data, essential in sensitive domains like energy, mobility, and healthcare.
Key highlights from the talk:
☑ Integration of FL into the 𝐈𝐃𝐒𝐀 𝐃𝐚𝐭𝐚 𝐒𝐩𝐚𝐜𝐞 𝐏𝐫𝐨𝐭𝐨𝐜𝐨𝐥
☑ 𝐀𝐮𝐭𝐨𝐦𝐚𝐭𝐞𝐝 𝐩𝐨𝐥𝐢𝐜𝐲 𝐞𝐧𝐟𝐨𝐫𝐜𝐞𝐦𝐞𝐧𝐭 in distributed FL pipelines
☑ Development of an 𝐨𝐧𝐭𝐨𝐥𝐨𝐠𝐲 𝐟𝐨𝐫 𝐅𝐋 to support governance, discovery, and interoperability
☑A POC implementation using the 𝐅𝐥𝐨𝐰𝐞𝐫 𝐅𝐋 framework (flower.ai), enabling policy-aware FL orchestration across organisations
🙏 A sincere thank you to Open Data Hub and NOI Techpark for organising the event and giving us the platform to share our work with the community.
Slides: https://lnkd.in/eBysJQ6b