Trusted Research Conversations
Episode 1: Federated Learning and Trusted Research Environments
In the first episode of Trusted Research Conversations, Aridhia Senior Product Manager Ross Stiven sits down with Chong Shen Ng, Research Engineer at Flower Labs, to discuss how federated learning is reshaping what’s possible in Trusted Research Environments and what it takes to make it work in practice.
Federated learning inverts the traditional data-sharing model. Rather than moving sensitive patient data to a central location for analysis, the model goes to the data. For health data research, where data governance, patient privacy, and cross-border regulatory requirements routinely block collaboration, this matters.
Ross and Chong Shen cover the practical realities of deploying federated learning across multi-site research networks. From the infrastructure complexity that typically slows projects down, to how the Aridhia DRE and Flower framework work together to address it. The Aridhia DRE provides the pre-certified, governed infrastructure at each participating site, with network-secure workspaces and organisation-level isolation built in. Flower provides the federated learning orchestration layer via SuperNodes at each data site, a SuperLink coordinating training runs, and SuperGrid for managing federations at scale. Together, they compress what typically takes six months to approximately one week.
They also discuss where federated learning sits in the broader evolution of health data research infrastructure, the remaining challenges around standardisation and governance, and what research teams should be thinking about before they start.
🔗 Flower – The friendly federated AI framework
🔗 Flower Hub – Flower Apps built by the community
🔗 Flower AI Summit 2026 – Our annual summit for federated and agentic AI