CSP Plenary Session on Science Platforms

Operational science platforms are now prevalent in the astronomy community, presenting the IVOA an opportunity to leverage their availability through interoperability mechanisms. The Canadian Advanced Network for Astronomical Research (CANFAR) Science Platform, developed by CADC, has been in full operation since 2021, serving a large portion of the Canadian astronomy community. I will highlight the principles behind the success of CANFAR, including the use of OCI containers as the unit of software. Also, in the context of SRCNet, we are introducing the use of software metadata on containers, which could provide a path for the VO to support FAIR principles, not only for data, but for software as well.

CADC is active in integrating aspects of CANFAR with other "large telescope networks": the SKA SRCNet, the Vera C. Rubin Observatory, and the LISA Consortium. As there are multiple sites involved in the operation of these networks, they act as precursors to science platform interoperability, providing insight into the complexities involved when dealing with platforms running on heterogeneous infrastructure, highlighting the need for an abstraction layer into platform functionality.

This combination of operational experience and collaborations with these networked science platform efforts offers CADC a unique perspective on the opportunities available to IVOA in advancing science platform interoperability.

To support these emerging workflows, ESA has developed ESA Datalabs and the Euclid Data Space (EDS). ESA Datalabs provides remote-analysis environments connected to ESA mission archives, while EDS integrates the Euclid archive, ESASky visualisation services, and ESA Datalabs into a more unified science-platform ecosystem for the Euclid scientific community.

Using ESA Datalabs and EDS as concrete examples, this presentation discusses emerging interoperability challenges for next-generation science platforms, including distributed storage and data locality, workflow portability, remote-analysis environments, authentication and authorization, federation across platforms, and cloud-native workflows.

We discuss where existing IVOA standards already provide strong foundations for archive interoperability and data access, while also highlighting areas where future science-platform ecosystems may require interoperability discussions extending beyond traditional archive-centric models toward workflows, compute, storage, and federated analysis environments.