June has been a milestone month for the global open research community. The 4Science team attended Open Repositories 2026, presenting six contributions that coincided with a landmark moment: the Official Release of DSpace 10, the first concrete step of the long-awaited merger between DSpace and DSpace-CRIS.
We are excited to share the key insights, technical breakthroughs, and strategic vision we brought to the stage this year.
🛠️ Inside the Developer Track: DSpace 10 & The Merger
Two sessions explored the platform’s structural evolution and how DSpace 10 is becoming more unified and transparent.
DSpace Merger: Authority Framework vs. Relationship Framework
DSpace-CRIS links entities through the Authority Framework, while standard DSpace relies on the Relationship Framework. 4Science presented a thorough analysis of the differences between the two mechanisms, the strategies adopted to ensure their coexistence during the merge, and best practices for the new unified architecture. Future releases will aim to deliver a fully converged solution.
Enhancing Transparency: The DSpace 10 Audit Trail
Accountability is essential for institutional compliance. We demonstrated the new Audit Trail feature, which tracks and records all actions on repository items, from metadata edits to bitstream events, through a well-defined data model, REST APIs, and UI integration. Repository administrators gain a complete, reliable history of platform activity.
🤝 Community Contributions & Strategic Partnerships
Institutional Equipment and Facilities in DSpace-CRIS
Supported by the Vietsch Foundation, a joint project with euroCRIS and the Technical University of Hamburg extends the DSpace-CRIS data model to include research instruments and facilities. The new entity type is modeled on CERIF and the work of the RDA PIDINST Working Group, and is currently being tested in production via TU Hamburg’s institutional CRIS, TORE.
Towards the Integration of CORE into DSpace
Through a strategic partnership with CORE, we are adding new capabilities directly into DSpace: the CORE Recommender, CORE live import and SDG classification suggestions delivered via the COAR Notify protocol: a concrete step toward a richer, more connected open research infrastructure.
🗺️ Digital Cultural Heritage: Space, Time, and Narratives
Historica: the digital library of the University of Bologna
We presented Historica, the digital library of the University of Bologna, built on 4Science’s DSpace-GLAM. Rather than managing digital files in isolation, Historica structures rich relationships between objects and contextual entities — persons, places, events — enabling users to explore historical scenarios through interactive timelines, maps, and IIIF-based image services, in full alignment with international preservation and interoperability standards.
⚡ Challenges and Solutions for Reliable and Performant Repositories
Our COO Andrea Bollini’s presentation on reliable and performant DSpace repositories resonated strongly with the conference’s opening keynote, where Kathleen Shearer, Kazu Yamaji, and others stressed that repositories must be highly interoperable and provide reliable services to the research community.
In 2025, 4Science began migrating its hosting infrastructure from traditional VM-based setups to a containerized deployment powered by Kubernetes and AWS cloud-native services. This transition addressed horizontal scalability, delivered consistent High Availability under load, and reduced environmental impact by eliminating unnecessary resource over-provisioning. Numerous improvements have already been contributed back to the DSpace codebase and the ongoing merger.
In the closing keynote, Kathleen Gregory emphasized that repositories represent a vital, distributed pillar of the research infrastructure. As such, they must remain resilient in the face of ongoing risks, including commercial buyouts, unexpected shutdowns, and geopolitical challenges. In this context, 4Science’s commitment to an “Open Source First” approach is our concrete contribution to true Digital Sovereignty. (We signed the Open Letter – Europe’s Digital Future)
🤖 The Dual Challenge of AI in Open Knowledge
Among the most discussed topics at OR2026 was the double-edged relationship between repositories and Artificial Intelligence.
On one side: how to embed AI capabilities to make repositories smarter — through recommendations, metadata enrichment, and semantic discovery.
On the other: the massive crawling challenge — AI agents aggressively harvesting knowledge from the very platforms that are the primary open source of that knowledge.
Because research discovery is increasingly AI-mediated, blocking crawlers is not an option; maximizing visibility remains our goal. Repositories must evolve to be seamlessly accessible to both humans and machines. At 4Science, we are addressing this directly through our new container-based infrastructure, designed to handle this shift efficiently and sustainably. Our mission is to deliver sustainable, cutting-edge architectures, making high-performance and high-availability solutions accessible to institutions of every size and budget, including those with more limited spending capacity.