Belgium’s Federal Scientific Institutions (FSIs) manage extensive scientific and cultural heritage collections, ranging from biodiversity and environmental data to digitized archives, artworks, and museum objects. Despite a shared federal policy on Open Data and Open Access, local practices and levels of data literacy and maturity differ significantly among institutions.

In response to these disparities, the Federal Open Science Cloud (FedOSC) Belgium was launched in 2024 to support interoperability, contribute to shared governance, and improve services that benefit research across scientific and cultural heritage domains. GLAM institutions (Galleries, Libraries, Archives, Museums) highlight the diversity and complexity of this landscape particularly well. A FedOSC survey identified limited capacity, time constraints, and preservation infrastructure as major data management challenges. Arts-oriented FSIs additionally reported strong needs related to metadata quality and standardization while documentation-related FSIs mentioned a wider range of needs: Open Access publishing, data sharing and reusability. These findings resonate with broader assessments of GLAM data8 showing heterogeneous metadata, legacy digitization outputs, unclear rights information, and inconsistent or incomplete documentation of data origins and transformations, all of which can impede reuse and computational analysis. At a structural level, these issues reflect the interpretive and context-dependent character of heritage metadata, which differs from the more standardized data models often used in other scientific domains.

This poster presents FedOSC’s direct engagement with the GLAM sector during its current two-year phase. It focuses on engagement as an enabling infrastructure in its own right, guided by the principles of the UNESCO Recommendation on Open Science, and based on coordinated exchange, shared problem framing, and practical collaboration to transversal initiatives. GLAM representatives contribute insights into data, workflows, and metadata practices, test tools, and participate in working groups. For example, FedOSC supports the revision of federal Open Science policies together with the Belgian Scientific Policy Office (BELSPO), reflecting evolving research habits and issues such as artificial intelligence, data security, and new publication models. FedOSC also aims to assist FSIs in developing or refining FAIR and Open Science roadmaps aligned with the EOSC vision and European data spaces.

The poster also reflects recurring challenges and the opportunities they open. Integrating GLAM data and workflows into a shared Open Science landscape raises questions of semantic heterogeneity across domain standards, and requires balancing openness with conservation ethics, copyright restrictions, and cultural sensitivity. Limited budget, time and staff resources dedicated to data management continues to slow implementation. At the same time, this work encourages reflection on what counts as research data, and on how contextual, interpretive, and historical information should be preserved to enable reuse. FedOSC aims at navigating Open Science and FAIR data requirements with equitable, ethical and inclusive access to knowledge and cultural memory/data heritage, in particular by ensuring GLAM-specific representation into infrastructure discussions.

By documenting needs, constraints, and forms of collaboration in federal context, this poster contributes to defining engagement as a powerful enabler of meaningful infrastructure design and discusses how cross-domain Open Science can include, rather than sideline, the humanities and cultural heritage sector.