Network services for Maritime Spatial Planning (MSP): "as-is" & "to-be" use cases

The document “Network services for MSP: as-is & to-be use cases” outlines a technical vision from the DG MARE/CINEA Technical Expert Group sub-group “Network services for MSP” on how Marine Spatial Planning (MSP) can access distributed data and transform it into maps and decision-ready knowledge. It starts from the current European interoperability foundation created by the INSPIRE Directive and its network services, framing today’s MSP reality as a landscape of spatial data infrastructures (SDIs) and “silos” that are made more interoperable through standardised metadata and services. MSP is presented as an analytical pipeline turning data into maps and then into knowledge, requiring not only technology, but also aligned processes and governance among stakeholders. AS-IS use cases (INSPIRE-based): · Data access and mapping via data virtualisation: a “metadata broker” approach (illustrated through the Marine Analyst platform) that stores and serves metadata to provide an abstraction layer over many heterogeneous sources, allowing data to be accessed across the network without replication. The key idea is to rely on machine-readable metadata (online resources, protocols) to integrate sources virtually and run analyses that generate maps and reports. · Data discoverability and availability via augmented data access: harvesting metadata (e.g., through catalogue services) to support uniform cross-repository search, and then enriching user experience with advanced services, including interpreting WMS-returned images to assess what data is actually available in a region of interest. The document highlights the potential of computer vision (object detection, segmentation, feature extraction) to complement classic metadata and make services “actionable” for MSP workflows. TO-BE use cases (beyond INSPIRE’s original 2007-era design): · Linked data and Semantic Web: positioned as a shift from a “syntactic ecosystem” of silos toward a web of interconnected data (RDF, SPARQL), improving machine-driven discovery, integration, and reuse of spatial data from diverse sources (including outside traditional SDIs). A showcased proof of concept involves harvesting linked data from many web pages to build indicators (e.g., environmental and socio-economic). The document also notes that modern AI/LLMs can extract information from unstructured web documents even when linked data is absent, changing the data-science paradigm. · Big data technologies and federated analysis: motivated by data volume growth and the collaborative MSP setting. The document advocates sending analysis to data (instead of moving data to analysis) via federated/distributed frameworks, referencing Elasticsearch-based approaches and an example “Chlorophyll Sandbox” providing long time-series access for analysis. Linked data is also described as compatible with federated approaches, with added benefits from machine-interpretable concepts. Conclusion:The main message is that “machine-readable and actionable” data and metadata are increasingly essential for MSP to fully benefit from emerging AI and advanced analytics. Preparing MSP data with better organisation and semantic annotation is presented as a path to more scalable discovery, integration, and evidence-based decision-making, enabling more operational and effective MSP outcomes.