Ai-Driven Data Ecosystems For Smart Olive Oil Sector

Official statistics and administrative registers play a central role in monitoring the olive oil sector. However, in Italy, these data are dispersed across multiple institutions, limiting interoperability and advanced analytical use. This research develops a structured framework to transform accessible institutional data into an AI-ready ecosystem capable of supporting intelligent decision-making in the olive oil sector. First, the authors systematically map statistical and administrative sources relevant to production, structural indicators, market dynamics, trade flows and traceability. Based on this assessment, the authors define a Minimum Interoperable Olive Statistics Dataset (MIOSD) to harmonise core variables and metadata across institutions. Second, the authors introduce a complementary Process-and-Quality extension layer (MIOSD-PQ), incorporating milling technology descriptors, extraction performance indicators, energy and water consumption metrics, and chemical quality parameters. This dual-layer architecture enables machine learning applications such as forecasting, anomaly detection and scenario analysis, while establishing the foundations for process optimisation and sustainability-oriented decision support. The analysis identifies key systemic barriers, including fragmented governance, heterogeneous definitions, limited machine-readable access, restricted linkage keys and uneven up-date frequencies. A FAIR-aligned data catalogue and benchmark AI tasks are proposed as practical intervention lines towards a scalable, interoperable and sustainability-driven olive oil data ecosystem.