As AIGODS approaches the end of its funded phase, one of its clearest contributions lies in the way it supports the regulatory and supervisory responsibilities of the Instituto dos Vinhos do Douro e do Porto (IVDP) in the Douro Demarcated Region (DDR). In a territory marked by steep slopes, terraced vineyards, small irregular parcels, and more than 104,000 individual parcels, maintaining an accurate vineyard register and supervising production areas remain highly demanding institutional tasks. From the outset, the project was designed to respond directly to this reality by addressing two practical needs with clear operational value: more reliable vineyard identification for cadastral support and earlier yield estimation for planning and regulatory interpretation.

In the vineyard mapping branch, the value of AIGODS lies not simply in the use of artificial intelligence, but in the generation of spatial outputs that can be interpreted against the existing institutional reference layer. By converting image-based prediction into parcel-oriented products, the project helps isolate areas where the predicted vineyard surface may diverge from the cadastral record. This creates a technically grounded basis for more focused territorial supervision and cadastral review, while preserving an appropriate institutional logic in which model outputs support expert validation rather than replace it. In a region where exhaustive manual checking is both difficult and inefficient, this represents a meaningful gain in regulatory support.

The yield estimation branch is equally relevant for IVDP because it strengthens earlier and more structured interpretation of production potential across the territorial units that matter most for institutional decision-making. Built around the DDR, its three sub-regions, and its 169 parishes, the forecasting line supports a more consistent reading of spatial variability for annual planning, quota interpretation, and benefit-related analysis. Particularly significant is the fact that the project demonstrated that this forecasting logic can now be sustained under an open-data climate configuration sources, reducing dependence on narrower station-based conditions and reinforcing reproducibility, continuity, and future institutional resilience.

Another important aspect of this contribution lies in the way the project prepared its outputs for institutional use. The LSTM forecasting workflow, the beta mapping workflow, and their associated outputs were stabilised for delivery to IVDP, while more exploratory and unpublished components remained internal pending further consolidation. This phased approach ensured that the materials prepared for institutional transfer correspond to the outputs with the strongest immediate practical relevance. In that sense, AIGODS should not be understood as the delivery of a finished production platform, but as the creation of a technically robust and institutionally meaningful basis for better cadastral support, more targeted territorial supervision, and more informed interpretation of yield variability in the Douro.

Related News: