Hybrid Territorial AI: Modeling A Decision Support System for Strategic Siting in Complex Territorial Contexts

This research aims to lay the foundations for a hybrid territorial artificial intelligence system designed to support strategic location decisions in a context of strengthened regulations (ZAN, SCOT, PLU) and complex interactions between public and private stakeholders. Based on a corpus of 39 Territorial Coherence Schemes, the methodology combines ontological formalization, an expert rule engine, and relational data modeling. The purpose is not to deliver a fully operational tool, but rather to build the theoretical and empirical framework required for the future development of an explainable, governable, and co-constructed decision-support system. The study demonstrates how territorial complexity can be structured into intelligible concepts, how divergent temporalities among actors can be modeled, and how conditions for dialogical algorithmic governance can be established. It represents a first step toward a transparent and collaborative collective intelligence tool that can mediate between private rationales and public planning objectives.