Agentic AI systems that operate on geospatial data require more than autonomous reasoning and tool use; they must also support explicit orchestration, spatial validation, governance, provenance, failure handling, and human decision authority. Earlier work introduced the ANASSA framework, which defined high‑level components and responsibility boundaries for spatial intelligence (arXiv:2609.14824), but it did not specify how to systematically decompose those responsibilities into an implementable low‑level architecture. This paper applies Design Science Research (DSR) to refine ANASSA’s high‑level architecture (HLA) responsibilities into a low‑level architecture (LLA). The resulting design preserves the original responsibility boundaries while translating them into executable structures, including 47 logical roles, versioned task graphs, capability‑based agent selection, typed contracts, explicit runtime states, spatial validation mechanisms, governance controls, failure and recovery paths, and provenance records. An end‑to‑end execution trace demonstrates the architecture, and it is analytically evaluated against requirements traceability, internal consistency, responsibility preservation, and technology independence. This high‑to‑low‑level design provides an implementation‑facing bridge between the ANASSA framework and future system realizations, enabling governed spatial agentic workflows to be inspectable, reconstructable, and testable without tying the architecture to a specific AI model, geospatial platform, communication protocol, or deployment environment.
Review