Jorge L. Argibay
Every automated system in existence solves the same underlying problem: translating a specification of desired behavior into actual behavior. Despite this universality, no general formal framework for the translation exists. Every domain — software, manufacturing, building automation, device interfaces, robotics — has built its own ad hoc solution, encoding human intent into domain-specific mechanical representations. This paper argues that the encoding itself is unnecessary. It defines behavioral ontology — a formal, persistent, machine-interpretable specification of what a system is and does — and introduces Dynamic Behavioral Ontology (DBO) as the name for this new layer of the computational stack. It identifies seven properties such a specification must have, presents OBI1 (Ontology Based Intent) as the first working implementation, and demonstrates through a continuous example — fleet management evolving from screen-based software through GPS tracking to autonomous vehicle control — that the generalization from software to physical automation requires no change to the formal system, only the expansion of the runtime interpreter’s physical capabilities. It introduces the concept of distributed DBO, in which behavioral specifications are transmitted to remote runtime interpreters across networks of physical devices — made practical by the extraordinary compactness of DBO as data — and identifies this as the missing intelligence layer of the Internet of Things. It proposes the foundational conjecture that the five components of the formal system constitute an irreducible orthogonal basis for formal intent specification, refines the conjecture by distinguishing computational orthogonality from descriptive orthogonality, demonstrates through formal language analysis that the five components constitute a context-free grammar for behavioral specification, and establishes a research agenda for the emerging discipline. It argues that AI is what makes behavioral ontology practically relevant: the general concept could have been conceived decades ago, but without AI to construct specifications at the scale of real-world systems, it would have remained a theoretical curiosity. The convergence of AI and formal behavioral specification is what transforms the theory into a practical discipline. A complete worked example — a fleet management system specified in the paper’s grammar — is provided as an appendix.