Functional Metacognitive Transfer for Artificial Cognitive Architectures: An Inverse Cognitive Engineering Approach

This work introduces the Functional Metacognitive Transfer Theory (FMTT), a design-oriented theoretical framework that explores how functional principles observed in human metacognitive experience can be abstracted and transferred into artificial cognitive architectures without assuming structural, neural, or ontological equivalence. The theory is explicitly positioned within an engineering and control-oriented epistemology, where metacognition is treated as a heuristic source for architectural design rather than as empirical evidence of biological mechanisms. Based on this framework, we propose the ACTA–AMA–SMCA architecture, a modular artificial cognitive system that separates primary cognitive dynamics from metacognitive supervision and formal stability control. The architecture integrates perception, memory, decision-making, and action under explicit stability guarantees derived from dynamical systems theory, including Lyapunov-based analysis and Input-to-State Stability (ISS). Computational simulations are presented as proof-of-concept evidence, illustrating how metacognitive-inspired regulation improves stability, interpretability, and controlled adaptation under perturbations. The contribution of this work lies not in explaining human cognition, but in providing a systematic, epistemologically conservative methodology for transferring metacognitive principles into artificial systems evaluated exclusively through formal and empirical engineering criteria.