Daphne Garrido
Preprint · Version 1.0 · September 2026 - Preprint on Zenodo
Status: Not peer reviewed. Technology perspective, conceptual framework, and research roadmap.
Series: The Human Embodied Intelligence Research Program · Preprint 5
Human digital twins are emerging as dynamic computational representations of individuals constructed from physiological, anatomical, behavioral, clinical, and environmental data. Current systems, however, are predominantly organ-specific, task-specific, or clinically bounded, and relatively few implementations satisfy strong definitions of a digital twin requiring personalization, dynamic updating, and predictive capability.
This paper develops a research framework for extending biological digital-twin methods toward whole-body computational models capable of reproducing selected closed-loop dynamics relevant to embodied cognition.
The framework follows from the Human Embodied Intelligence Research Program, in which full-body intelligence is treated as a proposed organism-level construct involving interactions among interoceptive, autonomic, neural, sensorimotor, and environmental processes. If those interactions prove empirically important, a computational representation intended to model embodied intelligence would require more than anatomical reconstruction or a static archive of biological measurements. It would need to represent internal states, update them dynamically, integrate multimodal signals, respond to perturbation, generate actions or predictions, and remain plastic under new conditions.
Three engineering layers are distinguished:
Contemporary digital-twin research, mechanistic physiological modeling, neuromechanical simulation, wearable sensing, and emerging interoceptive artificial-intelligence frameworks provide partial technological foundations for this agenda. They do not presently constitute a complete whole-body twin.
The term embodied digital continuity is introduced here as an operational research concept referring to persistence of validated subject-specific dynamical and behavioral properties across computational updating and interaction. It does not denote, and should not be interpreted as demonstrating, continuity of consciousness, preservation of personal identity, mind transfer, or literal survival.
The scientific objective is therefore deliberately narrower than “uploading a person”: determine which biologically validated dynamics can be reconstructed computationally, how reconstruction fidelity should be tested, which functions require closed-loop embodiment, and where biological-to-digital correspondence fails.
This formulation transforms embodied digital continuity from a metaphysical assertion into a falsifiable engineering and modeling program.
Keywords: human digital twin; biological digital twin; whole-body modeling; embodied cognition; full-body intelligence; computational physiology; interoceptive artificial intelligence; multimodal sensing; physiological modeling; closed-loop systems; embodied AI; digital continuity; model fidelity; computational embodiment; personalized modeling.
Digital-twin technology originated as an engineering strategy for constructing computational counterparts of physical systems that can be informed by observations of those systems and used for monitoring, simulation, prediction, or control.