Continuous Ambient Haptic World-State: A Skin-Worn Encoding Framework for Unified Sensory Substitution in Deafblind Users

Existing haptic assistive devices for blind and deafblind users operate on a discrete, reactive model: a stimulus fires in response to a specific trigger, then ceases. This paper proposes a fundamentally different interaction architecture — continuous ambient haptic world-state — in which a skin-worn actuator array delivers an uninterrupted, morphing haptic stream encoding the user's spatial and semantic environment in real time, without requiring active query. Built on existing flexible electroactive polymer actuator technology, the proposed system translates AI-interpreted scene data into a differential bump-pattern grammar across a finger- or palm-worn patch, substituting simultaneously for both vision and hearing in deafblind users. The architecture includes a bidirectional channel: the same surface that delivers environmental information receives pressure-pattern responses from the user, enabling two-way communication through skin contact alone. The contribution of this paper is not hardware — the actuator exists — but the system architecture, the continuous-ambient interaction model, and the semantic-to-haptic encoding framework that makes unified deafblind sensory substitution possible.