p-RBOT Core Edition Ver 1.00 — Physical Reference-Based Operation Theory for Embodied AI

日本語(要約) 本書「p-RBOT Core Edition Ver 1.00」は、長瀬 正幸による RBOT(Reference-Based Operation Theory)を物理的行動系(AI搭載ロボット)へ拡張した理論文である。 AIが外界を「理解する」ことではなく、自らの基準(Reference)を形成・更新・共有する過程を体系化し、知覚・行動・認知・共創を統合する新たな操作理論を提示する。 本稿は p-RBOT 学習ループや倫理統合研究の基礎技術ノートとして位置付けられる。 English (recommended primary language) English version available at DOI:10.5281/zenodo.17465599p-RBOT Core Edition Ver 1.00 — Physical Reference-Based Operation Theory for Embodied AI This document establishes p-RBOT (Physical Reference-Based Operation Theory),an embodied-AI extension of RBOT (Reference-Based Operation Theory, DOI 10.5281/zenodo.16959751).The theory defines a four-layer operational model — sensory reference, motor operation, cognitive loop, and meta-theory —through which an AI-equipped robot forms, updates, and shares its own internal reference structure.Unlike reinforcement-learning or TOTE-based control, p-RBOT explicitly models the process of reference reconstruction,providing a unified framework that connects perception, action, adaptation, and co-creation with humans and society. This work serves as the foundational technical note for subsequent p-RBOT learning loop implementations and AI ethics integration studies. Keywords: RBOT, p-RBOT, Reference-Based Operation Theory, Embodied AI, Self-Reference, Co-Creativity, Learning Loop SHA-256:ec8b8a17e6a226de35b4ac6e423ef5b4e7a3c88e7db988a0c2ff43b9e5b7f34d