hexadecimal-REBOOT · hexadecimal-REBOOT · Oct 5, 2025
diff --git a/LICENSE b/LICENSE
@@ -0,0 +1,51 @@
+VDSX RESTRICTED LICENSE (Company-Only)
+Version 1.0 — 2025-01-01
+
+Designated Entity: NIMBUS LLC, LANDON CRUTCHFIELD & MCCOY HOLDINGS
+
+1. GRANT OF RIGHTS
+Subject to continued compliance with this License, the Licensor grants the Designated Entity (“LANDON CRUTCHFIELD & MCCOY HOLDINGS, NIMBUS LLC”) a limited, non-exclusive, non-transferable, non-sublicensable license to:
+  (a) use, reproduce, and modify the Licensed Materials;
+  (b) create derivative works of the Licensed Materials; and
+  (c) deploy and operate the Licensed Materials and derivative works internally and in products or services offered by or for the Designated Entity.
+
+All rights not expressly granted are reserved by the Licensor.
+
+2. SCOPE & RESTRICTIONS
+2.1 This License applies to the source code, binaries, documentation, specifications, and assets included in this repository (“Licensed Materials”).
+2.2 The rights granted hereunder are limited to the Designated Entity and its employees or contractors acting on its behalf under written confidentiality obligations.
+2.3 You MAY NOT:
+   (i) redistribute, publish, or make the Licensed Materials (or substantial portions) available to any third party, whether in source or binary form, except as integrated into a product or service provided by the Designated Entity;
+   (ii) sublicense, assign, or otherwise transfer the Licensed Materials or this License to any third party;
+   (iii) remove or alter any notices of ownership, attribution, or this License;
+   (iv) use any Licensor names, marks, or logos except as necessary to exercise rights granted herein.
+
+3. DATA & MODELS
+3.1 You may use the Licensed Materials to process and generate data (including VDSX files and point clouds) and to train, fine-tune, and evaluate machine-learning models, provided such use is by or for the Designated Entity.
+3.2 You are solely responsible for all input/output data and third-party rights. No rights to third-party content are granted by this License.
+
+4. NO PATENT LICENSE
+No patent rights are granted under this License, whether by implication, estoppel, or otherwise.
+
+5. OWNERSHIP
+The Licensor retains all right, title, and interest in and to the Licensed Materials. This License does not transfer ownership.
+
+6. TERMINATION
+This License automatically terminates upon:
+  (a) any material breach by You that remains uncured for 15 days after written notice; or
+  (b) Your attempt to transfer or sublicense rights in violation of Section 2.3.
+Upon termination, You must immediately cease use and destroy or securely delete all copies of the Licensed Materials in Your possession or control. Sections 3–10 survive termination.
+
+7. WARRANTY DISCLAIMER
+THE LICENSED MATERIALS ARE PROVIDED “AS IS” AND “AS AVAILABLE” WITHOUT WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED. WITHOUT LIMITATION, LICENSOR DISCLAIMS ANY WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT, AND NON-INTERFERENCE.
+
+8. LIMITATION OF LIABILITY
+TO THE MAXIMUM EXTENT PERMITTED BY LAW, LICENSOR SHALL NOT BE LIABLE FOR ANY INDIRECT, INCIDENTAL, SPECIAL, CONSEQUENTIAL, OR PUNITIVE DAMAGES, OR FOR ANY LOSS OF PROFITS, REVENUE, DATA, OR GOODWILL, ARISING FROM OR RELATING TO THE LICENSED MATERIALS OR THIS LICENSE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. LICENSOR’S AGGREGATE LIABILITY SHALL NOT EXCEED ONE HUNDRED U.S. DOLLARS (US$100).
+
+9. GOVERNING LAW & VENUE
+This License is governed by the laws of the State of California, USA, without regard to conflicts of law principles. Exclusive jurisdiction and venue shall lie in the courts located in San Francisco County, California, USA.
+
+10. MISCELLANEOUS
+If any provision of this License is held unenforceable, the remaining provisions will remain in full force. This License constitutes the entire agreement regarding the Licensed Materials and supersedes any prior or contemporaneous understandings.
+
+© 2025 American Robotics IO. All rights reserved.
diff --git a/README.md b/README.md
@@ -1,2 +1,108 @@
-# Token-Matrix
-Token Matrix: A context-aware language model system using vector databases for efficient token management. Optimizes token usage, enhances response relevance, and supports multi-context environments. Ideal for AI-powered applications requiring intelligent, context-sensitive language generation.
+# 🌍 VDSX Family
+
+Visual Data Scene eXchange — Building the World in 3D, for Humans and AI
+
+⸻
+
+## 🔷 Overview
+
+VDSX is a next-generation 3D capture and reasoning framework that connects vision, geometry, and semantic understanding in real time.
+It turns any camera—mobile, embedded, or industrial—into a VDSX-enabled 3D scanner capable of generating structured, ontology-aware .vdsx scene files.
+
+Each .vdsx file encodes not just images, but depth, motion, segmentation, spatial alignment, and meaning—the building blocks of a living, searchable digital twin of the real world.
+
+The VDSX Family forms the infrastructure for planet-scale 3D understanding, where every captured frame contributes to a continuously learning world model.
+
+⸻
+
+## 🧩 Core Ecosystem
+
+| Module | Purpose | Description |
+| --- | --- | --- |
+| VDSX | Core Capture Protocol | Defines the core .vdsx standard — integrating RGB, depth, segmentation, and ontology layers into a unified format. It establishes the foundation for real-time 3D reconstruction and AI reasoning. |
+| VDSX Lite | Sensor-Enhanced Capture | Extends VDSX with IMU (Inertial Measurement Unit) and MIMO (Multi-Input Multi-Output) sensors for motion, orientation, and precision spatial tracking. Designed for mobile and field-level capture, enabling robust real-world scene reconstruction from handheld or moving cameras. |
+| VDSX Hardware | Edge & Embedded Runtime | Runs the VDSX protocol directly on FPGA/SoC devices and specialized boards (e.g., Nimbus N001/N010). Supports sensor fusion inputs (IMU, LiDAR, ToF, thermal, GNSS) and deterministic 3D scene encoding at the hardware layer for ultra-low-latency applications. |
+| VDSX Universal | Media, Gaming & Entertainment | The interoperability bridge connecting VDSX to creative industries. Import assets from Blender, Unreal, Unity, Maya, or GLTF, and export .vdsx data to cinematic, VR, or XR pipelines. Enables photo-realistic TSR rendering and hybrid workflows between real and synthetic scenes. |
+| VDSX Stereo | Multi-View Relative Depth | Solves relative depth estimation across stereo or multi-camera configurations. Performs disparity fusion, optical alignment, and motion-based parallax reconstruction for temporally consistent 3D data. |
+| VDSX Files | Data Structure & Retention | Defines the .vdsx file schema. A modular, quantum-signable container supporting retention modes (choose what to save) with file sizes from 300 KB to 4 GB. Supports depth bins, semantic layers, IMU/MIMO telemetry, and compression options for cloud or local storage. |
+| VDSX Global Indexing | Planet-Scale Mapping | Distributed 3D indexing system that aggregates all .vdsx captures. Supports real-time spatial search, semantic ontology linking, and geographic query resolution. Enables a global “Search the World in 3D” experience. |
+| 3DWOM | 3D World Observer Model | The AI model trained on VDSX data. Integrates geometry, semantics, and ontologies to create a reasoning model that understands the world physically — not just by pixels, but by relationships, function, and context. |
+
+⸻
+
+## 🧠 Why It’s Different
+
+Today’s AI sees pixels.
+
+Vision–language models like CLIP or GPT-4V recognize patterns, but lack geometry, permanence, and true spatial understanding.
+
+VDSX sees the world.
+
+VDSX integrates depth, motion, and meaning — allowing AI to understand how things exist and why they matter.
+
+It’s not just “That’s a fire hydrant.”
+It’s “That’s a fire hydrant, connected to a municipal water system, standardized in size, placed near road infrastructure, and governed by local utility regulations.”
+
+This leap — from semantic labeling to contextual reasoning — is what makes 3DWOM the first model capable of planning and understanding in true 3D space.
+
+⸻
+
+## 🧱 Architecture
+
+```
+📦 VDSX Family
+ ├── vdsx/                  # Core capture + reconstruction logic
+ ├── vdsx-lite/             # IMU + MIMO sensor extension
+ ├── vdsx-hardware/         # Edge + FPGA sensor fusion runtime
+ ├── vdsx-files/            # File schema, retention, quantum signing
+ ├── vdsx-universal/        # Blender/Unreal integration & exports
+ ├── vdsx-stereo/           # Stereo & multi-view disparity solver
+ ├── vdsx-global-indexing/  # Distributed 3D world search
+ └── 3dwom/                 # 3D reasoning model training pipeline
+```
+
+Each module interoperates through the NOLYN Protocol 0.1a for secure, post-quantum AI-to-AI communication.
+
+⸻
+
+## 🛰️ Example Workflow
+
+1. **Capture** — A VDSX or VDSX Lite camera captures RGB, depth, IMU, and segmentation data in real time.
+2. **Generate .vdsx File** — Data is fused into a structured container with calibrated intrinsics, extrinsics, and ontology metadata.
+3. **Upload & Index** — Files are uploaded to the VDSX Global Index, where they’re spatially and semantically indexed.
+4. **Reconstruct & Render** — Using TSR (Triangle Splatting Rendering), 3D scenes are reconstructed with photo-realistic fidelity.
+5. **Train 3DWOM** — The AI learns physical understanding from billions of .vdsx samples — enabling world-scale reasoning.
+6. **Search & Query** — Users and agents can “search the world in 3D” — by object, concept, or ontology class.
+
+⸻
+
+## 🚀 Scaling the Real World
+
+By deploying VDSX Lite or VDSX Hardware devices across fleets (e.g., delivery vans or municipal garbage trucks),
+the system can map 80 % of the U.S. in under a year, capturing stereo and depth data at scale.
+
+Each frame contributes to 3DWOM’s evolving understanding of Earth — creating a shared digital twin usable for urban planning, robotics, simulation, and AI training.
+
+⸻
+
+## 🔒 Restricted License
+
+**VDSX Internal Research License (VIRL-1.0)**
+
+Copyright © 2025 NOLYN / VDSX
+
+Redistribution, modification, or derivative use is strictly prohibited without prior written consent.
+
+Authorized for internal R&D and affiliated research projects only.
+
+For partnerships or commercial licensing contact
+[yannis@americanrobotics.io](mailto:yannis@americanrobotics.io)
+
+⸻
+
+## 🧭 Mission
+
+“We’re not just scanning the world —
+we’re teaching AI to understand it.”
+
+⸻