Vertical-Driven Architectures
Collection
Physical AI system designs by CodeNinja, reasoned on Praxis, made living with Hyper Ontology. • 47 items • Updated
choice stringlengths 6 18 | picked stringlengths 25 91 | why stringlengths 71 119 |
|---|---|---|
Frontier model | Mistral Large 3 675B, Apache-2.0, FP8 | No revenue trigger or attribution duty; one 141 GB HBM class node holds the 682 GB weights with about 350 GB KV ceiling |
Site model | Qwen3.8-27B, Apache-2.0, one 48 GB PCIe card | About 31 GB at FP8 keeps the assistant alive in degraded mode during WAN loss |
Detection model | RF-DETR, Apache-2.0 N to L checkpoints | XL and 2XL carry a platform license; N to L takes site fine-tuning for evaluation sets |
Tracking model | OC-SORT, MIT, motion-only | No re-identification branch, so the safety system never becomes a surveillance system |
Embedding model | BGE-M3, MIT, 8,192 token window | One pass yields dense, sparse and multi-vector retrieval on a small box |
Edge compute class | Rugged rack server with PCIe inference accelerators | Sized per site from the measured decode-plus-inference budget on camera substreams, with headroom on both |
Camera estate | The existing roughly 220 fixed IP cameras, reused | Six-gate survey per camera; purpose-placed fixed additions only where a view fails |
Sizing rules | Streams, calibration, meters, bandwidth | Size from streams not boxes; calibrate when speeds must survive challenge; budget video on the correlated peak |
Sensing | WMS event stream, labor extracts, yard appointments, link and GPU health telemetry | The twin's feeds and the failure signals the boundary behavior depends on |
Patterns | System of Context, camera analytics as feedback loop, baseline before anything, shadow mode | The assistant drafts and a person approves; analytics never act as a safety-rated stop |
Ground | The operator's own hardware, open weights, in-company environment | Domestic supply engages no export pathway; only events and short clips leave each building |
Warehouse Twin Command ontology and model register for the United States
The object model and the model and equipment register from Warehouse Twin Command: Live Slotting, Docks, Yards and Safety Vision for Warehousing and Intralogistics in the United States, an open reference architecture by CodeNinja Atoms, in warehousing and intralogistics. Part of the Vertical-Driven Architectures series; every design in the series is also a row in the cumulative dataset https://huggingface.co/datasets/CodeNinjatools/vertical-driven-architectures.
- Read the paper: https://codeatoms.ai/warehouse-digital-twin-us/
- DOI: https://doi.org/10.5281/zenodo.23556416
- Source files and PDF: https://github.com/muhammadumar89/codeninja-research
- Live view: https://huggingface.co/spaces/CodeNinjatools/warehouse-digital-twin-us
Files
| File | What it holds |
|---|---|
objects.json |
14 typed objects (Distribution Center, Dock Door, Yard Trailer, Storage Slot, SKU, Open Order, Pick Task, Associate, Labor Shift, Camera Zone, Safety Event, Forklift, Shift Lead, Slotting Recommendation), each with its properties and 14 typed links written from chapter 5. Format hyper-ontology/1: designed with CodeNinja Praxis, implemented with Hyper Ontology. |
models.csv |
The model and equipment register: every model, tool and hardware class with its licence and the reason for each choice. |
How to use it
import json
from huggingface_hub import hf_hub_download
p = hf_hub_download("CodeNinjatools/warehouse-digital-twin-us-ontology", "objects.json", repo_type="dataset")
objects = json.load(open(p))["objects"]
print([o["label"] for o in objects])
Citation
CodeNinja Engineering Team and Muhammad Umar. 2026. Warehouse Twin Command: Live Slotting, Docks, Yards and Safety Vision for Warehousing and Intralogistics in the United States. CodeNinja Atoms. https://doi.org/10.5281/zenodo.23556416. CC BY 4.0.
One of the CodeNinja Atoms sovereign AI reference architectures: every published design with its sector, country, compute and three-year cost. Its object model sits with the other industrial ontology examples.
- Downloads last month
- 27