LIVE DEMONSTRATION
UNS Smart Factory
Supervisor Gateway
This dashboard represents a Unified Namespace (UNS) architecture implementation.

Hardware Source: Data is streamed in real-time from an ESP32 embedded gateway simulating high-frequency industrial machine states.

Connectivity: Edge → MQTT Broker → Web Supervisor.

Note: All production data shown is simulated for demonstration purposes and does not represent an operating factory.
Node.js
Socket.io
ECharts
MQTT
ESP32
Invalid Access Credentials
Skip Login & View Profile →
UNS SUPERVISOR | Demo Cell
----/--/-- --:--:--
Day Output (PCS)
Power Load (W)
Cycle Time (S)
NG Rate (%)
📦 Last Piece Trace
WAITING
SERIAL NUMBER--
Model SKU--
Cycle Time--s
Torque--Nm
Angle--°
Diagnostic Code-
📊 OEE Trend (7 Days)Drag Slider
Environment
■ Temp ■ Hum
Dynamics
■ RPM ■ Flow
Avg Yield
98.5%
Avg OEE
86.2%
Energy (kWh)
4,230
Avg CT (s)
32.5
🔴 NG Pareto Analysis
🌓 Shift Performance
💠 Cycle Time Stability
Timestamp Serial No. (SN) Model Result Torque (Nm) Angle (°) CT (s) Error Code

UNS DataLink Engine

Connect Any Device · Deploy Anywhere · Data Freedom

1. Concept Overview

THE VISION

In traditional manufacturing, data is trapped in silos (PLCs, SCADA, MES, ERP). This is the "Integration Mess." UNS DataLink breaks these silos by implementing a Unified Namespace (UNS) architecture.

What is UNS?

It is a single, centralized location where all factory data exists. We structure data using the ISA-95 standard hierarchy:

Enterprise / Site / Area / Line / WorkCell

By acting as a central "Nervous System" (via MQTT Broker), smart devices publish data only once, and any application (Dashboard, AI, ERP) consumes it instantly. No more point-to-point integration nightmares.

Architecture Diagram

2. Core Business Value

DRIVERS
Data Assetization
Pain: Data locked in 'Black Boxes' (proprietary PLCs), requiring expensive fees for API access.
Gain: Return data ownership to you. Clean, standardized data is the fuel for future AI models.
Scalable & Low Cost
Pain: Traditional MES costs millions with slow deployment and rigid structures.
Gain: Decoupled architecture allows "Start Small, Scale Fast." Reuse existing hardware to maximize ROI.
Real-Time Visibility
Pain: Delayed reports. Finding out about scrap rates hours after the shift ends.
Gain: Second-level latency. Shift from "Post-mortem Analysis" to "Active Intervention."
AI Ready Infrastructure
Pain: AI projects fail due to unstructured data and protocol mismatches.
Gain: The UNS structure solves the "Last Mile" connectivity problem, making data Plug-and-Play for ML.
Business Value

3. Workflow Architecture

HOW IT WORKS
1
Edge Access & Cleaning
Universal Connectivity (Modbus, OPC UA, S7). Edge Governance includes deadband filtering, unit conversion, and anomaly detection to save bandwidth.
2
Unified Transport (MQTT)
Data is published to the Unified Namespace using the Topic structure: Site/Area/Line/Cell. Supports lightweight, bi-directional communication.
3
Contextualization
Raw data is enriched with context (Shift info, Order ID) before being stored or visualized.
4
Full Stack Visualization
Real-time Monitoring (Heartbeat), Analytics (Decision Support), and Traceability (Production Replay) as seen in this demo.

4. Approach

KEY ADVANTAGES
Extreme Flexibility
Hardware Neutral (Pi/IPC/Server). Environment Free (LAN/Cloud). Software migrates as network changes.
Self-Service Config
No expensive customization. Drag-and-drop Low-code backend. Build links without writing code.
UNS Ready
Standardized Topic Structure (Site/Area/Line/Cell). Foundation for Digital Twin.
Configuration

5. About the Author

TECHNICAL SHOWCASE
Carlos Chen
Carlos Chen
Industrial IIoT Enthusiast
This project is a Proof of Concept (PoC) demonstrating how open-source technologies can achieve real-time performance comparable to expensive industrial SCADA systems.

Open for technical discussions and architecture exchanges.
LinkedIn