Skip to content

LambertThink.md

Latest commit

 

History

History
61 lines (40 loc) · 3.69 KB

LambertThink.md

File metadata and controls

61 lines (40 loc) · 3.69 KB

LambertThink - Development Philosophy & DevLog 🧠

🌐 English Version | 繁體中文版本

"This is not just a stack of code; it is an experiment in the fusion of Industrial Automation and Edge Computing."

This document records my core thinking process during the three-month development of the Wago-Data project. My goal was not merely to create examples, but to cultivate a modern Industrial IoT Tech Stack on the foundation of WAGO controllers.

💡 Why go this "Crazy"?

In the traditional automation world, Programmable Logic Controllers (PLCs) are often viewed as closed-box logic switches. From my perspective, a WAGO controller—equipped with a Linux kernel and Docker support—is actually a hidden edge server inside a power distribution cabinet.

Over the past 90 days, my development has followed three core philosophies:

1. Data Sovereignty and "Near-Field" Processing

I do not believe in blindly pushing all raw data to the cloud.

  • The Necessity of SQLite: Data should be structured and stored locally first. This mitigates the risk of network outages and provides an "on-device" historical record for self-diagnosis.
  • Data Cleaning: Processing data at the edge and only uploading critical results is the hallmark of an efficient architecture.

2. Node-RED: The Bridge between IT and OT

Node-RED serves as the "Digital Glue" of this project.

  • It allows a PLC to speak to Discord and interact with Webhooks for the first time.
  • When a hardware alarm becomes a real-time mobile notification, the value of OT (Operational Technology) finally becomes visible to IT-level management.

3. From Standalone to Cluster: The Revelation of Docker Swarm

The move toward Swarm orchestration marks a major turning point in my thinking.

  • High Availability (HA): A factory shouldn't lose its data pipeline just because one controller fails. By clustering, I can make services "hop" between hardware, achieving true industrial-grade server resilience.

🚀 The Path to Advanced Edge Computing

My goal is to build a bottom-up comprehensive system:

  1. L1 - Perception Layer (Codesys): Ensuring stable communication and logic—the source of all data.
  2. L2 - Mediation Layer (Node-RED & SQLite): Processing, transforming, and persisting data to make it "talk."
  3. L3 - Orchestration Layer (Docker Swarm): Integrating multiple nodes into a single entity to provide server-level computing resources.
  4. L4 - Integration Layer (Cloud Hybrid): Connecting local clusters with private or public clouds for cross-regional management.

🛠 Solving Pain Points

  • Cross-Domain Communication: Resolved difficulties in Modbus access across different network segments and gateways.
  • Discord Integration: Bridged the gap between industrial protocols and modern encrypted APIs.
  • Scheduling Flexibility: Broke the limitation of needing to re-compile PLC code for logic changes by releasing operational flexibility through a visual interface.

🤖 The AI Frontier: The Next Conquest

Once the Swarm cluster is stable, the next step is to give these controllers the ability to "think." I plan to introduce Industrial AI to move beyond simple logic toward:

  • Predictive Maintenance: Letting equipment "call for help" before a failure occurs to avoid downtime losses.
  • Edge Vision: Integrating image processing (OpenCV/YOLO) to give the controller visual judgment capabilities.

We don't necessarily need expensive cloud storage. Through software-defined architecture, we can build powerful and flexible monitoring centers right at the edge.

Lambert Written on January 30, 2026 - The 3-Month Project Milestone