[!NOTE] TL;DR: For my third apprenticeship mission, I stepped away from pure software to lead an industrial automation audit. The objective: replace an outdated telemetry system with a modern Edge Computing architecture. On the agenda: reverse engineering 168,000 lines of logs, evaluating MQTT vs. OPC-UA protocols, and exercising technical diplomacy among diverging business stakeholders.
Can a developer ensure the success of an industrial project without ever leaving their IDE? The answer is no. To truly understand IT/OT convergence, one must physically engage with hardware on the floor. With this in mind, I led an audit for an industrial client in the home decor sector whose aging telemetry system needed replacing.
Modernizing a Legacy System Through Edge Computing
The initial requirement was clear: the client operated machinery using obsolete hardware managed by an external vendor. The goal was to stream reliable real-time production metrics directly to a centralized Cloud system (MES).
However, the challenge went beyond replacing hardware at a single pilot site. My role was to lay the foundation for a standardized, secure, and scalable architecture doctrine. I defined hardware acquisition using specific industrial controllers, electrical isolation strategies, and secure data transport. This was a natural extension of my architectural refactoring work on the APS integration.
Reverse Engineering & Mass Data Log Processing
To deliver an audit document that serves as a contractual baseline, I couldn’t rely on assumptions. The most intensive phase was reverse engineering.
I analyzed existing field hardware, then extracted and processed real machine log files. This involved processing over 168,000 lines of data across three heterogeneous formats. This evidence-based approach allowed me to define the exact compute distribution between local (Edge) devices and the Cloud, relying strictly on manufacturer specifications.
The IoT Tradeoff: MQTT vs. OPC-UA
Designing a target architecture often requires choosing between competing industrial standards. On this project, a key technical disagreement emerged regarding data collection from the controller:
- On one side, pushing an event stream via MQTT from the gateway to the Cloud.
- On the other side, synchronous polling via OPC-UA directly inside the controller.
Instead of imposing my preference, I objectively framed the debate. I documented both scenarios (benefits, limits, architectural implications) and presented them for decision.
Technical Diplomacy at the Heart of the Audit
This is where emotional intelligence (soft skills) supersedes pure technical skills. Engaging with the client’s production manager, sales lead, and engineering teams—each with conflicting priorities—requires technical diplomacy.
The audit was finalized across 6 iterative review sessions with the client (from v1.0 to v1.5), ensuring recommendations reflected operational realities rather than just theoretical models.
Key Takeaways
Humans Remain the Best Communication Protocol
This mission proved that my highest value lies where algorithms end. Demystifying advanced concepts (protocols, event-driven architecture) for non-specialist stakeholders is an essential skill. Translating complex business needs into precise software architectures remains a domain where humans hold a decisive advantage over AI (as discussed in my generative workflow analysis).