One of the most intensive phases of my internship was setting up the communication pipeline between trash rack cleaners and the cloud supervision platform. My mission was clear: extract performance metrics and stream them via the MQTT protocol.
The Schneider Electric Architecture: Industrial Computing Power
The project relied on a dual PLC setup with impressive hardware specs:
- Modicon M340 (Server): A processor capable of handling 1,024 I/O points, with a cycle execution speed of 0.12 µs per boolean instruction.
- Modicon M262 (IIoT Gateway): A next-generation “Edge” controller with a 0.5 ms execution cycle and 32 MB of RAM dedicated to program logic, featuring native MQTT and HTTPS support.
PLC Programming: Mastering Structured Text
Development required versatility across industrial programming languages (IEC 61131-3):
- Structured Text (ST) (~60%): Used for all complex computation logic in the M262, initializations, and fault handling routines.
- Grafcet / Sequential Function Chart (SFC) (~30%): Orchestrating operational modes.
- Ladder Diagram (LD) (~10%): Strictly for simple transition conditions.
Challenge: Data Extraction & the “Memorable” Address Bug
Communication was structured in two steps:
- Locally: Using Modbus TCP polling scanner. The M262 pulled data from the M340 across a dedicated local network (84.0.2).
- Addressing: A critical hurdle. I will long remember an addressing mismatch between decimal and hexadecimal offsets (%MW23006 vs 16#59DE) that taught me absolute precision in industrial automation.
Component Supply Bottlenecks & Agility
During the global chip shortage, sourcing an M262 in stock was near impossible (lead times announced between 3 to 6 months). We demonstrated agility by prototyping the architecture on simulators and backup hardware (M580) before deploying to the final target hardware acquired in extremis.
“Transitioning from raw memory addresses to JSON messages via MQTT was the true leap into Industry 4.0.”