DigitalFactoryLab is a complete industrial-data stack — from station-level telemetry to live analytics — designed, built and operated end-to-end by one engineer, and left running 24/7 for anyone to inspect.
Most shopfloor-visibility claims come as slideware: an architecture diagram, a few screenshots, a promise. This project takes the opposite approach. A simulated EV-powertrain plant — a rotor line, a traction-inverter line, an assembly line and a low-volume hypercar cell — streams OPC UA and MQTT telemetry through a MonsterMQ broker into PostgreSQL, at roughly 2.1 million rows a day. The dashboards you can open from this site query that database live. Nothing on this site is a recording.
The point is simple: if the architecture works, you should be able to open it at 3 a.m. on a Sunday and watch parts move through stations. You can. Not screenshots — the real thing, running now.
Built end-to-end by one engineer
Simulation engine, OPC UA servers, MQTT namespace design, broker configuration, ingestion pipeline, database schema and indexing, every dashboard and this website — one person, one coherent system. That is deliberate: the value of a reference architecture is that every layer was chosen and wired for a reason, and the person behind it can explain each one.
The background behind it is real industrial delivery — electric-motor and power-electronics manufacturing programs and enterprise shopfloor-IT projects — distilled into a system small enough to understand and honest enough to trust.