Machine monitoring is the automatic collection of a machine's operating signals, running versus stopped, cycle counts, speeds, temperatures, and fault codes, to build a real-time, accurate picture of what equipment is actually doing. It replaces the hand-written log and the "I think it ran okay" with an objective record. And because the signals can come from several sources, you can start monitoring a 30-year-old machine without replacing it.
Where Do the Signals Come From?
A machine broadcasts its state in more ways than most plants use. The main sources, roughly from richest to simplest:
- The PLC. The programmable logic controller already knows the machine's state, running, faulted, cycle count. Reading its tags is the richest, most direct source when access is available.
- Added sensors. Vibration, temperature, proximity, and current sensors bolt onto equipment that cannot be tapped directly.
- Power draw. A current clamp on the machine's supply distinguishes running, idling, and off with almost no integration.
- The stack light. The simplest of all, read the green/yellow/red tower the machine already has, and you know run, warning, and stop.
What Should You Monitor First?
Not everything. The plants that succeed start narrow and grow. A sensible order:
- Run vs. stop on the constraint. Monitor the bottleneck machine first, the one that sets plant output. Uptime here matters most.
- Cycle count and rate. Are we running at the speed we think we are? Slow cycles are hidden losses.
- Stop reasons. Pair automatic stop detection with an operator reason code to feed downtime analysis.
- Condition signals. Add vibration and temperature once the basics are trusted, to move toward predictive maintenance.
Why Automatic Beats Hand-Logged
Hand-logged machine data has two fatal flaws: it is incomplete (the biggest events happen when nobody has a free hand to write) and it is disputed (two people remember the shift differently). Automatic monitoring removes both. The number is the number, captured the same way every shift, so the OEE meeting stops being an argument about the data and starts being a conversation about the process. That objective baseline is the real product of machine monitoring, it feeds OEE reliability, and scheduling with a single trusted source.
By the Numbers
The recoverable losses that monitoring exposes are large and well documented, the U.S. Department of Energy's operations and maintenance research ties significant capacity and energy losses to unmonitored, reactive equipment management (PNNL O&M Best Practices). You cannot recover a loss you never recorded. Where Harmony fits: Harmony connects to your machines, through the PLC, sensors, or simpler signals, and streams their state into one operational layer alongside your other systems, so monitoring is not a standalone dashboard but part of the whole plant picture. See smart factory technology or a real deployment.