On a grinder, not cutting is often still working.
Dress cycles, spark-out, gauging pauses and thermal holds are all part of producing the component. Treat them as downtime and the number will be disputed — correctly.
Grinding breaks the assumption that a machine which is not removing metal is not working. A dress cycle, a spark-out, a gauge-and-compensate pause and a thermal stabilisation hold are all part of producing the component, and a monitoring system that treats them as downtime will report a grinder as chronically unavailable while it is doing exactly what it should.
This is the machine type where naive monitoring most reliably produces an argument with the shop floor.
| State | What is happening | How it must be treated |
|---|---|---|
| Dress cycle | Wheel being trued or dressed, no part being cut | Productive — part of the process, not a stop. Trend its frequency. |
| Spark-out | Wheel in contact, minimal removal, achieving size and finish | Productive. Removal rate near zero is the intended condition. |
| Gauge and compensate | In-process gauging, offset applied, cycle resumes | A pause inside a cycle, not the end of one. |
| Thermal stabilisation | Machine held at temperature before precision work | Planned, and should be visible as its own state rather than as idle. |
| Wheel change | Wheel replaced and balanced | A genuine stop, and one worth measuring separately from breakdown. |
If a grinder spends fifteen per cent of its productive time dressing and sparking out, and monitoring counts that as downtime, the reported availability is wrong by fifteen points. The supervisor knows the machine was working, disputes the number, and reasonably concludes the system is unreliable — which then discredits the figures that were correct.
The fix is not clever software. It is agreeing at commissioning which states are productive on this machine and mapping them, which requires someone who has ground components to be in the conversation. This is the clearest example on the site of why machine-type-specific configuration is not a marketing distinction.
Wheel changes and balancing
Infrequent but long. Worth measuring separately so wheel life and changeover discipline improve independently.
Setup and first-off on precision work
Grinding tolerances make first-off approval slower than on milling or turning. Frequently the dominant loss on short batches.
Waiting for inspection
A finished component held pending measurement, on a machine that is ready to continue.
Coolant and filtration attention
Short, frequent and almost never logged, but material on a floor running continuous grinding.
A networked grinding control reports machine state, cycle timing, counts and alarm presence. What it does not do is label which of the states above it is in — dress, spark-out and a genuine idle can appear the same from outside the control unless the machine is configured to distinguish them.
Where a dress cycle is signalled — by a program call, a discrete output or a mode indication — it can be mapped and treated correctly. Where it is not, the honest options are inferring it from cycle structure, which is a model rather than an observation, or accepting a coarser state picture and saying so. We would rather tell you the second than present the first as fact.
Two grinders on a live dashboard this week.
Bring a machine with a dressing-heavy process. Getting its productive states mapped correctly is the whole exercise, and it is worth doing before anyone sees an availability figure.
Book a free 2-machine pilot →WhatsApp us