The VMC is simple. What surrounds it is not.
Cycle structure and the five things that break it, the real job-shop loss profile, and why setup must be measured as its own state rather than as downtime.
The vertical machining centre is the default machine of Indian general engineering, and the one most monitoring projects start with. Its cycle structure is the simplest of any CNC — usually one program run, one component — which makes it the machine where a plant's first honest OEE figure is most likely to be trusted.
That simplicity is also why VMC data exposes scheduling problems so clearly. When the machine itself is straightforward, what remains in the loss profile is everything around it.
| Situation | Effect on counting | Handling |
|---|---|---|
| One program, one component | None — the default assumption holds | Count program completions |
| Fixture holding several blanks | Output understated by the fixture count | Parts-per-cycle multiplier set per part number |
| Program stop for in-process gauging | Cycle appears to end and restart | Minimum-cycle threshold and pause tolerance, learned per machine |
| Tool change mid-cycle | Brief pause inside a running cycle | Absorbed by pause tolerance; not a stop |
| Cycle interrupted for chip clearing | Two cycles counted where one part exists | One part, one cycle, one stop recorded inside it |
Setup and first-off approval
The dominant loss in job-shop work. A three-hour changeover on a machine running two-day batches is a scheduling problem worth measuring precisely.
Waiting — material, crane, fixture, operator
The category that surprises plants most. On unmonitored floors the machine is frequently ready and waiting for something that is not the machine.
Shift-boundary losses
Late starts, early cleanups and the gap at handover. Individually defensible, collectively often the single largest recoverable block.
Tool changes and offset corrections
Short and frequent. Worth separating from setup so the two improve independently.
A networked VMC control provides machine state, cycle timing, part counts and alarm presence, with operator identity where enabled. That covers availability and output completely.
What it cannot supply is why the machine sat idle for forty minutes between two cycles, and on a VMC that is where most of the recoverable loss lives. A machining centre that is switched on and idle looks the same to a control whether it is waiting for material, waiting for a crane or waiting for a supervisor's approval. Reason capture at the machine is what separates those three, and they have completely different fixes.
OEE on a VMC, both ways.
The standard definition first, then the variant MachineWise computes and why it differs. Quote the ISO figure externally; use the variant internally.
| ISO 22400-2 — the standard | MachineWise variant — internal | |
|---|---|---|
| Availability | Planned production time 435 min, actual production 341 min → 78.4% | Reporting time 435 min, machine on-time 372 min → 85.5% |
| Performance / Effectiveness | 212 parts × 88 s ideal cycle ÷ 341 min → 91.2% | Productive (in-cut) time 268 min ÷ 372 min on-time → 72.0% |
| Quality | 206 good of 212 → 97.2% | 206 good of 212 → 97.2% |
| Result | 69.5% | 59.9% |
Note the gap between the two Performance figures. The ISO measure says the machine ran close to its rated speed while cutting; the variant says it spent a substantial share of its powered time not cutting at all. Both are true, and together they point at setup and waiting rather than at feeds and speeds. The two numbers answer different questions and are not interchangeable. The ISO figure is comparable to published benchmarks; the variant is not, and should be treated as your own baseline to improve against. Both are computed side by side in the OEE calculator.
A VMC changeover is a genuine event: fixture off, fixture on, indicate, tool offsets, program proving, first-off inspection, approval. Two to four hours is common in Indian job-shop work, and on short batches it can exceed the run time it enables.
That makes changeover measurement more valuable on a VMC than almost anywhere else, and it needs to be measured as its own state rather than as downtime. A machine in setup is not broken, and an availability figure that conflates the two will be disputed by the first supervisor who sees it — correctly. Mapping the mode selector at commissioning prevents that argument entirely.
Two VMCs on a live dashboard this week.
Start with your bottleneck machine and the one with the most changeovers. The contrast between the two usually settles the rollout plan by itself.
Book a free 2-machine pilot →WhatsApp us