OEE Summary¶
The Summary is OEE's landing screen. By default it shows the whole plant at a glance for a shift and a date — one overall OEE figure, which machines are performing worst, and how each shift row compares. From there you can drill into a single machine for the detailed view.
(This page was previously called the Dashboard, and showed only the single-machine view described further down. Old links still work.)
The plant summary (what you see by default)¶
You don't have to choose anything. The Shift is picked for you from Shift Mapping, the Date defaults to today, and the summary appears.
Overall OEE — four cards covering every machine on that shift: OEE, Availability, Performance and Quality. Beside them you'll see how many machines went into the figure, for example "Weighted across 4/6 machines measured".
Each card shows its figure in a ring that fills as the figure rises, and the card is tinted by how the figure is doing, so the one worth looking at stands out without reading every number:
| Card | Meaning |
|---|---|
| Green — On target | 85% or better |
| Amber — Needs attention | 65% up to 85% |
| Red — Below target | under 65% |
| Plain — No data | nothing logged, so there is no figure to judge |
Every card says which of these it is in words as well as colour, so the status is never carried by colour alone.
The same bands are used for all four cards
85% and 65% are the usual marks for OEE. Availability, Performance and Quality are judged against those same two numbers here, which is worth knowing because the usual expectations for those three individually are higher — a Quality of 90% is poor in most plants but shows as On target on its card. Read the colour as a guide to where to look first, not as a target your plant has agreed to.
Why it says 4 of 6
A machine with nothing logged against it has no OEE — not 0%. Those machines are left out of the overall figure rather than counted as zero, because a machine nobody logged anything for isn't a machine that performed badly. The count tells you how much of the plant the number actually covers: if it says 2 of 9, the headline figure isn't yet telling you much.
How the overall figure is combined
Overall OEE is not the average of the individual machine figures. It adds up the real minutes and piece counts across all the machines first, then works out the three parts from those totals. This means a machine that only ran for twenty minutes doesn't swing the plant figure as hard as one that ran a full eight hours — which is what you'd want, and is also why the overall figure can sit lower than the average of the numbers in the list below it.
Click any of the four cards to see that same measure for every machine on the shift, lowest first — so if Performance looks poor, one click tells you which machine is responsible. Machines with nothing logged are listed at the bottom rather than ranked as zeroes. Click a machine in that list to jump straight into its detailed view.
Underneath, the raw totals the cards came from are shown as two bars:
- Where the time went — the planned production time split into Runtime and Downtime. The runtime share of the bar is Availability.
- What came out — the total pieces split into Good, Scrap and Rework. The good share is Quality.
Each part shows its own figure and its share of the whole.
If downtime exceeds the planned time
The bar will say so. It almost always means a stoppage was left running — see Finding a stoppage you left open.
OEE history¶
Below the totals, OEE history is a grid: one row per machine, one cell per period, so you can see at a glance which machine has been struggling and when. The plant as a whole is the first row, with each machine underneath it.
- Range — 30d (the default), 90d, 180d or 365d, counting back from today.
- Machine — one machine, or All machines (the default).
- Shift — one shift, or All shifts (the default).
Hover any cell for that period's OEE, Availability, Performance and Quality.
What \"All\" covers
All shifts means the shift set up under Shift Mapping — not every shift the organisation has. Work logged against a shift that isn't mapped to OEE isn't part of OEE and isn't counted here.
All machines means the machines assigned to that shift in your Shift settings, falling back to every machine if none have been assigned — the same rule as the Machine list everywhere else in OEE (see Which machines appear in the Machine list).
If no shift is mapped to OEE yet, the grid says so instead of showing figures, because there is no OEE scope to draw.
Darker means lower OEE, so the cells worth your attention are the ones that stand out; a good period fades into the page. The scale underneath the grid shows the bands, which are the same 65% and 85% marks the cards at the top of the page use. A cell with no figure at all is drawn as a hatched square, never as a pale one — so "nothing logged" can't be mistaken for "nearly perfect".
Longer ranges are grouped
30 days shows one cell per day. Longer ranges group into weeks, and a full year groups into months — 365 separate columns can't be read at any size. The caption says which grouping is in use, and a grouped cell is worked out by adding up the real minutes and piece counts across the whole period and then re-working the figures, not by averaging the daily percentages. That matters: averaging would let a twenty-minute Sunday count as much as a full Monday.
History comes from the nightly summary
The grid is built from the summary Enture compiles each night, which is what lets a year load quickly. Periods the nightly job hasn't reached yet — today especially — show as Not summarised yet rather than as a gap or a zero, and a week or month that's still filling in is marked as such, because its figure will change as the rest of the days are added. Today's live figures are the cards at the top of the page; the grid is history.
If the grid says there is no history
The nightly summary covers the previous day, each night. It does not reach backwards on its own, so there is no history for any date before OEE started running at your site — a new setup shows an empty grid until the first overnight run, even while the cards at the top of the page are already showing today's figures. The grid says which of the two it is rather than just appearing blank.
If it stays empty after a few days of logging, ask your administrator to check that the nightly summary job is running. Earlier dates can be filled in on request — it's a one-off run over whatever date range you need.
Lowest OEE machines — the five worst performers, lowest first, so the machine most worth looking at is at the top. Click any machine in this list to drill straight into it. Machines with nothing logged are listed separately underneath rather than ranked, for the same reason as above.
Each machine also shows its Availability, Performance and Quality, with the lowest of the three marked as the weakest. That's the useful part: OEE is those three multiplied together, so the weakest one tells you where to go next — a weak Availability means stoppages, so head for the Downtime Log; a weak Quality means scrap, so look at Production Entry.
\"Not yet computable\" is not a low score
If one of the three shows as not yet computable, something needed to work it out hasn't been logged — it is not counted as the weakest. See When a figure is blank.
OEE by shift row — where a shift is made of more than one row (see A shift can have more than one row), each row gets its own figure, drawn as a ring inside a ring — the outermost ring is the first row, and the plant's overall OEE sits in the middle. So you can see whether, say, nights run worse than mornings. Each ring's figure is listed beside it, and a row with nothing logged shows a dash.
Read the numbers, not the ring lengths
Because the rings have different sizes, the same percentage draws a longer arc on an outer ring than on an inner one — so two rings can't be compared by how far round they go. Each ring shows how that one row is doing; the figures listed beside them are what to compare.
Downtime Pareto — on the summary this covers all machines for the date, so you can see what's costing the plant the most time overall.
Analyse — the four measures over time¶
Analyse, beside Filters at the top of the page, opens a view of OEE and its three parts charted over a date range. It answers a different question from the rest of this page: not "how is this shift doing", but "how have we been trending, and which machine is behind it".
Four charts, one under the other, in the order the figures multiply:
- OEE
- Availability
- Performance
- Quality
One set of filters at the top applies to all four, so they always show the same slice and can be read against each other:
- From / To — the date range. Defaults to the last 14 days.
- Shift and Shift Row — narrow to one shift, then to one of its rows. Shift Row becomes available once a shift is chosen.
- Compare machines — pick machines to chart individually.
With no machines picked, each chart shows the plant as a whole — one line, the same weighted figure the cards use. Pick machines to compare them and each gets its own line, keeping the same colour in all four charts so you can follow one machine down the page.
Up to four machines at a time
Four is the limit, and it's deliberate: beyond that the lines can't be told apart reliably — including by readers with colour-vision deficiency, who would see two series as one. If you pick more, the view charts the first four and says so. Compare in smaller groups.
Tip
All four charts use a fixed 0–100% scale rather than zooming to fit, so a small wobble looks small. That's on purpose — it's what lets you see that, say, Availability is the leg costing you OEE while Quality has been steady.
A break in a line is a day with no figure, not a zero — the line stops rather than dropping to the floor and inventing a reading that was never taken. Like the history grid, Analyse reads from the nightly summary, so it shows history rather than today.
Drilling into one machine¶
Click Filters, then choose a Machine and a Shift Row. The page switches to the detailed single-machine view. Click Clear to go back to the plant summary.
The Machine list shows the machines assigned to the selected shift in your Shift settings, or every machine if none have been assigned — see Which machines appear in the Machine list. The Shift Row list shows that shift's rows for the date chosen; if the shift has only one row, it's picked for you.
Once drilled in you get the same four cards for that one machine on that one shift row:
- OEE — the three parts multiplied together
- Availability, Performance and Quality — see the overview for what each one means
They read exactly as they do on the plant summary above — ring, figure, and a tint and word for how it's doing.
Underneath, the same Where the time went and What came out bars as the plant summary, for this one machine and shift row. The Downtime Pareto at the foot of the page narrows to just that machine.
The cards aren't clickable here — they're already a single machine, so there's nothing to break down.
Where the Planned figure comes from
Planned is usually the shift's length minus any stoppage marked as a planned stop. But if you've set up a Machine Production Planner entry for this exact machine, shift and date, Planned comes from that plan instead — the figure you planned ahead of the shift, not the shift's raw hours. See How each figure is actually worked out for the full explanation.
Note
If a card reads No data rather than 0%, something needed to work the figure out is missing. The page says which — see When a figure is blank.
Downtime breakdown¶
Below the cards, downtime for the chosen date is grouped by reason, longest first, with the number of times each reason occurred. This is what turns a low Availability into something you can act on: it's the difference between "the line lost three hours" and "the line lost three hours, almost all of it waiting for material".
Stoppages that were never given a reason appear as Unclassified. A long unclassified bar usually means operators are starting stoppages but not choosing a reason when they resume — worth correcting, because those hours can't be acted on.
The trend¶
Below the figures, OEE Trend charts the same machine and shift over time — OEE alongside Availability, Performance and Quality — with a switch for the last 7, 30 or 90 days. This is the question the cards can't answer: not "how did today go", but "is this getting better or worse".
Before you've picked a Shift Row, the trend shows the whole shift's figures for each day — every row's numbers added together and re-worked-out as one figure, the same way a machine-wide trend works. Once you pick a Shift Row, the trend narrows to that row specifically.
Use the table button beside the range controls to read the same figures as numbers instead of lines.
Where the figures come from
Past days are read from a summary Enture builds each night, which is why a long range loads quickly. Days the nightly job hasn't reached yet — today, most obviously — are worked out on the spot, and the chart says how many of those it's showing. Both are calculated the same way, so a day's figure doesn't change when the summary catches up with it.
If a day says it isn't summarised yet, that usually means the nightly job hasn't run for that date. It will fill in on the next run; the underlying logs are never lost.
Closing a shift¶
Close Shift marks the selected shift row as reconciled for that machine and date. It does two things:
- Closes any stoppage still running on that machine that falls within this row's own window, ending it at the moment you close the shift. The confirmation tells you how many were closed this way — if that number is regularly above zero, stoppages aren't being ended when the machine restarts.
- Records that this row was closed, and by whom, so there's a clear line between shift rows that have been checked and shift rows that haven't.
Each shift row can only be closed once, and closing one row never affects another — closing "Shift A" doesn't close or block "Shift B" for the same date. Closing a row does not freeze or store the OEE figures — they continue to be worked out from the logs, so correcting a mistaken entry afterwards still corrects the figures.
Tip
Close the shift at handover. It's the moment someone is looking at the numbers anyway, and it stops a forgotten stoppage running on into the next shift and distorting both.