Blog / OEE
OEE · Guide

OEE on a New Line: When to Start the Clock

OL By OEE Lab Editorial |Updated 2 November 2026 |6 min read
Short answer

Start the OEE clock at the first shift scheduled to make saleable product, normally after acceptance tests, because commissioning and planned trials are excluded time. During ramp-up, measure against the full accredited rate, expect a low number, and track progress with a separate ramp-up plan.

The clock starts when the line is asked to produce

OEE judges how well a line runs when it is scheduled to make product. While engineers are still proving the line, it is not yet being asked to do that.

Planned trials sit in the time outside OEE, and TEEP brings them into view.

This time model comes from Overall Equipment Effectiveness (R. C. Hansen, 2001). It gives you one simple test:

Is this shift on the production plan, with product meant for customers?

StageIn OEE?Log it?
InstallationNoProject plan only
Commissioning, dry and wet runsNoYes, snag list
Planned trials, acceptance testsNoYes: rates, stops, quality
Ramp-up, first saleable productYesYes, everything
Steady productionYesYes, everything
The line between "no" and "yes" is the handover to production. Agree its date before the project starts.

Grey cases get decided by the plan. A trial that makes saleable product on a production schedule counts, and it should be booked as production from the start.

A trial added at short notice to explain a fault is another matter. If the line was on the plan to produce that shift, the time is a stop.

Log everything before the clock starts

Excluded from OEE does not mean unmeasured. Commissioning data gives you the first look at how this line really fails, and it costs little to collect.

Connect your counts and stop signals during commissioning, so you know they work before they matter. PLC signals for OEE and machine state detection show what to test.

Which ideal rate during ramp-up?

This is the decision that sets the tone for the whole project. Use the rate the line is accredited for, for each format, from day one of counting.

  • Best source: the rate shown in acceptance tests on your own product.
  • If you have none yet: the design rate for that format, marked as provisional and reviewed after the first weeks.
  • Not this: the ramp-up target for the week. That is a plan and not a capability.

The rules for choosing a rate are in ideal cycle time. What a soft rate does to the number is easy to show.

Example: in week 4 of a ramp-up a line is scheduled for 2,400 minutes and makes 115,200 good units. Its accredited rate is 120 per minute, and the ramp plan for that week is only 80.

Against the accredited rate, 120 per min40.0%
Against the week's plan rate, 80 per min60.0%
Same week, same output. The higher number only looks better, and it falls the day the rate is raised.

Keep the fixed rate. Report the plan as a separate number, as output against the ramp plan, and see plan attainment.

That gives two honest statements side by side: "we beat the ramp plan this week" and "the line is at 40% of what it can do".

What the first weeks look like

Here is that week 4 in full, so you can see how the 40% breaks down.

Scheduled2,400 min
Stops900 min
Slow running420 min
Scrap120 min
Good output960 min
Week 4: availability 62.5%, performance 72.0% and quality 88.9%, which multiply to 40.0%.

The factors are 1,500 ÷ 2,400 for availability, 1,080 ÷ 1,500 for performance and 960 ÷ 1,080 for quality. Ramp-up weeks often lose in all three at once.

Plot OEE every week against the handover date. The shape you want is a rising curve, and the slope tells you more than any one point.

100%50%0%Week 4: 40%14812
Hypothetical curve, measured in weeks from handover. Your line's shape will differ.

Reason codes for a new line

A new line fails in new ways, so your normal code list will not fit. Add a few codes for the ramp-up period.

  • Snag list item, tied to the item on the project's list. This is what you use to hold suppliers to their work.
  • Waiting for supplier or engineer. Often a big loss early on, with its own owner.
  • Recipe or setting not final. Stops to tune settings are real losses, and they should fall quickly.
  • Training. Operator-caused stops show where more practice is needed, with no blame attached.

Keep these inside your wider reason codes, and drop them when ramp-up ends. Log a lower stop threshold than usual at first, as short stops are common and cheap to fix now.

Do not confuse this period with startup losses. That page covers the minutes after each restart in normal running, and it carries on for the life of the line.

When does ramp-up end?

1
Holds the target
OEE at or above the agreed ramp-up target for four weeks in a row.
2
All main formats
Each main product has run at its proven rate, not only the easy one.
3
Snag list closed
Open items are closed or moved to normal improvement work.
4
Normal crew
The line runs on normal crews without project engineers on the floor.
Our suggested exit test. Agree yours with production and the project team before the handover.

Until then, report it apart from mature lines. One new line at 40% can pull a plant roll-up down hard, and that hides how the old lines are doing; why OEE drops after a new machine shows the sum.

After that, set normal staged goals using how to set OEE targets. The first long-term losses to work on come out of your OEE loss tree.

How ramp-up numbers get spoiled

  • Waiting until the line is "stable". You lose the early data, and nobody can say what stable means.
  • Calling problem days trials. A planned trial is on the plan before it starts.
  • Lowering the ideal to hit a project goal. It moves the yardstick and makes the trend useless.
  • Using the nameplate of another product. A rate on water is not a rate on sauce.
  • Mixing it into plant OEE silently. Show the plant figure both with and without the new line.
See it live in Fabrico

A ramp-up changes week to week, so a live record of stops and rates helps more than a weekly summary. Fabrico is a light MES for OEE that takes stops, speed and counts live from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line. See how Fabrico works as a light MES for OEE.

What people ask next

Do commissioning and trial runs count in OEE?

No, if they were planned as such and no saleable output was scheduled. They are excluded time, as in planned vs unplanned downtime, but they do count in TEEP.

What OEE should I expect from a new line?

We know of no neutral figure for ramp-up curves, so we will not quote one. Agree the curve with your supplier and team up front, and use how to set OEE targets for staged goals.

Which comes first, OEE or the acceptance test?

Run the acceptance test first, as it proves the rate. Start OEE at the handover that follows, using the rate you just proved. Ideal cycle time explains how to record it.

Should the ideal rate rise during ramp-up?

Only when a new rate is proven. Mark the date on the chart, since OEE dips when the rate rises even though nothing got worse.

Does a new line count in plant OEE during ramp-up?

It counts once it is scheduled for production, but show the plant figure both ways. The sum is in how to average OEE across machines.

See OEE per line and shift on your own machines

Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.

Book a Fabrico demo

← More OEE Lab guides · Why OEE drops after a new machine →

OEE essentials: OEE calculator · What is OEE · Six big losses · Availability · TEEP · Cost of downtime · OEE benchmark · Free OEE tracker · Free OEE software · How to track downtime · How to implement OEE · Light MES vs full MES