What counts as a unit
01
The item the process is measured against. Counting at the batch rather than the element hides single-unit failures inside an otherwise acceptable batch.

Advanced Manufacturing · Metrics
First pass yield is the proportion of items that complete a process correctly the first time, with no rework, repair or reprocessing. It is expressed as a percentage of units started, and it counts only those that pass without intervention.
The number matters because it separates output from effort. Two plants can ship the same quantity in the same week while one did it once and the other did a portion of it twice, and only first pass yield makes that difference visible.
Zabidou measures every item as it is produced, so the figure is built from the line itself rather than from what was written down afterwards.
Zabidou in production today
QLD + VIC
Operating in factories across Queensland and Victoria today.
100%
Inspection coverage on finished items as they come off the line.
Silver
Associate Member, Australian Steel Institute.
Bronze
Member, National Precast Association Australia.
First pass yield is simple to define and easy to calculate in a way that flatters. Three decisions determine whether the figure means anything.
01
The item the process is measured against. Counting at the batch rather than the element hides single-unit failures inside an otherwise acceptable batch.
02
Passing means completing without rework, repair or concession. Where a minor touch-up is treated as normal production rather than rework, the figure rises without anything improving.
03
A yield measured at the end of the line tells you the plant total. Measured at each step it tells you which step is costing you, which is the part that can be acted on.
First pass yield is one of the most quoted numbers in a plant and one of the least reliable, for reasons that have nothing to do with anyone being dishonest.
01
A correction made at the station in a few minutes rarely generates a record. It is absorbed into the normal rhythm of the shift, and the yield figure never sees it.
02
A fault created at set-out and found at demould is usually logged against demould. The step that caused it keeps a clean figure and the improvement effort goes to the wrong place.
03
Where pass and fail are judged by eye, the yield figure inherits that judgement. It varies between inspectors and between shifts, which makes trends difficult to trust.
A first pass yield figure is only as good as the data underneath it. Zabidou produces that data from the line itself, measuring every item as it is made.
Where every item is measured as it is produced, pass and fail stop being a judgement and become a comparison against a stated tolerance. The figure is then the same number regardless of who was on shift.
Measuring at the step also attributes failures to where they occurred rather than where they surfaced, which is the difference between a yield report that describes a problem and one that locates it.
Zabidou measures each item inline against a Golden Sample or the drawings and holds the result against that item, so first pass yield is derived from measurement rather than from recollection. The same check runs on a precast concrete bed and on a steel fabrication line.
For the plant, that means a yield figure the production meeting can act on, with the step responsible named rather than argued about, and a trend that holds up when the number is challenged.
Measuring a number does not by itself improve it, and the gain is in finally having a figure worth acting on.
Every failed item is a potential non-conformance, the first article inspection sets the reference the rest are measured against, and automated visual inspection is what makes checking every item practical.
Five checks run on every item as it is produced, and each one contributes measured data to the yield figure rather than an opinion about it.
Complete 3D measurement and verification of as-built output against a Golden Sample or design intent, for full dimensional compliance without stopping the line.
Automated, real-time verification that components are correctly positioned before and during production, while the mould is still open or the section is still in the jig.
Inline detection of surface defects, voids and irregularities at industrial speed, fully automated and without manual inspection.
Precise measurement and verification of critical dimensions against a Golden Sample or the drawings, inline, in real time, at scale.
Manufactured-unit centric digital provenance that automatically delivers the full audit history operations, financiers and customers demand.
These four get used interchangeably in conversation and mean materially different things on a report.
| First pass yield | Rolled throughput yield | OEE | |
|---|---|---|---|
| What it measures | Units through one step with no rework. | First pass yield multiplied across every step. | Availability, performance and quality combined. |
| What it reveals | Where a single step is losing units. | The true probability of a defect-free unit end to end. | How much of the planned production time became good output. |
| Where it misleads | A high figure at one step hides the plant total. | A long process gives a low figure even when each step is strong. | A single quality factor can disguise a serious step-level problem. |
| What improves it | Removing the cause of rework at that step. | Improving the weakest step first. | Any of availability, speed or quality. |


First pass yield is the percentage of units that complete a process correctly the first time, without rework, repair or reprocessing. It is calculated as units passing without intervention divided by units started.
It is sometimes called first time yield or right first time, and the three are generally used interchangeably.
Divide the number of units that completed the step with no rework by the number of units that entered it, then express the result as a percentage. A step that started 100 units and completed 94 without intervention has a first pass yield of 94 per cent.
The common error is counting reworked units as passes, which produces a figure closer to throughput than to first pass yield.
First pass yield applies to a single step. Rolled throughput yield multiplies the first pass yield of every step together to give the probability that a unit passes the whole process without intervention.
A process of five steps each at 95 per cent has a rolled throughput yield of about 77 per cent, which is why long processes look worse than any individual step suggests.
Quality is one of the three factors in overall equipment effectiveness, alongside availability and performance. First pass yield is the factor that sits closest to that quality component.
OEE describes how much planned production time became good output. First pass yield describes how much of the output needed a second attempt.
It depends entirely on the process, so a target taken from another industry is rarely useful. What matters more is whether the figure is measured consistently and whether it is trending in the right direction.
A figure that has not moved in a year usually indicates a measurement problem rather than a stable process.
Ready to explore how ultra-vision can transform your operations? Send us a message and our team will be in touch.