Yield & Grading
The Borderline Bird Problem: Where Grading Margin Quietly Disappears
Every processing plant loses money on birds that should have graded A but didn't. Not because the birds were wrong — because the call was close, the line was moving, and the safe answer was B.
The call nobody talks about
Walk any grading station and you'll see the same thing. Most birds are easy calls. A clean carcass is an A. A torn skin flap or a bad bruise is a B. The grader doesn't hesitate, and they're right.
The problem is the bird in between — the minor discoloration, the small skin tear right at the spec limit, the bird that's an A on Monday morning and a B at hour seven of a shift. When a human is making that call hundreds of times an hour, the rational move is to downgrade. Nobody gets written off for being conservative. So borderline birds default to B.
Where the calls actually fall
The borderline zone is small as a share of birds — and it defaults to B under line speed and fatigue. That's the margin leak.
Small percentage, real money
The math is simple enough to run on a napkin. Take a plant running a typical line speed across two shifts. Even if only a low single-digit percentage of birds are borderline calls that get downgraded when they shouldn't be, you're multiplying the A-to-B price differential across thousands of birds, every day, all year.
Illustrative — run your own line's numbers
2% misgraded × A/B differential × annual volume
At commercial line speeds, this compounds into six figures per line, per year.
Every plant's numbers are different — line speed, bird size, your actual A/B price spread, your grader staffing. But the structure of the loss is the same everywhere: it's invisible. Downgraded borderline birds don't show up as waste, condemnations, or givebacks. They show up as B-grade revenue that should have been A-grade revenue, and no report flags it.
Why fatigue isn't a training problem
The usual response is to train graders harder or rotate them more often. Both help at the margins. Neither fixes the underlying issue: a human eye making subjective close calls at line speed will drift over a shift. That's not a discipline problem — it's biology. Consistency at hour one and consistency at hour seven are different things for any person on any line.
Machine vision doesn't drift. A camera grading every bird against the same spec applies the identical standard to bird one and bird one hundred thousand. The borderline bird that legitimately meets A spec gets called an A — at 7 AM, at 3 PM, on a Friday before a holiday.
What it takes to find out
You don't need to commit to anything to know what borderline birds are costing you. The honest way to evaluate a grading system is on your own product:
- Bench test on your birds. Not a demo reel, not someone else's plant — your carcasses, graded against your spec, side by side with your current calls.
- No line disruption. Our system mounts to the ceiling above the line and integrates with your existing PLC over EtherNet/IP. Installation is one day. Production doesn't stop.
- Judge the deltas. Where the machine and your graders agree, fine. Where they disagree, those are your borderline birds — and now you can count what they're worth.
We don't ask anyone to take performance on faith. We prove it, not promise it.
See what your borderline birds are worth
A plant walk takes under an hour. We'll look at your line, your grading station, and your spec — and tell you straight whether the numbers are worth pursuing.
Request a Plant Walk