Harvest Quality Control

Batches inspected

–

Downgrade rate

–

Superior-grade yield

–

passed to spec first time

Points out of control

–

weeks beyond ±3σ

Harvest volume

–

gutted weight

Downgrade rate by week — p-chart with control limits

Weekly downgraded-batch proportion vs the centre-line and ±3σ limits (limits widen when a week harvests fewer batches). Points outside the band are special-cause — a sea-lice treatment, a bloom, a crowding incident — not routine variation. Limits recompute on your current filter; n small ⇒ wide limits.

Weekly harvest — to spec vs downgraded

Tonnes gutted weight landed per week, split by QC outcome

Top reason

–

Top reason code

–

Vital few

–

codes driving 80% of downgrades

Active codes

–

distinct downgrade codes

Downgrade code Pareto

Each code's share of downgrades (bars) + cumulative % (line) · dashed = 80% line · the vital few on the left

Reason mix

Share of downgrades by defect family · click a slice to drill

Superior-grade yield

–

across the current filter

Worst pen

–

Best pen

–

Pens above average

–

worse than the group mean

Superior-grade yield by pen

Yield = 1 − downgrades ÷ batches, per pen, grouped in site order · coral bar = worst pen · min. 30 batches

Downgrade rate by site

Where downgrades concentrate across the farm estate · click to drill

Worst feed lot

–

Lots above average

–

worse than the mean

Group downgrade rate

–

baseline

Batches traced

–

to a feed supplier lot

Downgrade rate by feed supplier lot

Worst 15 lots by downgrade %, min. 30 batches · redder = worse · feed lot is the strongest upstream input signal you can trace

Lot volume vs downgrade rate

Each dot a feed lot · x = batches fed · y = downgrade % · top-right = high-volume & high-downgrade = raise with the supplier first

Total value loss

–

price gap vs superior spec

Loss per tonne

–

across all harvest

Costliest reason

–

Trim-out vs re-grade

–

flesh-quality vs cosmetic

Value loss Pareto — by reason

Total £ lost by reason (bars) + cumulative % · this re-ranks against the count Pareto — the most frequent defect is not the most expensive one

Value loss by pen

Where the money is lost across the estate · top 12 pens

Concentration (Gini)

–

across reason × site

Open actions

–

corrective actions raised

Action closure

–

downgrades with closed action

Fix first

–

highest frequency × value

Defect concentration — reason × site

Downgrade count per reason × site · darker = more · is it one cell or spread across the estate?

Corrective-action priority

Each dot a downgrade code · x = frequency · y = value lost · top-right = fix first · size = open actions

Worst harvest shift

–

Shift spread

–

best → worst (pp)

Worst smolt intake

–

Grader spread

–

find-rate range (pp)

Downgrade rate by harvest shift

Crowding and pumping at night bruises more fish than it does at first light

Downgrade rate by smolt intake batch

Production cycle, oldest intake first · longer at sea = larger, more mature fish = more downgrades · dashed = group average

Downgrade rate by QC grader

Find-rate per grader vs the group average (dashed) · variation in grading, not necessarily fault — but a grader well below the line is passing what others catch · min. 30 batches

Harvest QC records

Aggregated on current filters · click a header to sort · click a row to filter

Group by

Group	Batches	Downgrades	Downgrade % ▼	Tonnes	Value loss £