Recipe-driven batching
An ingredient maps to one or more motors, not one-to-one: one site runs three ingredients across five motors, another three across three. Same software, no branch.
Batch Run Report
| Ticket | Time | Target | Actual | Deviation | Status |
|---|---|---|---|---|---|
| B-260909-014 | 10:21 | 833.00 | 833.50 | +0.50 kg | In spec |
| B-260909-013 | 10:04 | 833.00 | 833.00 | 0.00 kg | In spec |
| B-260909-012 | 09:47 | 833.00 | 833.00 | 0.00 kg | In spec |
| B-260909-011 | 09:30 | 833.00 | 832.50 | −0.50 kg | In spec |
| B-260909-010 | 09:13 | 833.00 | 833.00 | 0.00 kg | In spec |
Recipe selected
The operator picks the recipe on the indicator. That one choice sets a target weight for each ingredient and decides which motors feed each of them.
Calcium
First ingredient. One motor, one auger, until the hopper reads its target. A recipe with no calcium skips this step outright — the motor never starts and the cycle moves on.
Soy
Second ingredient, on top of the first. The hopper is never emptied between ingredients, so each dose is measured as a difference — which is why an error on one ingredient does not cancel another.
Maize — three motors
One ingredient, three stores, three motors running together. This is the whole reason the recipe is modelled against ingredients and not against motors: the same build runs this five-motor site and the three-motor one, with configuration deciding which motors serve which ingredient.
Batch complete
The hopper discharges and the batch becomes a record: target against actual for every ingredient, scored against the scale’s own division of 0.5 kg. A tolerance finer than the division would mean nothing.
The run, as a record
The shift as the supervisor sees it: every batch against its target, in the scale’s own half-kilo steps. The last row is the batch the drawings built, and the ingredient table under it names the motors that served each one — the recipe itself never does.
