The optical sorter stitched to the ERP

The problem

One wrong setting on the sorter sends export fruit down the wrong lane.

The optical sorter decides the commercial channel for each piece of fruit (export, domestic, industrial). When the destination changes, someone types the setting into the sorter panel by hand; one wrong setting can send export-grade fruit down the wrong channel, costing margin or creating a quality risk.

• The optical sorter decides the commercial channel of every avocado by size and color: export, domestic market or industrial for pulp and oil, and each channel has a very different price per kilo.
• The destination changes with the work order — different markets ask for different size counts — and someone types the new criteria into the sorter panel by hand, often reading them from a printout at the side of the machine.
• The ERP already knows the active order and its specification. The sorter already has an API. Nobody connected them.
• A wrong setting sends export-grade fruit to the domestic lane or lets undersized fruit into an export box: lost margin or a quality claim, and usually nobody finds out until the boxes are counted.

How it fits the IRIS system

Connect stitches ERP and sorter over the API — the order sets the criteria.

Connect stitches the ERP and the sorter together via API: when the ERP marks the active commercial destination, it automatically publishes the size/color criteria to the sorter. Destination change = one tap in the ERP, not a manual panel adjustment.

With a perishable fruit, the sorting decision is the price decision: once an avocado lands in the wrong lane, there is rarely time to send it back. Taking the keyboard out of that decision is the fastest payback in the matrix. It also leaves a record of which criteria every lot was graded under.

See the full IRIS architecture →

Before and after

The sorter set by hand versus the sorter fed by the order

Aspect	Today	With iLEAN Connect
Changing the destination	Manual panel adjustment	One tap in the ERP
Size and color criteria	Typed from a printed spec	Published from the active work order
Export fruit in the wrong lane	A permanent risk	Removed at the source
Market-specific size counts	Remembered by the operator	Carried by the order
What the sorter really applied	Assumed	Confirmed back to the ERP
Replacing the sorter or the ERP	—	Neither

Manual adjustments with channel-error risk → criteria automatically synced with the active work order.

Impact estimate

Impact estimate — to be validated with your numbers.

The block below is an estimate to be validated against your plant's actual data. We put it forward so the committee has an order of magnitude; we refine it during the assessment.

• Fast payback, estimated at 3-6 months, the shortest in this matrix.
• From the reduced risk of margin loss when export fruit is misrouted to a lower-value channel.
• And from claims avoided when undersized or off-color fruit never reaches an export box, which protects the next shipment to the same customer as well.
• Manual criteria adjustments with channel-error risk become criteria synced automatically with the active order, with no extra work for the line.

Fast payback, 3-6 months, from reduced risk of margin loss due to misrouted export fruit. *Estimate to be validated.*

And the fair question from the production manager

“What if the ERP publishes the wrong criteria?” — then the error sits in the order, where it is visible and correctable, not hidden in a panel setting. Mapping an order's specification to the sorter's parameters is an anchored task, where the best models drop below 1.5% error [1], and the sorter confirms back what it applied, so a mismatch shows up before the first box is closed. The person at the sorter still sees the active criteria on the panel, only without typing them.

^[1] OpenAI paper "Why Language Models Hallucinate", 2025 — on the reliability of AI in anchored tasks.