Ink color control with a spectrophotometer — ΔE measured and validated batch by batch, not by eye against an aging master.

The problem

The light booth approves, the spectrophotometer measures — and the data stays on an island.

Approving color by visual comparison worked for decades, but it has a flaw at its root: the reference you compare against is not stable, and the outcome is not a number. In a typical ink plant:

1. The physical master ages — the reference drawdown loses strength and shifts hue with light and time. Every batch is approved against a target that has moved.
2. The eye depends on everything — on the observer, the shift, the condition of the booth lamps, and on metamerism: two inks that match in the booth can split apart under the supermarket shelf lighting.
3. The spectrophotometer is disconnected — measurements happen, but the reading lives in the instrument or in an island software. What lands on the batch sheet is a signature, not a ΔE.
4. Every customer has its own tolerance — the brand demands ΔE00 ≤ 1 on its corporate color and accepts more on the secondary one. That gets managed from memory, not in a system.

The quality manager knows it; the color matcher knows it. The classic system works 99% of the time. That 1% is the batch the customer returns because the printed packaging does not match the previous run — and then you re-shade, replace, and try to explain why there is no numeric record of the batch that shipped.

How it fits the IRIS system

iLEAN does not add a new spectrophotometer — it ties the one you already own to the batch.

Color control does not fail for lack of instruments; it fails because the measurement is never tied to the batch or to the customer's tolerance. iLEAN acts as the putty that fills the gap between the spectrophotometer, the production order sheet and the ERP, without forcing you to change anything.

Edge identifies the batch. Connect captures the spectrophotometer reading. The agent checks the ΔE against the tolerance for that reference and that customer. The person signs — never the other way round.

The iLEAN pieces applied to per-batch color control:

• Edge — a terminal in the QC lab or at the line. It reads the QR on the production order and identifies the batch and the reference. With a voice earpiece, the color matcher says “batch 4812 measured” and the system ties it in with no typing. It works without a network. If the plant loses its connection, it keeps capturing.
• Connect — integrates with the spectrophotometer you already own (X-Rite, Konica Minolta, Datacolor, Techkon) over USB, Bluetooth or the file its software exports, and captures L*a*b*, ΔE, illuminant and observer. If the instrument only shows the value on its screen, the color matcher photographs it and Connect reads it.
• Color agent — cross-references ΔE + reference + customer tolerance + batch. If the batch does not pass, it warns the color matcher and holds the release until a new measurement passes or the quality manager signs off the exception. The corrections applied are kept as a pattern. The person signs — the batch never ships on its own.

See the full IRIS architecture →

Before and after

Light booth + physical master vs. ΔE measured and traced with iLEAN

Aspect	Booth + physical master	With iLEAN (spectrophotometer tied to the batch)
Color reference	Aging master drawdown	Digital L*a*b* master values per reference
Approval criterion	Color matcher's eye, per booth and shift	ΔE computed against the customer's agreed tolerance
Record per batch	Signature on the order sheet	Full reading tied to the batch number
Out-of-tolerance batch	Caught only if the eye catches it	Release held at second zero
Customer complaint	History reconstructed by hand	Per-batch color dossier ready to send
Veteran's judgment	In his head	Historic corrections recorded as a pattern

Impact estimate

Impact estimate for your plant — to be validated with your numbers.

The block below is an estimate to be validated with the specific data of your plant. We put it forward so the committee has an order of magnitude; we refine it during the diagnostic.

• Printing ink plant (flexo, gravure, offset) or color paste and concentrate maker, with a spectrophotometer already in use in the QC lab.
• Pilot on one line's QC lab (spectrophotometer integration + batch identification by QR). First value expected within a few weeks.
• Indicative payback between 4 and 9 months, depending on batches/month and the average cost of a color return (re-shading, freight, urgent replacement, printer penalty).
• Hard levers: ≥ 30% reduction in batches returned or reworked for color; a per-batch color dossier ready to hand over; consistency across repeats of the same reference.

And the quality manager's reasonable doubt

“What if the system misreads a measurement?” — false positives exist, and that is why the system never decides alone. Hallucination is a problem of free generation, not of anchored tasks. In tasks where the AI merely receives the digital value from the spectrophotometer and compares it against an agreed tolerance, the best models brought error below 1.5% ^[1]. And even so, what is critical is never decided alone: iLEAN warns and the color matcher or the quality manager sign. The three safety rings exist precisely for this.

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