Child dermo pH 5.5 and hypoallergenic — an excipient that slips into a bath or diaper product is not an incident, it is a damaged brand.

The problem

Child dermo does not allow "more or less" — and the classic control falls short at the two key moments.

A child dermo product (bath gel, tear-free shampoo, body lotion, diaper-care wipe) carries two hard requirements that the classic system only half-watches:

1. pH 5.5 ± an agreed range in the final product — and, what almost nobody controls, across the whole mixing curve, not just the closing value. A mix that spends 20 minutes in an alkaline stretch can degrade sensitive molecules even if it ends at the correct pH.
2. A hypoallergenic formulation — a closed list of permitted excipients and preservatives. What is banned (long-chain parabens, fragrance allergens from Annex III of 1223/2009, harsh sulfates, certain isothiazolinones) cannot sneak in through a silent supplier lot change.

The quality manager knows this. But the classic control takes a single pH measurement at batch closing and looks at the supplier's certificate when there is time. The classic system works 99% of the time. That 1% is what comes back as a cosmetovigilance ticket — irritations reported by pediatricians, returns at the pharmacy, a precautionary recall that stains the entire child dermo brand. In pediatric dermo, trust is built over years and broken in one.

How it fits the IRIS system

iLEAN Edge does not replace your pH meter — it captures the whole curve and cross-checks it against the formula and the actual lot.

The child dermo control problem is not a lack of sensors: it is that pH gets checked at closing while the supplier's certificates live in another folder. iLEAN acts as the putty that seals the crack between the tank, R&D, purchasing and packing, without asking you to change anything on the floor.

Edge captures the tank's pH curve. Connect captures the lot's actual INCI. The agent cross-checks both against the current formula and the hypoallergenic list. If something drifts, the batch is held. The person signs — never the other way around.

The iLEAN pieces applied to pH control and hypoallergenic verification in child dermo:

• Edge — a terminal in the mixing room with an input for the tank's existing pH meter. It captures the complete curve throughout the whole mix, not just the closing value. A CNN trained on curves from batches that reached the customer in good condition: it learns which curve shape produces stable product. When the curve drifts (a prolonged alkaline stretch, an anomalous oscillation, an out-of-range closing value), it holds the batch. It works without a network.
• Connect — captures the supplier's lot certificate whether it comes from the EDI portal, a PDF email or a WhatsApp from the sales rep. It extracts the actual INCI and cross-checks it against the child dermo product's current formula and the brand's hypoallergenic list. If a non-permitted excipient or an unexpected INCI appears, it raises the alert before the lot enters the mix.
• Tracer agent — assembles the per-batch dossier: formula, actual INCI of every raw material, pH curve, in-line controls and human signature. When a cosmetovigilance ticket arrives, the team opens the batch and the root cause is visible in minutes.

See the full IRIS architecture →

Before and after

Manual child dermo control vs. cross-checked control with iLEAN Edge

Aspect	Closing check + certificates in email	With iLEAN Edge + Connect
pH measurement	Single value at batch closing	Complete curve throughout the whole mix
Silent supplier lot change	Certificate in a PDF, depends on who opens it	Captured and cross-checked against the hypoallergenic list
Banned excipient detection	When someone reviews the certificate by hand	Before the lot enters the mix
Traceability for cosmetovigilance	Manual reconstruction, weeks	Per-batch dossier, opened in minutes
Degrading pH curve	Invisible if the closing value adds up	Learned from the curves of good batches
Operation without a network	n/a	Edge keeps running on the cabinet's own power

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.

• A child dermo line with two or three families (bath gel, lotion, diaper care), surfactant and preservative suppliers with frequent lot changes, occasional cosmetovigilance tickets today.
• Edge pilot on one mixing tank (input to the existing pH meter + Connect for lot INCIs + a per-batch dossier agent). First value expected within a few weeks: the pH curve captured and cross-checked against outcomes for the first time.
• Indicative payback between 4 and 9 months, dominated by the elimination of the cosmetovigilance recall risk and the reduction of response time when a ticket does occur.
• The hard lever is a single avoided cosmetovigilance recall: in child dermo a mass recall is not measured in product alone, it is measured in years of brand. That pays for the entire pilot.

And the quality manager's reasonable doubt

"What if the AI misreads an INCI or a pH curve?" — hallucination is a problem of free generation, not of anchored tasks. In tasks where the AI merely recontextualizes structured data (reading a certificate, comparing against the hypoallergenic list, reading a curve against a pattern), the best models brought error below 1.5% ^[1]. And even so, what is critical is never decided alone: Edge holds the batch and the person signs. The three safety rings are there for this.

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