Paperless shift start on the filling line

The problem

The ERP has the order. What actually happened takes a shift to arrive.

The problem is not that the paper exists, it is that it takes a whole shift to reach the system. It stays in a drawer until someone collects it, and those are exactly the first documents an auditor asks for, or the ones you hunt for when a complaint lands on a batch made three weeks ago. Meanwhile the ERP has the order, but not what actually happened.

• Between the three documents they define the whole batch: what is filled, on which line, in which container and with which raw material.
• They sit in a folder in the office until somebody collects them — a full shift of latency.
• They are precisely the first documents an auditor asks for, and the ones you hunt for when a complaint about a batch from three weeks ago arrives.
• The syrup room sheet is the most critical and the one that travels worst: it is where the concentrate and the sugar that went into the finished syrup are written down.

How it fits the IRIS system

Connect in photo mode — the operator keeps filling in the same paperwork.

Three phone photos, zero change of habit. With Connect, the operator takes three photos at start-up. iLEAN recognises the structure of each document and extracts the fields: batch, line, format, container, target Brix, concentrate lot, sugar lot, start time and any handwritten notes. The operator still writes by hand and still signs on paper. Nothing about the routine changes. The batch exists in the system at second zero. From that photo onwards, the batch lives in central memory from the first minute of the shift rather than at the end of the day. Any later question starts from a record that built itself: what was filled, with what, who was there and what was noted by hand. Manual reconstruction of batch records disappears.

Three photos at start-up and zero habit change. On a filling line that is not about convenience: it is the only way the record still exists in month six, because anything that asks for typing between bottles gets filled in at the end, from memory.

See the full IRIS architecture →

Before and after

Start-up on paper versus start-up captured

Aspect	Today, on paper	With iLEAN Connect
Latency to the system	A full shift	Second zero
Which concentrate and sugar went in	On the syrup sheet, in the office	Tied to the batch from start-up
Answering a three-week-old complaint	Hunt for paper	One query by batch
What the auditor sees	A folder	The batch file with the photo of the original
What the ERP holds	The order, not what happened	Both
Operator's routine	—	Unchanged: three photos

Impact estimate

Estimated impact — to validate with your own 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.

• Estimated payback 4-9 months.
• Transcription time currently spent by the production supervisor.
• The hunt for paper disappears at every complaint and every audit.
• From a shift of latency to second zero in the batch file.

The return comes from two places: the transcription time the production supervisor spends today, and the end of paper hunting every time a complaint or an audit arrives. Estimated payback between four and nine months depending on batches per day and active lines. Estimate to be validated against your own figures.

And the fair question from the production manager

«What if it misreads the concentrate lot?» — reading fields from a document with a known structure is an anchored task, where the best models drop below 1.5% error [1]. And there is a second net: the summary goes to the tablet and a person confirms or corrects it before anything crosses into the system. Nothing enters on its own.

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