Digitalising shop-floor production tracking
-
Project name
Food industry · contract packing
-
Industry
Food industry · packing / co-packing
- Development time
Quick overview
The client is a contract food-packing company (co-packing): it receives goods, processes and packs them per client packaging orders, organised into runs.
Its challenge: make weight control and traceability reliable, and move off paper without disrupting the teams in place.
For a contract food-packing company, Quai des Balises designed a tablet production-tracking app: weight-control calculations (mean, standard deviation, T2 limits) become automatic with real-time alerts, a dynamic run summary aggregates CCP, traceability and yield, and the tool connects to the existing WMS — to digitalise the shop floor without disrupting habits.
Context
Tracking relied on a stack of paper documents (tracking sheet, metal-detector logs, net-weight logs, anomaly sheets, primary packaging, picking list, exit control, labels). Weight calculations (mean, standard deviation) were done by hand by line leaders — hard and often skipped. Data was re-keyed twice (spreadsheet + WMS), with no automatic summary (CCP, traceability, yield).
The need / the problem
The need: a production-tracking app, usable on the shop floor, covering:
Automatic weight calculations (mean, std dev, T2 limits) with real-time alerts on non-compliance
A dynamic run summary: CCP status, traceability, weight control, yield calculation
Pre-fill of known information (packaging refs, quantities) from the packaging order
Tablet photo of labels (run start / end) instead of paper gluing
Online client request management, with AI pre-fill from a plain e-mail
Connection to the existing WMS (API / CSV) for a single source of truth
Specific challenges and constraints
Our approach
Field immersion
Understand runs and documents directly on the shop floor before designing.
As-is first
Faithfully reproduce the process, then evolve — not both at once.
Interactive mock-ups
Project the concrete very early to validate before final pricing.
AI as assistance
Pre-fill and visual recognition to reduce entry, never validation.