The result to aim for
Sepalumic, an aluminium manufacturer, has re-launched a project that has been blocked for six years thanks to the AI: identify in a few seconds some 3,500 similar references the profile that enters into lacquering.
What was done
In an article published by CCI Nice Côte d'Azur on 7 October 2026, Jean-Philippe Picon, DSI of Sepalumic, explains that the shareholders of the group requested an AI cell in each unit. The case chosen: automatically recognize the aluminium profiles on the lacquering line, among about 3,500 nearby references, to launch the right program and ensure traceability. The slowing of the building freed up time; a pre-diagnosis of the ICC and the work of two young experts in mathematics resulted in a prototype presented to the Governing Board, which gave its green light to invest. A first entry into service is envisaged before the end of the year. The company estimates the return to about 80 times the bet, an unverified internal estimate in production. The precision requirement is close to 100%, as one day missed in 30 tons would cost between 300,000 and 500,000 euros.
Before you start
Tool cited by the source: Custom computer vision. Gather the information the prompt asks for. Strip confidential data that is not needed and check your tool’s access terms. Features and plans can change.
The method, in three steps.
Gather the context
Gather the information described in the prompt.
Adapt the prompt
Copy the prompt below, replace the bracketed items, then paste it into the AI assistant you use. The cited tool (Custom computer vision) is not necessarily available to the public..
Reread before using
Reread the result, correct it if needed and keep the final decision. Caveats are detailed below.
Your starting point.
Replace the bracketed items, then adapt the proposal to your situation. This prompt is written by the iadujour desk It is not a quotation from the source.
I want to evaluate an automatic shares recognition project on a production line. Data: 1. List the questions to be decided before a prototype. 2. Proposed a test protocol with a confidence threeshold and a reference to an operator below this threeshold. 3. Build a conservative return on investment calculation, separate assumption and measured facts.
What deserves your attention
Proof level: Testimony published by an identified source. Source checked on 10 November 2026. Use described by the source, not tested by the desk. Original summary written by iadujour.fr: figures and claims are the source’s; the prompt is a desk proposal, not a quotation.
Comments.
A field note, a question, a precision on this use case? Comments are reviewed before publication.
Published comments
Loading comments…
Leave a comment
Moderation charter. Every comment is reviewed before publication. We refuse hate speech, insulting or discriminatory language, advertising, spam and promotional links, off-topic posts and other people’s personal data.