AI USE CASES YOU CAN REUSE.English edition · Individuals, independents, small businesses
Spotted on the web

Targeting water leakage on the most likely sections

Use described publicly by Numérique360 — Banque des Territoires (28 February 2025), summarised by the desk. Not tested by iadujour.

The result to aim for

Sicoval, an agglomeration of 36 communes south-east of Toulouse, used Leakmited's AI to target its research: 12 leaks found in two weeks, 113 000 m3 of avoided losses and approximately 93,000 euros saved.

What was done

According to Digital360 (28 February 2025), Sicoval manages 1,000 km of drinking water with a 78% yield, or about one million cubic metres lost each year, a level deemed unacceptable in a context of stress on the resource. The community already had about 50 connected sectorization flow meters. Leakmited has applied an algorithm that operates a base of more than one million leaks, as well as the geological context, weather and topography, to designate the 20 to 30% of the network that concentrates 80% of the leaks. In two weeks of targeted research, twelve leaks were found and repaired immediately, avoiding 113 000 m3 of losses, equivalent to the annual consumption of 950 households, and about 93,000 euros of expenditure. The benefit, billed for the result according to the volume of water saved, cost 20,000 euros. The work on the algorithm was conducted by the claimant, with the operator Network 31 for the interventions.

Before you start

Tool cited by the source: Leakmited. Fill in the bracketed items: [tableau]. Strip confidential data that is not needed and check your tool’s access terms. Features and plans can change.

The method, in three steps.

  1. Gather the context

    Spot the bracketed items in the prompt ([tableau]) and prepare what they ask for.

  2. Adapt the prompt

    Copy the prompt below, replace the bracketed items, then paste it into the AI assistant you use. The cited tool (Leakmited) is not necessarily available to the public..

  3. 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.

The prompt to customise
Here are the data from our water network sectors (sector, length, age and material of pipes, night flow, past leaks): [tableau].

1. Classifies areas by leak priority, explaining the criteria.
2. Estimated the order of magnitude of the recoverable volume if the 20% priority is treated.
3. Proposes a two-week research plan and indicators to measure its outcome.

What deserves your attention

These results cover a two-week campaign; They don't say how many leaks the tool would have left on the rest of the network.
Origin of the case. Use described publicly by Numérique360 — Banque des Territoires (28 February 2025).
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.
TO GO FURTHER

Other ideas to put to work.

The whole Public services & local government theme
Public services & local government

Respond to residents 24 hours a day on the website and citizen portal

Asnières-sur-Seine first limited a classic chatbot to three services, then migrated to the general Wikit AI: he answers on the site and the citizen portal, civil status come at the top of the questions, including off-time opening.

Tool: Wikit SemanticsType: Spotted
Public services & local government

Propose a conversational assistant on the new site of a municipality

In Saint-Genis-Laval, Jacky Béjean integrated a Wikit citizen chatbot into the site's redesign: the city compared Mistral and OpenAI, retained OpenAI for the quality of the answers, with data hosted in France.

Tool: Wikit Semantics (OpenAI)Type: Spotted
DISCUSSION

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.