Blog

As part of its Innovation Contract with the Occitanie regional economic development agency (AD’OCC), Kalisio signe une convention avec le SMMAR. Indeed, one of the objectives of this ambitious R&D program is to evaluate the possibility of creating predictive water level models based on artificial intelligence, in order to better prevent the risk of flooding. Thanks to this agreement, Kalisio will be able to exploit the hydrometric measurements of SMMAR, which owns a number of measuring stations within its perimeter, in addition to the stations already accessible via the portal Hub’Eau.

By synchronizing forecasts with real-time visualization of the extent of a zone below a certain level (look video) we could know precisely, and several hours in advance, the potentially flooded areas of a region.

As anOpen Source andOpen Data player in the field of geographic data, Kalisio offers cartographic backgrounds based on OpenStreetMap data.Nevertheless, OpenStreetMap remains a non-profit organization with limited resources. It is therefore not possible to use the openstreetmap.org tile servers : at will policy on the use of tiles):

OpenStreetMap data is free for everyone to use. Our tile servers are not.

Until now, we used a third-party supplier to provide us with optimized data (i.e. tiles). They were ready to use and hosted on our servers, or those of our customers. Today, we're delighted to be able to generate your own tiles from the OpenStreetMap database in full autonomy. And it's all thanks to the fabulous ecosystem help to contribute to:

• kaabah for automated deployment of computing infrastructures,
• krawler for the orchestration of computational tasks,
• openmaptiles for importing the OpenStreetMap database and generating tiles,
• tippecanoe for tile aggregation,
• kargo for making tiles available via various standard OGC protocols.

A project that might seem trivial in 2021, but it's not. No less than 256 processors and several days of calculations required to produce this 90-gigabyte data set ! This represents :

• 508 million buildings
• 602 million transport-related entities (roads, railroads, etc.),
• 80 million points of interest (public facilities, tourism, stores, etc.),
• 113 million administrative entities (countries, states, regions, cities).

Nothing revolutionary or that doesn't already exist (e.g. Google Maps), but this is also what digital sovereigntyis all about: know-how that enables independence and resilience.

View the result in Kano.

Read a press article about our Akt’n’Map solution in a feature dedicated to fire risk, and take a look at our videos:

• The first illustrates the use of EFFIS services from the COPERNICUS program to obtain fire risk forecasts for the coming days,
• The seconde illustrates the coupling with weather forecasts from Météo France (ARPEGE model) to better anticipate future dangerous situations.