Edge Computing pioneers

Since AzurİA was founded in 2021, we've been committed to pushing the boundaries of innovation by creating smart, eco-responsible solutions.
Our expertise in artificial intelligence and bold vision have positioned us as pioneers in near-sensor processing, with a first successful research project deploying Aİ on a nanosat led by our founders in 2019.

Edge Computing

Image processing is deployed close to the sensors, extracting useful information from the source and transmitting only useful data.

Eco-designed

We achieve this with low power consumption and in a small volume capable of dissipating the energy consumed.

Frugal

With very few labeled images, our Aİ can adapt to a new use case or detection class.

Multispectral

Addition of spectral bands useful for the specific application, significantly increases performance compared to RGB/visible.

Systemic approach

Evaluated not only on Aİ metrics but also on higher-level system constraints such as inference time and overall consumption.

Agnostic

Our Aİ adapts to the number of input spectral bands, being able to infer images from different types of sensors in both the visible and infrared spectrum.

Eco-responsible and ethical by design

 

Dataset selection

If we test external datasets, we make sure to check their origin and associated licenses.

Aİ training

The semi-automatic annotation of our images is verified by our in-house data scientists, and training is then carried out on the GPUs integrated into our local server, with continuous measurement of power consumption to keep our carbon footprint under control.

Solutions utilisation

Images are inferred in real time close to the sensors on low-power boards, and only the alert containing the thumbnail image of the detection of interest is sent in real time to the end user, respecting RGPD by design and reducing carbon impact by transmitting only useful information and not a video stream with a need for energy-intensive storage and processing.


CEO and Cofounder of AzurIA

Frédéric FERESIN

  • University degree (DEUG S) before entering an engineering school (Polytech Orléans).

  • 30 years of engineering and project management in major space industry groups, from phase A to phase E.

  • Set-up in 2018 then management of the IRT project “Chaîne Image Autonome et Réactive” (CIAR) with Aİ embedded on NanoSat and terrestrial drone. 

CTO and Cofounder of AzurIA

Lionel Daniel

  • 4 years of research in probabilistic logic and biometrics.

  • 6 years of software engineering in the field of space cameras, optical payload test benches and in-flight calibration software.

  • 3 years of architecture in Aİ on the CIAR project.

2020

  • Integration of the PACA Est Incubator

2021

  • i-Lab national pre-selection validates the concept

  • AzurIA created on the basis of strong values enshrined in its articles of association and the rules of the Social and Solidarity Economy

  • French Tech Emergence grant to develop our in-house Aİ multi-spectral frugal solution

2022

  • Confiance.ai winner (French Grand Défi project) for Aİ deployed on low-energy targets

  • Winner of the European SecurIT call for projects to develop a tethered aerostat embedding a real-time detection and alert system

  • "AI Talent” selection in the European BonsApps project

2023

  • RAPID signature for Aİ integration on observation satellite constellation

  • Selected for a live demonstration during the i-Naval event for detecting events at sea

  • Crédit Agricole Innovation trophee

  • Hi France label

  • Real-life demonstration of Helia for forest fire detection

  • AMO phase 1 architecture protection of the Sologne forest (SDIS 18/41/45)

2024

  • AMO phase 2 deployment protection of the Sologne forest (SDIS 18/41/45)

  • SME Trophy for the most initiatives in favor of ecological and CSR transition

Expertise recognized by

BPI France, DGA,  Confiance.ai, SecurIT / Horizon2020, Bonsapps,  several SDIS.

Learn more

Discover Aİ
Discover CamIA
Discover Helia