The "Eclipse AIDGE"
open source platform 

 The DeepGreen collective is engaged in developing the first dedicated independent
and sovereign embedded AI software platform.
 
 
The Eclipse AIDGE platform facilitates the design of powerful, efficient deep neural networks tailored to the needs of the constrained environment in which they are intended to operate. Its goal is to provide a set of tools and functions that reduce model computational complexity and memory consumption, select the best hardware target and generate optimised implementation in the relevant programming model. 

Committed to AI that is...

Open

We are building a digital commons whose software building blocks can be reused and adapted by all users. The embedded AI revolution demands an approach that offers guarantees around transparency, trust and efficiency and delivers clear benefits for industry and society.

The platform is hosted by the Eclipse Foundation, a global community recognised for its commitment to open source projects and developers, and provides an environment that encourages and facilitates collaboration, knowledge sharing and innovation.

Responsible

We are driving innovation to design and implement a frugal and reliable model of AI.

By developing methods that reduce model size and/or the amount of data needed for learning, we are promoting lean and efficient approaches at every stage of the intelligent system development and operation cycle.

The platform is also designed to respond positively to the quality, robustness and reliability issues specific to mission-critical systems to ensure compatibility with certification criteria.

Enter ECLIPSE AIDGE

High-potential applications and technology demonstrators

As part of the DeepGreen project, the Eclipse AIDGE platform will be used to develop 7 demonstrators in key areas for embedded AI.

Energy

Health

Defence

Aerospace

Aviation

Factory of the Future

Key project stages:

SPRING 2023

Launch
of DeepGreen

Incubation
& Maturity

Autumn 2023

Creation
of the Eclipse AIDGE
platform

Spring 2024

Eclipse AIDGE platform v1

Deployment
& Optimisation

Spring 2025

First demonstrators
and integrations of
sovereign hardware solutions

Innovation
& Europe-wide Dissemination

2026-2027

Ongoing innovation
focused on maximising frugality
and trust in embedded AI

Discover our latest news