Periodic Reporting for period 2 - AI4Copernicus (Reinforcing the AI4EU Platform by Advancing Earth Observation Intelligence, Innovation and Adoption)
Período documentado: 2022-07-01 hasta 2023-12-31
AI4Copernicus reinforces the European AI on-demand Platform (AIoD) within the Copernicus ecosystem by facilitating the development of useful, on-demand Earth-observation artificial intelligence solutions to a number of technological, economic/business, and societal challenges and domains.
The overarching objective of the project is to make the European AI-on-demand platform, the platform of choice for users of Copernicus data along the value chain (scientists, SMEs, non-tech sector).
Via technological integration and a series of open calls, the project enables innovative use-cases and small-scale experiments, with the aim to increase the uptake of both AIoD and Copernicus data and services, while at the same time push the capacity for AI-driven innovation across multiple business and social domains.
AI4Copernicus achieves this by exposing AIoD resources on DIAS platforms, making it easy to procure computing power and large EO data, as well as to access training material and expertise.
The overarching objective of the project is to make the European AI-on-demand platform, the platform of choice for users of Copernicus data along the value chain (scientists, SMEs, non-tech sector).
Via technological integration and a series of open calls, the project enables innovative use-cases and small-scale experiments, with the aim to increase the uptake of both AIoD and Copernicus data and services, while at the same time push the capacity for AI-driven innovation across multiple business and social domains.
AI4Copernicus achieves this by exposing AIoD resources on DIAS platforms, making it easy to procure computing power and large EO data, as well as to access training material and expertise.
In order to provide a bridge between the European AI and EO communities, AI4Copernicus delivered a technological framework comprising a series of AI services and tools to bootstrap the development of AI and EO applications.
We have developed EarthQA, a question answering engine for discovering Copernicus data stored on CREODIAS with their metadata stored in a Knowledge Graph, expressed via the new Copernicus Ontology developed in the project.
A service enabling the deployment of AI pipelines created by AI4Experiments to CREODIAS platform has been developed, providing a tight integration between AIoD and the DIAS platforms.
A number of bootstrapping services and tools have been customized and made available, which empower users of the AI4Copernicus framework to innovate by making use of AI in their own domains.
A series of 4 open calls had been planned and carried out, leading to 8 small-scale experiments (smaller, single-beneficiary experimental projects targeting technology-advanced users) and 9 use-cases (larger-budget projects, involving at least one non-technology user).
An additional 5th open call was carried out, leading to 10 micro-projects (smaller, single-beneficiary projects for testing the AI4Copernicus services).
All results from the open call projects were published in the AIoD Catalogue.
Apart from providing financial support to the open call projects, we offered them mentoring services for technical, business, and ethical AI services.
In particular, all open call projects were introduced to a novel AI Ethics assessment, advocating for a Responsible “Ethics By Design” approach.
The open calls necessitate the utilisation of DIAS platforms, Copernicus data, the AIoD platform and the services and resources provided by AI4Copernicus.
Through organising, facilitating and mentoring these Open Calls, AI4Copernicus reaches out to new user domains and boosts the use of the AIoD platform.
The open call process allowed us acquire feedback for the AI4Copernicus services, and collect requirements and propositions for future extensions and optimisations.
Several communication and dissemination activities, including the AI Ecosystem Forum event in Athens, allowed us to grow an ecosystem comprising the SMEs participating the open calls, the European Association of Remote Sensing Companies (EARSC), as well as the ESA network.
Exploitation and sustainability are a priority and were being constantly monitored in line with progress made in other ICT-49 projects as well as with the AIoD platform.
The services, tools, and lessons learnt from the AI4Copernicus project are going to be exploited in future actions, and particularly the DeployAI project, where the outcomes of AI4Copernicus will serve as the building blocks of the offering of new AI-powered in the EO and environmental domain.
We have developed EarthQA, a question answering engine for discovering Copernicus data stored on CREODIAS with their metadata stored in a Knowledge Graph, expressed via the new Copernicus Ontology developed in the project.
A service enabling the deployment of AI pipelines created by AI4Experiments to CREODIAS platform has been developed, providing a tight integration between AIoD and the DIAS platforms.
A number of bootstrapping services and tools have been customized and made available, which empower users of the AI4Copernicus framework to innovate by making use of AI in their own domains.
A series of 4 open calls had been planned and carried out, leading to 8 small-scale experiments (smaller, single-beneficiary experimental projects targeting technology-advanced users) and 9 use-cases (larger-budget projects, involving at least one non-technology user).
An additional 5th open call was carried out, leading to 10 micro-projects (smaller, single-beneficiary projects for testing the AI4Copernicus services).
All results from the open call projects were published in the AIoD Catalogue.
Apart from providing financial support to the open call projects, we offered them mentoring services for technical, business, and ethical AI services.
In particular, all open call projects were introduced to a novel AI Ethics assessment, advocating for a Responsible “Ethics By Design” approach.
The open calls necessitate the utilisation of DIAS platforms, Copernicus data, the AIoD platform and the services and resources provided by AI4Copernicus.
Through organising, facilitating and mentoring these Open Calls, AI4Copernicus reaches out to new user domains and boosts the use of the AIoD platform.
The open call process allowed us acquire feedback for the AI4Copernicus services, and collect requirements and propositions for future extensions and optimisations.
Several communication and dissemination activities, including the AI Ecosystem Forum event in Athens, allowed us to grow an ecosystem comprising the SMEs participating the open calls, the European Association of Remote Sensing Companies (EARSC), as well as the ESA network.
Exploitation and sustainability are a priority and were being constantly monitored in line with progress made in other ICT-49 projects as well as with the AIoD platform.
The services, tools, and lessons learnt from the AI4Copernicus project are going to be exploited in future actions, and particularly the DeployAI project, where the outcomes of AI4Copernicus will serve as the building blocks of the offering of new AI-powered in the EO and environmental domain.
The AI4Copernicus results that offered progress beyond the state-of-the-art can be summarised as follows:
• A technical framework for bridging the European AI and EO communities by offering interoperability between the AIoD platform and the DIAS platforms.
• A question-answering engine for discovering EO datasets stored in CREODIAS, with metadata stored as a Knowledge Graph
* A deployment service which enables automated execution of AI4Experiments on the CREODIAS platform, providing interoperability between AIoD and DIAS.
* An AI ethics and trustworthy AI assessment tool for conducting self-assessments of AI-based solutions in the EO domain.
* Several audited services and tools of high impact, as a result of the work done by the awarded projects, in domains such as Security, Agriculture, Health, Maritime, and Energy.
Overall, the AI4Copernicus project delivered its main outcomes and fulfilled its expected impacts.
• A technical framework for bridging the European AI and EO communities by offering interoperability between the AIoD platform and the DIAS platforms.
• A question-answering engine for discovering EO datasets stored in CREODIAS, with metadata stored as a Knowledge Graph
* A deployment service which enables automated execution of AI4Experiments on the CREODIAS platform, providing interoperability between AIoD and DIAS.
* An AI ethics and trustworthy AI assessment tool for conducting self-assessments of AI-based solutions in the EO domain.
* Several audited services and tools of high impact, as a result of the work done by the awarded projects, in domains such as Security, Agriculture, Health, Maritime, and Energy.
Overall, the AI4Copernicus project delivered its main outcomes and fulfilled its expected impacts.