Periodic Reporting for period 2 - AI4EU (A European AI On Demand Platform and Ecosystem)
Reporting period: 2020-07-01 to 2021-12-31
AI4EU project overarching goals are to build an AI On-Demand Platform and mobilise the European AI Community to take ownership of the platform and lower the barriers to AI adoption, particularly for SMEs and business companies. Indeed, Europe's main weakness lies in its ecosystem's fragmentation, a lack of European tools and sovereignty in a very competitive and fast-moving worldwide landscape. Notably, the USA and China are allegedly leading AI Deployment. As a result, Europe seems to lag due to its fragmentation and seemingly more minor ability to transform its research assets into breakthrough innovations and big tech companies.
By offering a core node to access excellent AI personalised resources and tools, the AI4EU On-Demand Platform will strengthen the European community and bridge the gap between research and innovation. AI4EU proposal is built with the following strategic vision. First, we create an AI On-Demand Platform with On-Demand services allowing the user to find resources, assets, talents and test and experiment models, tools, and algorithms seamlessly to pave the way for a future AI Factory floor for the European AI Ecosystem. Second, the project will bring into the AI4EU-On-Demand platform excellent AI assets and resources (model, algorithms dataset sample etc.) and promote the AI European trustworthy model as a strong differentiator compared to other platforms. Third, we showcase the ability of the platform and the community to bring an ecosystem and use the AI-On-Demand Platform to solve AI challenges through industrial pilots on eight domains and a technology transfer program targeting SMEs and start-ups (41 solutions providers have tested the AI-On Demand Platform).
Finally, the project goal is to build a sustainable platform after the end of the project with a governance model supported by the community and the European Commission. A strategic research and innovation agenda is foreseen to work on the future (s) of the AI On-Demand Platform and contribute to this endeavour.
By offering a core node to access excellent AI personalised resources and tools, the AI4EU On-Demand Platform will strengthen the European community and bridge the gap between research and innovation. AI4EU proposal is built with the following strategic vision. First, we create an AI On-Demand Platform with On-Demand services allowing the user to find resources, assets, talents and test and experiment models, tools, and algorithms seamlessly to pave the way for a future AI Factory floor for the European AI Ecosystem. Second, the project will bring into the AI4EU-On-Demand platform excellent AI assets and resources (model, algorithms dataset sample etc.) and promote the AI European trustworthy model as a strong differentiator compared to other platforms. Third, we showcase the ability of the platform and the community to bring an ecosystem and use the AI-On-Demand Platform to solve AI challenges through industrial pilots on eight domains and a technology transfer program targeting SMEs and start-ups (41 solutions providers have tested the AI-On Demand Platform).
Finally, the project goal is to build a sustainable platform after the end of the project with a governance model supported by the community and the European Commission. A strategic research and innovation agenda is foreseen to work on the future (s) of the AI On-Demand Platform and contribute to this endeavour.
The project has released several versions of the platform over the three years of the project. During the first 18 months, the software development model was implemented by a core team using agile methods. The project delivered in less than 12 months a first version of the platform with most of the community features. A tipping point was to change the software development model and implement a distributed development model, involving more partners and Wps in the platform's development. The connections with the ICT 48 (network of excellence) and ICT 49 (communities driven project) has been strengthened, particularly to anticipate sustainability issues after the end of the project. The AI On-Demand Platform is operational (https://www.ai4europe.eu/) and comprises two environments and four main components.
First, the community environment was developed under the Drupal Content Management System. Here, the user can find A.I. community news, events, a catalogue of components (89 assets are available), a demonstration of the pilots, business domains and other features. The hosting of the new CMS system is performed by Netcup, an infrastructure provider based in Germany following the European GDPR. Second, the AI4EU-Experiments provide an environment to design, create and deploy composite A.I. applications built of modular A.I. tools. The software is based on the Acumos framework, an open-source initiative of the Linux Foundation. Two others components have been developed. First, a Single Sign-On to have central user registration and login capabilities. The AI4EU On-Demand platform uses the EU-Login ECAS provided and operated by the European Commission. Second, a Search Component contains several SoA tools for text indexing and retrieval (including NLP capabilities and a chatbot). It acts as an enhanced aggregator of different information sources (textual content of the AI4EU platform, public content related to artificial intelligence provided, scientific index and other external A.I. content when available)
The AI4EU On-Demand Platform is interoperable with six other platforms: BEAT, MUNDI, WEKEO, OpenML, IDSA, Hub'Eau. Other significant results are the creation of an Ethics observatory and a value-based A.I. design methodology, deployment of 8 industrial pilots, and a technology transfer program mentoring 41 solutions providers. In addition, a technical governance Board has been installed. It should continue after the end of the project to ensure sustainability in the short term of the AI On-Demand Platform.
First, the community environment was developed under the Drupal Content Management System. Here, the user can find A.I. community news, events, a catalogue of components (89 assets are available), a demonstration of the pilots, business domains and other features. The hosting of the new CMS system is performed by Netcup, an infrastructure provider based in Germany following the European GDPR. Second, the AI4EU-Experiments provide an environment to design, create and deploy composite A.I. applications built of modular A.I. tools. The software is based on the Acumos framework, an open-source initiative of the Linux Foundation. Two others components have been developed. First, a Single Sign-On to have central user registration and login capabilities. The AI4EU On-Demand platform uses the EU-Login ECAS provided and operated by the European Commission. Second, a Search Component contains several SoA tools for text indexing and retrieval (including NLP capabilities and a chatbot). It acts as an enhanced aggregator of different information sources (textual content of the AI4EU platform, public content related to artificial intelligence provided, scientific index and other external A.I. content when available)
The AI4EU On-Demand Platform is interoperable with six other platforms: BEAT, MUNDI, WEKEO, OpenML, IDSA, Hub'Eau. Other significant results are the creation of an Ethics observatory and a value-based A.I. design methodology, deployment of 8 industrial pilots, and a technology transfer program mentoring 41 solutions providers. In addition, a technical governance Board has been installed. It should continue after the end of the project to ensure sustainability in the short term of the AI On-Demand Platform.
AI4europe On-Demand Platform is existing and provides a first core node for the community in Europe that should be further strengthened by the contributions of the next Europe projects, chiefly the ICT 49. Significant progress has also been achieved for the community, more structured and more interested to use the On-Demand Platform. The creation of the TGB and efforts put to work with these projects have created good conditions to further expand the On-Demand Platform in the next years. We believe the platform is a common digital good and should be owned by the Community and not by vested interests. A critical stake is to make this vision alive.