The new platform accelerates the development of artificial intelligence in Europe

Until recently, there was no widely used common platform on which tools, data and computing power could be easily exchanged, but that is all changing now.

European research and innovation in the field of artificial intelligence is now coming together on a common digital platform called “AI-on-Demand (AIoD)”, an online hub funded by the EU and designed to help; researchers, start-ups and other companies and public bodies to collaborate and experiment more easily.

The unique center of European artificial intelligence

Tanvir Singh Badwal from the University of Cork in Ireland has been helping to develop AIoD since 2022.

“In short, AIoD is a place where researchers and industry can access resources, use services, and even develop new ones,” he said.

The platform did not appear overnight. Its development began in 2019, as part of the EU-funded AI4EU project, which was coordinated by Thales in France and which laid the foundations for a joint European center for artificial intelligence. It was then significantly expanded as part of the AI4EUROPE project, involving a consortium of 24 institutions from 15 countries led by Badwal and the University of Cork from 2022 to 2025.

Today, the next phase is initiated by the cooperation within the DeployAI project, which is financed from the Digital Europe program and whose goal is to introduce the platform to the market and expand its use to industry and the public sector.

It is clear that there is a need for such a platform. While the US and China dominate the development of artificial intelligence with their tech giants and large platforms, the European artificial intelligence environment is more fragmented. It aligns the active smaller players each with their own different national rules, funding programs and data standards. AIoD was created to make it easier to connect all of these with each other.

It is not a replacement for existing initiatives such as Gaia-X or the European Open Science Cloud, but their superstructure as it serves as a practical entry point for artificial intelligence tools and collaboration.

The platform makes it easy to find datasets, algorithms, computing resources and partners, and its goal is to accelerate innovation and help European artificial intelligence solutions move more quickly from research laboratories to real use.

All of this fits into the EU’s broader strategy for artificial intelligence, which seeks to strengthen research and industry in Europe while maintaining the alignment of artificial intelligence with democratic values, fundamental rights and the rule of law.

Photo: 123RF – ILLUSTRATION

From platform to practice

How does it all work in reality? Imagine a small agricultural technology start-up company that develops a tool to help; farmers in optimizing irrigation and fertilizer use. His team could use AIoD to mine satellite and ground datasets and prototype a predictive model using the platform’s AI Builder component.

This online tool enables users to create workflows with artificial intelligence using a visual interface, without the need for painstaking programming.

Users could then go to the Research and Innovation AI Lab (RAIL), an online environment where experiments can be conducted directly on the platform. In it, users can access integrated high-performance computing resources to test and refine their models.

When the tool is ready, start-up companies can create a virtual community on the platform to connect with mentors, find collaborators in other countries and put their solution into use in all parts of Europe. The result of all this would be that farmers can get a weather forecast several days in advance and thus reduce the use of water and fertilizers.

“Everything in one place” in the field of artificial intelligence

AIoD combines several functions in one place: it is part search engine, part trading platform, and part Internet laboratory.

Users can browse datasets, scientific publications, educational materials, software components and pre-trained models. They also have access to advanced computing power thanks to connections with European supercomputing centers such as the Center for Supercomputing in Barcelona and the LUMI system in Finland.

It is important to note that not all of these resources are on the platform. Instead, it brings together content from other established AI and open source platforms, including Hugging Face, Bonseyes, and OpenML.

“Everything that’s uploaded to those other platforms gets pulled into ours,” explained Badwal. “The idea is that researchers can search in one place and access many more resources.”

This is important because high-quality data and large computing resources are often concentrated in well-funded corporations or elite institutions. By facilitating access to these resources, AIoD seeks to ensure equal conditions for smaller laboratories and start-up companies.

Continued use of the tool even after the end of the project

Another goal is to ensure that the tools developed in EU research projects do not disappear after the end of funding.

One example is the Compass for Responsible Robotics, or RoboCompass, developed under an earlier EU-funded project and now available through AIoD. This self-assessment tool is dedicated to the non-technical aspects of responsible robotics. It is intended to help researchers assess the extent to which their robots are aligned with public expectations and doubts.

“You evaluate your robot in relation to various socio-economic, environmental, ethical and legal aspects,” said Joana Martinheira, a consultant at the Portuguese communications agency LOBA, which was involved in the technical development of the Internet platform.

“Everything is similar to a quiz. At the end, you get a grade and recommendations for improvement.”

By adopting such tools, AIoD extends their useful life, and research results are opened to a wider audience.

Creating a UI ecosystem

For those who want to network, the platform also includes mentoring sections, discussion forums and directories of projects and facilities in the field of artificial intelligence in all parts of Europe.

“AIoD simplifies cooperation and synergy between all actors in the ecosystem,” said Joana Martinheira.

By turning a multitude of separate tools and isolated projects into a common, searchable and usable resource, AIoD seeks to deepen collaboration and accelerate the development of artificial intelligence in Europe.

Considering that the DeployAI project is now oriented towards the expansion of the platform and its better connection with industry and public services, the ambition is clear: to provide Europe not only with intensive research in the field of artificial intelligence, but also to provide it with a functional artificial intelligence ecosystem that can turn ideas into practical solutions for companies, governments and everyday life.

This article was originally published in the EU research and innovation journal Horizon.

The research in this article was financed from the EU’s Obzor program.

European researchers in the field of artificial intelligence are joining forces using an EU-funded digital platform to exchange tools, data and computing power to foster collaboration and innovation.

By Editor