AI-MATTERS
AI in manufacturing: Test and experimentation facilities for European SMEs
Your contact person
Muhammad Saeed
Research coordinator
ARENA2036
AI-MATTERS
Networking, testing and optimizing AI in production
AI-MATTERS puts artificial intelligence (AI) at the heart of European manufacturing innovation and strives to integrate, optimize and make transformative AI technologies accessible, especially to Europe's small and medium-sized enterprises (SMEs), through extensive testing and experimentation.
Background
The EU is investing €139 million in actions dedicated to artificial intelligence under theDigital Europeprogram, including sectoral testing and experimentation facilities. EU Member States are investing, preparing and mobilizing their national research and technology base and resources to remain competitive when it comes to the global transformation that human-centric and trustworthy AI will enable in different areas of society.
Project objective
The main objective of AI-MATTERS, with a funding volume of €60 million, is to implement concrete steps to strengthen Europe's positioning and leadership in the introduction of human-centered AI systems. These will enable robotics and other advanced technologies to transform the manufacturing sector in the EU. However, this overarching goal must be achieved without compromising on environmental protection, human safety and quality of life.
The project, launched in January 2023 and expected to end in December 2027, aims to establish a European network of interconnected large reference test and experimentation facilities for AI technologies and concepts that have the potential to increase the competitiveness of the European manufacturing industry.
Beyond the potential of the technologies and applications, this requires careful consideration of aspects such as scope, project size, critical mass, specialized equipment and services, and secure virtual access. For certain services, a definition of use cases within the TEF is extracted from the service portfolio.
The following services are presented at ARENA2036:
- Wireless real-time data transmission
- AI evaluation in the automated robot process
- Consulting for robot IT architecture
- Consulting for Asset Administration Shell
- Provision of store floor data for logistics applications
AI-MATTERS is coordinated by CEA-LIST in Paris-Saclay and brings together a consortium of 25 major research, technology and innovation organizations around seven nodes and one satellite in eight European countries. The AI-MATTERS players are already regional and national leaders in the use of artificial intelligence in the manufacturing industry. Together, they bring in-depth expertise to achieve the objectives of the tender.
Project structure
The overall concept of the AI-MATTERS work is divided into 5 work packages, which focus on the following topics: Management (WP1), Setup (WP2), Operation (WP3), Impact (WP4) and Dissemination (WP5). Although the arrangement of WP2-5 is theoretically sequential, all WPs run throughout the project.
AP1
Deals with the cross-cutting issue of work management of the consortia. It monitors the progress of the project, supervises the technical work and manages the financial contribution from EC and national sources.
AP2
Includes the physical and organizational construction of the seven nodes and the satellite. The main focus here is on preparing the physical facilities and setting up the various test benches based on the partner's existing infrastructure. A key result is the definition of the service catalog and the common operating procedures across all AI-MATTERS nodes so that all customers have access to the same high standard of service.
AP3
Manages the day-to-day operation of the network after it is released. While AP3 conceptually follows AP2, the two overlap heavily to enable iterative refinement of the physical and logical services offered (AP2) and continuous improvement of the service offering process (AP3).
AP4
Bundles all activities to achieve an impact. This includes raising awareness of the benefits of AI in general and AI-MATTERS services in particular in the relevant sectors. It connects the network with relevant stakeholders and policy makers to promote an AI-friendly political, economic and social environment. WP4 covers the entire project duration.
AP5
To pool dissemination activities for the general public and specific interest groups. It will work closely with all other WPs and keep the interested European public informed about the progress of AI-MATTERS. While WP5 is mostly unidirectional, WP4 is bidirectional as it guides the actions of the AI-MATTERS network based on the current state of the AI landscape in Europe and supports European decision makers (from local to regional to European level).
Planned services
Service 1: Application for error detection and container picking
This service integrates AI-driven error detection with the picking of bins by robots. In manufacturing, bin picking refers to the process by which robots remove parts or items from a bin. By incorporating defect detection, AI systems can ensure that only defect-free components are selected, improving overall quality control. By using machine vision systems and machine learning algorithms, AI can automatically identify, analyze and sort faulty parts, increasing the efficiency of production lines for SMEs. This service will be presented at the European Robotics Forum (ERF) 2025 in Stuttgart.
Service 2: Digital twin of a robot system with 5G
A digital twin is a virtual image of a physical system - in this case a robotic system - that is used to simulate and optimize its performance. In combination with 5G technology, this service enables the monitoring and control of robotic systems in real time with minimal latency. With the digital twin, manufacturers can predict system behavior, perform tests and make adjustments before making changes to the actual system, significantly reducing downtime and increasing productivity.
Service 3: Shop floor optimization with 5G
This service uses 5G connectivity to optimize the processes in a workshop, i.e. the physical production area in a manufacturing plant. The enhanced communication capabilities of 5G enable faster data exchange between machines, sensors and control systems. This leads to better coordination of production tasks, faster responses to problems and the ability to implement AI-driven decision-making processes to optimize workflows, reduce waste and improve production efficiency.
Service 4: AI in the generation of synthetic production data
In many cases, there is not enough real-world data to train AI models for complex manufacturing tasks. This service generates synthetic data - artificially created data sets that mimic real-world conditions - to train and test AI models in manufacturing scenarios. By creating realistic data, AI systems can learn to deal with different production challenges without the need for extensive and costly data collection from real machines, making the development of AI-driven manufacturing solutions faster and more flexible. This service will be presented at the European Robotic Forum (ERF) 2025 in Stuttgart.
Project partners
Sponsor
Our colleagues from the research coordination team have compiled more information on the EU AI Act. The white paper can be downloaded here:
Whitepaper on the AI Act
(PDF, English)