Data transfer and data management made easy! The Gaia-X 4 KI project shows how it's done.
The Gaia-X 4 KI project has been conducting research for two years under the consortium leadership of the German Aerospace Centre (DLR) on the development of a data and services ecosystem for training and validating artificial intelligence (AI) applications. The 13 project participants are researching a total of nine use cases in the automotive sector in which they are developing applications for the Gaia-X ecosystem. The project has now reached its first milestones in practical testing. As consortium leader Sascha Knake-Langhorst from the DLR Institute of Transportation Systems states: "We are moving full steam ahead towards the home stretch of the project and are showing what we have achieved. If we can finalise our work according to the current plans, we will have helped to take Gaia-X from science into practice."
In the fields of "Production" and "Automated and connected driving", the researchers have already successfully trialled Gaia-X in two of the defined use cases:
Securely transmit data in production
Gaia-X 4 KI has focussed on data transfer in the production of lidar sensors for the "Automated Optical Inspection" use case. The production systems there are largely automated. This also includes automated optical inspections (AOI), which provide information on whether the sensor surfaces produced are free of defects. The evaluation is carried out by an AI model on the production computer. In order to dynamically adapt to any quality deviations in the supplied components and to detect any unknown defects, it is essential to retrain the algorithm. This retraining usually takes place off-site at a partner company that specialises in this task. This means that safety-critical data must be transferred to an external company. It is also necessary to transfer the updated AI model back to the manufacturing company. This is where the project shows how Gaia-X can provide support as a shared data architecture. A digital contract for the service is negotiated between the companies via Gaia-X and a trustworthy data room is used.
Ensuring the accuracy of infrastructure data
The digitalisation of the transport system continues to progress. For example, fleet operators now have to analyse and process a large amount of data from their fleet. This also includes data from infrastructure facilities such as traffic lights, which can be used for numerous automated driving applications. As the accuracy of this data is of crucial importance for road traffic and individual road users, the monitoring of data sources is a fundamental issue. Gaia-X 4 KI offers an approach to this with the "SIEM" (Security Event Monitoring) use case. The aim of the researchers is to implement continuous monitoring using Gaia-X's streaming functionalities. In this context, they use Gaia-X to transmit data securely. It also allows infrastructure operators to flexibly subscribe to corresponding services in a data room.
The examples in the areas of "Production" and "Automated and connected driving" show just a small selection of the Gaia-X application scenarios that the project participants are developing and testing. The researchers have prepared an overview of all the use cases investigated in the project as videos on the project website: https://www.gaia-x4ki.eu/en/Demo. In the two minimal examples developed in the project, "Carla" and "AI training pipeline", the project consortium also shows the technical foundations they have developed for the use of Gaia-X in the various areas of application: Report on the first milestone.
Gaia-X 4 KI at the Hannover Messe 2024
At the end of the project, Gaia-X 4 KI will present its results at the Hannover Messe in April (22-26 April). The researchers will be presenting their work in person at the Gaia-X 4 Future Mobility project family stand. Further information will be published on the project website in due course.
About the Gaia-X 4 KI project
The Gaia-X 4 KI project was launched in June 2021, making it the first of the Gaia-X 4 Future Mobility project family in the mobility domain. 13 companies and research institutions are working together on the project. It will run until mid-2024, has a budget of 18 million euros and is funded by the Federal Ministry for Economic Affairs and Climate Action (BMWK).
