Insights from an Automotive Use Case for Integrating Services in a Gaia-X Data Ecosystem

Kurzfassung

In many modern business use cases, achieving a goal requires collaboration among multiple companies and organizations, particularly when data-oriented assets and services are involved. These collaborations occur because individual companies either lack essential resources or prefer to concentrate on their core competencies due to different areas of expertise and the separation of concerns. For instance, for the homologation of an automated vehicle, various stakeholders, including original equipment manufacturers (OEMs), regulatory bodies, and testing services, must work together to ensure that all safety standards are met. As part of this process, extensive testing, both in the field and through simulation, is conducted to validate vehicle behavior under different conditions. Various type of data, such as high-definition maps and driving scenario description files, is required, along with data-oriented services like test field, simulation, and monitoring services, to conduct and analyze a driving test. The Gaia-X initiative aims to facilitate the collaborations among companies by enabling a federated data ecosystem. However, our observations and practical experience within the automotive sector of a Gaia-X ecosystem, reveal that this collaborative process is inefficient and tedious. This is because, although Gaia-X federated services support compatibility, secure sharing, and proper data governance, additional challenges must still be addressed by Gaia-X ecosystem participants. Many of these challenges originate from a traditional market mindset, including individually negotiated contracts, isolated metadata queries, and fragmented service integration, all of which slow down processes and delay innovation. This work aims to address how data and services can be integrated within a Gaia-X ecosystem, and how various metadata serve as the interface between services. Among the various use cases in Gaia-X, we focus on a specific automotive application that is sensor model validation. In developing this use case, we faced significant challenges, particularly in data exchange and aligning stakeholder expectations. Misunderstandings around data and service requirements led to problems. For example, inconsistencies arising at the interfaces between services due to differing data formats and deviations from agreed performance parameters delayed the processes. Furthermore, integration was complicated by the later realization that additional data inputs that initially overlooked were in fact very important. This work presents the solutions developed as part of a collaborative effort within the sensor model validation use case. We demonstrate how standardized approaches to specify data and services can overcome the challenges. We show how data and services can be effectively offered and accessed within Gaia-X through the specification of detailed metadata, policies, and contracts. In particular, we emphasize the importance of standardized descriptions. For services, this includes defining the expected input and output formats, the parameters involved, and the operational characteristics. For data, this involves specifying the data format, its semantic meaning, indicators of quality, and the policies that govern access. Our findings provide concrete insights into how Gaia-X can enhance the visibility, usability, and seamless integration of services and data within a federated ecosystem, and improve the processes within a collaborative data-driven ecosystem.