2.3 Data sharing approaches
This framework targets three different approaches to data sharing:
- traditional data sharing (i.e., transferring data, including the rights to use the data, from one actor to another)
- on-site or remote data sharing (i.e., granting a user access within a controlled environment, or an external user access to a remote desktop or server, with access to data and tools within an infrastructure governed and controlled by the data provider)
- federated data sharing (i.e., granting an external data user access to a small component within a trusted federated ecosystem, governed by the data provider, with control mechanisms for what data can be accessed and what is allowed to be extracted)
It should be noted that C-ITS peer-to-peer data exchange (C2C or C2X) is per se a federated data sharing approach (edge to edge), already deployed on the market (VW Golf and ID.x using ETSI ITS G5 / IEE802.p/bd). However, the usage of federated data sharing in this document concern data sharing between centralized data centre nodes.
Traditional data sharing (data download)
This approach includes transferring data to another stakeholder using any media or via network access. The data is being copied to the data users’ environment and then used for any purpose agreed within the data provider in an agreement. This approach has been the most common, however, the volume and methods sharing the data have changed over the years. The methods include sharing a physical media (e.g., a hard drive), using or using secure network protocols (https, ftps, ssh).
On-site or remote desktop
The benefits of remote desktop or on-site access is that the raw data must not necessarily be exchanged since the user connects to an environment in full control of the data provider. This means that any operation on the data is done within the environment and the data provider can decide on what data can be accessed. The environment is often restricted for data extractions (by different means depending on the sensitivity of the data), and if an extraction is accepted, the traditional data sharing approach is used.
The benefit is that the user can get access to relevant data, but the data provider is in full control. By allowing remote data access, the user need not be at the same office (or even country). The data user must, however, accept the conditions and data protection measures stipulated by the data provider.
Federated data sharing
Federated data sharing represents a method where various parties can share data without centralizing it. It is often associated with the concept of ‘data spaces’, which refers to collaborative environments for sharing data across different sources under governed conditions. Two commonly used terms are:
Federated Data Space (FDS), which refers to an integrated approach to managing and accessing data that is distributed across multiple systems or locations. In a federated data space, different data sources maintain their autonomy but are interconnected in a way that allows for unified data access and analysis.
Federated Database System (FDBS), which is a type of database management system that allows for the management and integration of multiple autonomous databases into a single federated database. Each participating database in a FDBS remains independent but can be accessed and queried as part of the federated system.
The concept of federated data sharing is an approach that could be seen as a compromise between the two previous methods of traditional data download and on-site or remote desktop access. However, it introduces a fundamental shift in the direction of data flows. The principle is that trust is established between actors in a network, and common principles for data access, description, and formats facilitate data exchange. This approach includes infrastructure and software tools, authentication and authorization, cybersecurity and data protection, taxonomy, data catalogues, and governance. These aspects have been described in European Data Strategy and GAIA-X technical specifications (https://docs.gaia-x.eu/technical-committee/architecture-document/23.10/) and implemented in various software platforms, such as international Data Space (https://docs.internationaldataspaces.org/ids-knowledgebase/v/ids-ram-4/), X-Road (https://docs.x-road.global/Architecture/arc-g_x-road_arhitecture.html), and Eclipse (https://eclipse-edc.github.io/docs/#/README). Compared to data download or sharing access to a common database, the principle of FDS is granting access to specific data based on an agreement that governs which data the user is allowed to access. This means that original data does not necessarily need to be sent to a user; instead, only the product of a computation is shared.