Non-Functional Requirements

Defining the quality attributes for how a data spaces should work

DEFINING QUALITY ATTRIBUTES

Non-functional requirements are in many ways even more relevant than functional requirements, as in most practical cases, they cannot be jointly optimised. The table below mostly contains non-functional requirements that are either directly mentioned or that can be readily inferred, from EU policy documentation.

More detailed analysis of policy priorities relating to common European data spaces, combined with stakeholder interviews, may lead to the identification of additional significant non-functional requirements. In fact, some non-functional requirements can be directly derived from the perspective (pertaining to the pure technical dimension) of data spaces as platforms, on top of which end-users can build solutions, whether these are data provision services or ready-made AI products.

This implies a need to consider certain aspects of usability, notably creating an environment that effectively supports developer productivity or habitability, e.g., by providing comprehensive tools such as software development kits (SDKs) and ensuring that user interfaces are clearly defined, meaningful and accessible from the developer perspective. Other implicit non-functional requirements extend beyond technical considerations and include governance-related aspects, e.g., the need for long-term sustainability of data spaces, as these digital infrastructures are expected to outlive the initial investment phase leading up to their development and operationalisation. 

With the aforementioned taken into consideration, the list provided below is not meant to be exhaustive, yet start you on you data space journey.


NON-FUNCTIONAL REQUIREMENTS 

Security and confidentiality. Security by design principles are required in order increase trust and eliminate barriers to sharing private data, including personal data as well as intellectual property. 

Interoperability. On the non-functional side, common vocabularies and standards are expected to provide a basis for interoperability of the different data and services built by ecosystem participants.  

Maintainability. Data spaces functionality is expected to be developed gradually and be in constant evolution as new technical developments emerge and the priorities of the different participants and domains shift.  

Variability. This quality supports the development of different variants of core building blocks in order to support the specific priorities of different sectoral data spaces, allowing for their tailoring and evolution in accordance to the characteristics of their respective users and data. 

Scalability. Data spaces shall have the capacity to adapt and continue to function in the presence of potentially massive changes in terms of data available and exchanged, as well as in terms of numbers of users and services deployed. 

Performance. Data Spaces shall support high-performance data analytics use cases, including those based on High-Performance Computing (HPC). Similarly, high-speed data transfer between infrastructures and participants are expected to be required.  

Auditability. Data spaces are expected to be transparent in their operation, enabling users, governing bodies and regulators to monitor activities taking place within them and to enforce rules.  

Portability. A key quality attribute for the seamless transfer of services and data assets between different infrastructures, including across deployment tiers (field devices, edge, multi-cloud environments, including private and public components), effectively preventing vendor lock-in. 

Energy efficiency. Data spaces should favour the use of digital infrastructures with an optimised energy demand, and provide orchestration mechanisms involving low-power devices when feasible, effectively reducing carbon footprint, notably when hosting intensive data processing workloads.  

Inclusivity. Data spaces should allow for open and broad stakeholder participation in decision-making processes concerning access, control, sharing and use of data.

Fairness. Data spaces should maximise the value of data in the economy by ensuring that a wider range of stakeholders gain control over their data and that there is a greater and fairer flow of data in all sectors. They should thus enable fairness in the allocation of value from data among actors in the data economy and foster access to and use of data.

Sustainability. Data sharing infrastructures should have (i) a viable business model, (ii) appropriate governance model, and (iii) open participation by the relevant actors that would ensure they can be sustained in the long term.

Trustworthiness: Trust is a crucial factor in any data sharing arrangement. Participants and stakeholders in data spaces should feel confident that the data is governed responsibly, used with integrity and in accordance with the law, and that the technology being used is also trustworthy and secure. The idea of retaining control over one's own data and its use is an important basis for establishing trust amongst data space participants and is also in line with EU rules and values regarding data sovereignty. Trust can also be built through transparency, accessibility and participation.

Transparency: A certain degree of transparency is needed for the proper functioning of a data space. This should include information about what is happening within the data space, including what data is provided or received, for what purpose and for what duration. The policies, rules, standards and participation modalities in a data space should be clearly and transparently defined - individuals should know how to contribute data, who has access to it, and how they can interact with it.

Data sovereignty: Data sovereignty involves enhancing control by organisations and individuals over data that they contribute to generating. It implies participation in data governance and allows individuals and organisations to self-determine how, when and at what price others may use their data across the value chain. It means that data holders can safeguard user data, and ensure that it is used only in accordance with strictly defined rules. 



Suggested Section: Data sharing & reuse

Learn how to leverage on existing data space examples 

Disclaimer: The views expressed are purely those of the authors and may not in any circumstances be regarded as stating an official position of the European Commission. 

  • No labels