Sovereign data, secure and interoperable data thanks to Data Spaces.

Until recently, industries were characterized by building tangible (I mean physical) assets that were then sold: cars, airplanes, food, pharmaceuticals… although now any company that wants to be competitive in the manufacturing industry must incorporate and consider data as an inevitable complement. Data as elements of the final product and data as part of the intermediate processes, that is, as a differentiating tool to gain productivity.

The value of data in industry and business ecosystems

Those data by themselves possess an intrinsic value, a value for the company and now, as incremental (and greater) value, for the other members of its ecosystem: customers and suppliers. For customers as a guarantee of quality, the origin and traceability of the product, and for suppliers as a source of knowledge to improve and optimize their participation in our final product. And in reality, the physical products of the next decade will look more like services. And this enrichment of data will also affect many other sectors, for example, mobility and transportation. A physical package being transported or a passenger traveling carries a large amount of time-sensitive information associated with it that “travels” between organizations.

What are Data Spaces and why they are necessary

This is the purpose and need for Data Spaces, that is, to accelerate the exchange of data in order to improve and innovate in my product, because in truth I am going to need many other data that I do not have or generate, or perhaps it will be other third parties interested in obtaining them from me. And thus a relevant Data Economy is created, based on the exchange and the generation and capture of this additional value. For example, the value of the Data Economy in EU27 has gone from 301 billion euros in 2018 to 579 billion in 2024 and represents 4.4% of GDP with an expected CAGR 2025-2030 of up to 7.5% according to EU reports.

Definition of Data Spaces

And we will conclude with a definition of Data Spaces: “A data space is an ecosystem where, voluntarily, the data of its participants are pooled (public sector, large and small technology or business companies, individuals, research organizations, etc.). Thus, and under a context of sovereignty, trust and security, data can be shared, consumed and products or services can be designed from these data spaces.” (Data Office, SEDIA)

In this definition two key concepts are shown: because if we want to share data we need a) Trust and security between the parties that exchange the information and b) Interoperability, that is, a language, a common language that allows identifying and accelerating this process.

From traditional APIs to new data exchange models

Historically we come from 1-to-1 exchange interfaces, the “old-school” APIs between companies. They were valid when interacting between few parties. Made to measure, with very clear fields in the exchange and perhaps standardized for clear economic processes. Who doesn’t remember EDI? Now “times have changed” and those processes are much more complex. The exchange is not just about an invoice, there is very sophisticated powerful information to exchange, and collaboration models between several parties where agreements must be signed “on the fly.” Imagine: production serial number, serialization of medicines, sensitive patient information in a study, end-to-end traceability of a food. But also third parties that can offer us services using our data and comparing them with theirs: carbon footprint calculation, probability that an animal contracts a disease, defects in battery chemistry that we have sent to our manufacturer. That is, services based on AI models. Many companies seek to build their own model… but what if they could rely on third parties that already have one developed or at least provide key information in certain subprocesses?

The challenge of storage and data sovereignty

And from the above it follows that we have another additional challenge: Where is that learning data stored? Obviously we want it “well protected,” “controlled” and “limited in its use.” We want it “sovereign”… and not “autarkic.” I imagine the difference is understood. An isolated data point is never useless. But a pirated data point is terrible. And Data Spaces were born to solve this trust problem between organizations, even between departments, and to accelerate their use and exploitation.

That is why I say that Data Spaces enable sovereign, secure, and interoperable data.