Deploying the common European manufacturing data space: The UNDERPIN and SM4RTENANCE frameworks’ perspective

Abstract

The growing interest in using data in the manufacturing sector has led to an increased need for efficient and secure data sharing among stakeholders. However, technical, business, and regulatory barriers often prevent the full exploitation of manufacturing data. To address these challenges, the European Commission (EC) has introduced the concept of Common European Data Spaces (CEDS) as part of its broader data strategy. These data spaces aim to ensure data sovereignty and interoperability, in compliance with European regulations, facilitating a trusted environment for data exchange. This paper explores the deployment of two key European-funded projects, UNDERPIN and SM4RTENANCE, which are designed to accelerate the development of the CEDS for the manufacturing sector. These initiatives represent the current efforts towards enabling system integration and strengthening data security, while aiming to enhance cross sector collaboration within the industrial ecosystem. The development of UNDERPIN and SM4RTENANCE data spaces play a crucial role in overcoming technical and business barriers, enabling organizations to unlock the full potential of data-driven innovation and improve operational excellence.

1. Introduction

In recent years, a significant increase in data generation across various industry sectors, including manufacturing, has been observed. Data analysis provides valuable insights into operations, uncovering inefficiencies in processes and helping companies optimize resource allocation and reduce overall costs [1]. However, data often remain locked in individual companies’ or vendors’ systems and are not fully exploited [2]. Even though many companies have invested in specialized equipment for data production and collection, these systems often operate independently and are not designed to easily share data with external stakeholders. In addition, from a technical point of view, there is a lack of integration between companies’ systems and those of suppliers or third parties who may be interested in using the generated data [1]. Furthermore, several data assets -such as those in manufacturing- often contain sensitive information related to proprietary processes and designs, and companies are reluctant to share them [3]. To address these limitations and unlock the potential of industrial data, policy and regulatory frameworks have been introduced at the European level.

In 2020, the European Commission (EC) published the European Strategy for Data [4], which puts people first in developing technology and promotes European values and rights in the digital world. In addition, the European strategy for data aims at creating a single market for data that will ensure Europe’s global competitiveness and data sovereignty. To protect the rights of citizens while simultaneously fostering technological development, the Data Governance Act (DGA) [5] and the Data Act [6] have been introduced. The DGA offers a framework to enhance trust in data sharing across different sectors, ensuring compliance with European principles of privacy and sovereignty. It outlines rules and safeguards intended to allow for the reuse of protected data when permitted under specific European Union (EU) or national legislation. The DGA introduces a sharing model that emphasizes transparency among data intermediaries and seeks to ensure that individuals and companies retain control over their data. The Data Act initiative aims to address challenges related to data in the EU, emphasising fair data access, user rights and data protection. It defends rules for a fair, transparent and equitable access to industrial data with the main goal of ensuring that data generated from various sources (e.g. industrial machinery and IoT) are accessible and available for sharing. It also clarifies the rights and responsibilities, as well as the conditions under which public sector bodies and/or private businesses can request access to privately held data. Furthermore, the Regulation on High-Value Datasets [7], a component of the Open Data Directive [8], defines the free availability of specific categories of public-sector data, such as geospatial, environmental, and mobility data. This regulation aims to remove barriers to accessing crucial public datasets, opening the way for new information products and data-driven innovation. Building on these policies, the EC has launched targeted initiatives for the implementation of data sharing across strategic sectors. The creation of Common European Data Spaces (CEDS) [9] in fourteen strategic fields, as outlined in Fig. 1, is a key part of the European Strategy for Data, with manufacturing being one of these fields.

Fig. 1.

The Common European data spaces sectors.

Key objectives of CEDS, presented in Fig. 2, include preserving data sovereignty, creating a decentralized infrastructure for trustworthy data sharing and exchange, and promoting openness and collaboration Fig. 2.

Fig. 2.

Common European Data Spaces key objectives.

To support the growth of CEDS, the EU established the European Data Innovation Board (EDIB) [10], which provides guidelines and recommendations for their development. Furthermore, the establishment of data pools for big data analytics and machine learning applications aims to ensure compliance with applicable legislation such as the AI Act [11], which sets quality and governance requirements for training data, and the Interoperable Europe Act [12], which provides a legal framework for cross-border interoperability of data and services.

In addition, the EC has funded several initiatives to support the implementation of CEDS, the most notable being the Smart Open-source Middleware (SIMPL) [13], an open source, smart and secure middleware platform that supports data access and interoperability among European data spaces. SIMPL provides modular components and sector-specific deployment environments, enabling both public and private stakeholders to experiment with and implement interoperable data infrastructures. Another important initiative funded by the EU is the Data Space Support Centre (DSSC) [14], which defines common requirements and establishes best practices to accelerate the formation of sovereign data spaces as a crucial element of digital transformation in all sectors. The DSSC offers resources such as starter kits, blueprints, and glossaries to guide stakeholders through the setup and governance of data spaces. Both initiatives are funded under the DIGITAL Europe programme [15], which is considered as key driver for data space development, as it funds pilot projects that demonstrate the value proposition of data spaces and explore technical solutions for secure and interoperable data exchange. It supports various initiatives some of which aiming at digitalizing manufacturing processes, improving efficiency, and driving innovation in the manufacturing sector. It also aims to bridge the gap between digital technology research and market deployment, focusing on five key areas: supercomputing, artificial intelligence, cybersecurity, advanced digital skills, and the widespread use of digital technologies across the economy and society.

In the manufacturing domain, the research and development activities have included data spaces as one of their key topics. The EC funded project BOOST 4.0 [16] introduced the concept of European Industrial Data Space [17] to improve the competitiveness of Industry 4.0 and to guide the European manufacturing industry in the adoption of Big Data in factories through 10 large-scale lighthouse pilots. In another effort, the Manufacturing-X [18] initiative, funded by the German government, implements the DataSpace Industrie 4.0 principles [19]. This initiative, which aligns national strategies with the broader EU vision for interoperable data spaces, seeks to accelerate value creation in industrial manufacturing and production through support and cooperation among several projects, such as the Factory-X [20] and others. In alignment with the goals of the Data Space 4.0 [21], which was a preparatory action to develop multi-stakeholder data governance scheme, EC funded two Deployment Actions projects focused on the manufacturing sector. They are UNDERPIN [22] and SM4RTENANCE [23], both of which started in late 2023. UNDERPIN consists of a consortium of 11 partners from 5 EU countries and has a duration of 24 months, while SM4RTENANCE has a duration of 36 months and includes 44 partners from 11 EU countries. Both projects provide the foundation for implementing the CEDS for the manufacturing sector, and play an important role in overcoming technical and business barriers, allowing organizations to unlock the full potential of data-driven innovation. The development of these initiatives is supported by a broader ecosystem of organizations that provide the technical and governance frameworks necessary for secure and interoperable data sharing. Among the most notable is the International Data Space Association (IDSA) [24], which offers a reference architecture for secure and standardized data sharing and is considered a cornerstone of data spaces concept. In addition to IDSA, the Big Data Value Association (BDVA) [25], the FIWARE foundation [26] (both members of the DSSC consortium) and the Gaia-X [27] contribute to advancing interoperability, data governance and the development of reliable data infrastructures across Europe. These organisations have formed the Data Spaces Business Alliance (DSBA) [28] to speed up business transformation in the data economy.

This paper presents the two deployment action projects, UNDERPIN and SM4RTENANCE, that target to accelerate the development of the CEDS, specifically for the manufacturing sector. In addition to introducing the projects, the paper outlines a series of challenges related to manufacturing data spaces and explains how these challenges are addressed by the two projects. It also highlights the benefits of the actions undertaken by UNDERPIN and SM4RTENANCE, including enhanced data interoperability and improved operational efficiency. The remainder of this paper is organized as follows: In Section 2, the challenges related to manufacturing data spaces are briefly explained. Section 3 presents the two projects and outlines their roles in addressing the identified challenges. Section 4 presents a discussion on the progress made toward implementing data spaces for manufacturing through the UNDERPIN and SM4RTENANCE projects, along with the concluding remarks of this work.

2. The Challenges in Adopting Data Spaces for Manufacturing

Manufacturing-targeted data spaces provide a secure and interoperable environment where industrial companies, organizations and SMEs can share and utilize data collaboratively, fostering new opportunities for innovation, efficiency, and competitiveness. However, their widespread adoption faces significant challenges. These challenges can be broadly categorized into two areas: technical and business-related. On the technical side, issues related to interoperability, security, and the standardization of data formats pose substantial obstacles. On the business side, the challenges are more strategic, involving the establishment of effective governance structures, raising industry awareness, defining the tangible value of participating in data spaces and preserving sustainability.

2.1. Technical challenges

2.1.1. Interoperability

In the manufacturing industry, where a variety of legacy systems and modern digital tools co-exist, achieving interoperability is crucial. Manufacturing companies often use legacy systems, typical siloed and proprietary [29], that were not designed to interact with modern, interoperable platforms. Integrating these systems into a data space requires significant customization, which can be time-consuming and costly. Additionally, the plethora of platforms and software used by the different organizations creates a heterogeneous IT landscape in which development efforts are necessary to ensure the communication of the various systems. To ensure that diverse data sources can be integrated without compatibility issues, common ontologies and knowledge graphs should further be developed to define shared vocabularies and relationships. Moreover, semantic hubs and tools are crucial in aligning the interpretation of data across different systems and sectors, especially when data is exchanged across manufacturing supply chains or between production and maintenance systems.

2.1.1.1. Standardisation of data formats

The absence of universally accepted standards for data exchange within the manufacturing industry also complicates the interoperability. Different organizations may store and process data in various formats, making it difficult to integrate and analyse data across the entire data space. Therefore, organizations may need to convert data into a common format manually or through complex software solutions, increasing the risk of errors and inefficiencies. This complexity can be a limiting factor for companies to join data spaces, especially smaller manufacturers with limited IT resources.

2.1.1.2. Web-of knowledge-related technologies

The manufacturing industry already values the engineer profile experienced in numerical solutions of complex differential equations [[30], [31], [32]] for instance, interlinking thermodynamics, chemistry and their effects in the motion and deformation of engine components. In the context of data spaces, semantic interoperability requires a common language for data representation [29]. While ontologies and semantic web are not new concepts [33], they have jumped to prominence with the advent of the internet [34]. As an example, there has been work on the XML syntax since the late '90 s [35,36] and the Resource Description Framework (RDF) language was initially drafted in 1997, with the first recommended specification appearing in 1999/2000 [37,38]. Therefore, there is a need for either specialist’s incorporation or workforce retraining in these new trends, as personnel must become familiar with knowledge-based formalisms. This is particularly relevant in manufacturing, where domain-specific semantics- such as machine states, production events, and quality metrics- must be consistently interpreted across systems.

2.1.2. Security concerns

Industrial companies often handle proprietary and sensitive information such as production parameters and supplier information. The risk of unauthorized access or data leaks can make companies reluctant to share data, particularly if the consequences of such breaches are severe. Even with legal frameworks in place, trust must be built through robust cybersecurity measures tailored to industrial environments, which often includes systems and operational technologies with known vulnerabilities. In manufacturing data spaces, this is especially critical due to the interconnected nature of production networks and the potential impact of breaches on operational continuity.

2.2. Business challenges

2.2.1. Governance structures

When multiple organizations share data, disagreements can arise regarding data usage, ownership, or breaches. In the industrial sector and the manufacturing ecosystems the challenges involve complex relationships between OEMs, suppliers, and service providers. This is where governance structures come into play to address these concerns, especially when data flows across borders or between companies with differing compliance obligations. Clear rules and dispute resolution mechanisms are essential to ensure trust and sustained collaboration.

2.2.2. Industry awareness

The concept of data spaces is relatively new, and many organizations may be unfamiliar with its potential benefits. This lack of awareness often leads to hesitation in adopting the technology, despite the potential benefits such as improved supply chain visibility, predictive maintenance, and streamlined quality control. To overcome this barrier, it is essential to educate organizations about the benefits of data spaces and provide clear examples of how they can be applied to their specific needs. Organizations like IDSA, BDVA, and Baidata [39] regularly conduct events and other means of dissemination to raise awareness.

2.2.3. Defining tangible value

Quantifying the benefits of data spaces, such as improved decision-making and efficiency, can be challenging [29]. Additionally, to justify participation in a data space -especially when upfront investments in technology and training are required- clear metrics and cases studies must be defined. While data spaces promise long-term benefits like efficiency and innovation, these must be translated into measurable outcomes, such as reduced downtime, improved throughput, or enhanced product quality.

2.2.4. Sustainability

Sustainability encompasses both financial viability and the continuous technological relevance and adoption by the industry. Business models must accommodate the diverse capabilities of participants, from large OEMs to small suppliers in order to incentivize participation. This is particularly important in the early stages of a data space when the critical mass of participants is not yet secured, and there is reluctance to pay a fee for joining. Additionally, sustainability depends on continuous technological evolution and alignment with emerging industry standards to remain secure and useful. In manufacturing, this also includes adapting to evolving production technologies and regulatory requirement.

3. Deploying Manufacturing Data Spaces

Building on the foundations laid by BOOST 4.0 and Data Space 4.0, and with the goal of accelerating the wider adoption of the data spaces, two Deployment Actions projects- UNDERPIN and SM4RTENANCE- have been funded under the DIGITAL programme [15]. Both projects are contributing to the advancement of data-driven solutions within the manufacturing industry and to shaping the EC’s vision regarding the technological, governance, and business frameworks required for their implementation.

These projects are set to leave a foundation after their completion, in the form of a variety of sector-specific small-scale data spaces with room to further grow and incorporate additional participants. Having concrete implementations in several sectors (energy, textiles, automotives, machine tools and others) they pave the way and establish best practices, templates and standards for others who intend to follow in the same footsteps. UNDERPIN emphasizes on common semantics and ontologies to support data space integration. It works towards creating a shared understanding of data across different systems, organizations, and domains so that data can be exchanged, interpreted, and used meaningfully. SM4RTENANCE is working on the implementation of a federation, which will enable participants from a given data space to browse the data offers of a different one and to conduct the corresponding data contract negotiation and subsequent exchange, without having to be an onboarded member of the target data space that holds the data of interest.

3.1. Existing manufacturing data spaces

The Catena-X Automotive Network [40], a secure and standardized data exchange platform for the automotive industry is a network aligned with IDS and Gaia-X standards. It promotes sovereign data exchange, allowing users to maintain control over their data. The primary goal of Catena-X is to create a unified data ecosystem that collects and analyzes data throughout a vehicle's lifecycle. This comprehensive data collection can lead to valuable insights in both the utilization phase and the manufacturing process. As a collaborative data space, Catena-X aims to develop new data-driven business models and improve efficiency within the automotive sector.

The Dutch Smart Connected Supplier Network (SCSN) [41], is a data exchange solution supported by the Netherlands Organisation for Applied Scientific Research (TNO). Companies register once with an SCSN service provider and can subsequently exchange data with all other affiliated companies along the production chain. Examples of the shared data include orders, invoices and technical product information. As with any data space, a company keeps a grip on its own data, and determines which data is shared to what extent with whom. SCSN bolsters productivity along the supply chain through fast, secure, and interoperable information sharing between participant companies.

There have been some discussions about reaching a degree of interoperability between the two large data spaces [42], a significant challenge, as both spaces operate under different business models and have their own governance structures. Cross-data space interoperability, if achieved, can facilitate data transactions and gives rise to new opportunities, as it would require less effort for companies in one data space to be able to negotiate for data held by a participant of a different one.

3.2. UNDERPIN: advancing data-driven manufacturing excellence

3.2.1. UNDERPIN use cases

The UNDERPIN project was initiated at the end of 2023 and has a 24-month duration. UNDERPIN focuses on dynamic asset management and predictive/prescriptive maintenance and establishes a secure, user-friendly platform for manufacturers and SMEs, ensuring compliance with legal considerations, promoting trust and security, enabling technical and semantic interoperability, and deploying sustainable business models. Fig. 3 summarizes these foundational domains, each supporting the deployment of secure and interoperable manufacturing data space.

Fig. 3.

Embracement of data spaces pillars on UNDERPIN.

UNDERPIN validates its approach through two real-world pilot demonstrations in oil refineries and wind farms. These pilots showcase how manufacturing companies and SMEs, can leverage shared data for improved decision-making, better asset utilization, and enhanced maintenance strategies. Both pilots focus on demonstrating the benefits of data-driven decision-making for SMEs involved in the supply chain of these industries. In refineries, the main business goal is to optimize operational efficiency and minimize downtime. Predictive maintenance models that forecast equipment failures and optimize maintenance schedules are developed through the project. These models identify potential anomalies and predict equipment behaviour using historical data and past events and enables refineries to proactively address maintenance issues, reduce downtime, and improve overall operational performance.

On the other hand, wind farms face unique challenges in terms of asset management and maintenance due to the remote and harsh environments in which they operate. Within the project data from various sensors installed on the wind turbines are collected and analysed. These data are used to develop predictive maintenance models that can monitor the health of individual components, such as gearboxes, bearings, and generators. Wind farm operators can schedule maintenance performed by external vendors more efficiently and enhance the overall reliability and profitability of their operations. To support these pilots and ensure their scalability, UNDERPIN addresses key adoption challenges, as outlined below.

3.2.2. Addressing adoption challenges

A key aspect of UNDERPIN’s contribution is its focus on semantic interoperability, which has already been demonstrated through technical workshops and publications [[43], [44], [45]]. The project has developed a semantic layer integrated into the IDS reference architecture, enabling a structured and standardized way to harmonize metadata and facilitate seamless communication between diverse IT infrastructures [46]. This integration leverages Ontology-Based Data Access (OBDA) technologies, which have been tested in the project’s use cases of predictive maintenance in refineries and wind farms [47]. In addition, UNDERPIN uses a semantic technology platform that supports the creation of knowledge graphs, taxonomies, and ontologies, facilitating semantic enrichment of industrial data [48]. These tools are designed to improve data discoverability and interoperability across manufacturing ecosystems and enable the integration of data from different sources. This approach aligns with the proposed layered architecture for federating vocabulary hubs across data spaces, which enhances semantic services and supports scalability, governance, and data sovereignty as outlined in the IDS-RAM framework [49].

Additionally, the project enhances data security by incorporating blockchain-based smart contracts, which provide cryptographic guarantee and transparent tracking of data transactions. These smart contracts are deployed on a private Ethereum network using a Proof-of-Authority consensus mechanism, ensuring non-repudiation of key events such as contract acceptance and data access [50] . While this component is still under development, it is part of the project's roadmap to ensure secure and verifiable data exchanges in the future.

On the business side, UNDERPIN has established a foundation for trust by exploring and addressing the legal requirements and regulations that govern data sharing, compliance, and governance within the data space [51] . The project highlights the use of standardized legal agreements, which are aligned with the General Data Protection Regulation (GDPR) [52], the DGA and DA requirements, ensuring that all participants are legally accountable for their actions and that data exchanges are transparent and compliant. These legal agreements play a key role in enabling trusted cross-organizational data exchanges. For example, a wind farm operator and a service provider can formalize a contract that specifies how operational data from turbines will be shared for predictive maintenance purposes. This contract would clearly define who owns the data, who can access it, and what actions can be performed on the data, ensuring that all parties are legally protected, and that the data is used in accordance with the owner's preferences.

In addition to the legal framework, the project aims to implement business models that reinforce the sustainability of the data space. Current models include both subscription-based and pay-per-use approaches, designed to incentivize participation and ensure economic viability for all stakeholders, while embedding compliance and transparency as core principles [53]. Looking forward, ongoing work explores scenario-based financial planning, new revenue streams, and strategies for scaling the ecosystem, with a continued focus on aligning business practices with evolving regulatory requirements and stakeholder needs.

Nevertheless, the successful deployment of manufacturing data spaces requires also strong value creation beyond the technical foundation. UNDERPIN demonstrates this through two industrial use cases. In a wind farm predictive maintenance models trained on multi-year SCADA data from a 10-turbine onshore wind farm have shown promising results [54]. These models successfully predicted generator bearing temperature anomalies, with several failures detected days in advance. The use of tree-based ensemble methods such as CatBoost and XGBoost yielded high recall rates, identifying all known failures during the evaluation period. Although precision remains a challenge due to false positives, the models achieved over 99.6% accuracy, demonstrating their potential for real-world deployment [55]. Similarly, in the refinery domain, sensor data collected over a one-year period from a centrifugal compressor was used to train statistical and machine learning models. These models, including linear regression and ARIMA, achieved acceptable error margins in forecasting key operational parameters such as temperature and pressure. In many cases, anomalies were flagged before threshold violations occurred, allowing for early intervention and minimizing downtime [56].

To support structured and standardised approach to data exchange, UNDERPIN is working on streamlining the process of gathering, storing, and sharing product-related data -such as materials composition, lifecycle information, and sustainability metrics- through the creation of Digital Product Passports (DPPs). This includes the development of a Software DPP, which is currently a work in progress; however, some initial samples are already available on UNDERPIN’s onboarding page [57]. As the EU moves toward making DPPs a mandatory requirement in the coming years [58], their integration into manufacturing data spaces ensures that companies are well-prepared for future regulations, reducing the risk of non-compliance.

3.3. SM4RTENANCE - trusted data sharing for manufacturing equipment industry

3.3.1. SM4RTENANCE project overview

The SM4RTENANCE project, a 36-month-duration DIGITAL SME Support Action started in October 2023, involves 44 partners from 11 EU countries, and is dedicated to establishing an industry data space ecosystem. This will be accomplished through the federation of -purpose- specific newly created data spaces and platforms setting up the stage for future interoperability with other existing data spaces, enabling cross-industry data exchange, and therefore tearing down a significant barrier to data valorisation. SM4RTENANCE is implementing demonstrators along multilateral data value chains in both process and discrete manufacturing industries, i.e., textile, e-mobility & electronics, steel & aluminium and automotive. The project includes nine active implementation instances, which we will interchangeably refer to as pilots or trials, dedicated to a wide range of manufacturing sectors: e-vehicles, industrial machines, textiles and metrology. SM4RTENANCE trials are designed to showcase the advantages and business value of the upcoming SM4RTENANCE data space federation within specific use cases. This approach addresses the current operational challenges while paving the way for long-term improvements and innovation in the manufacturing sector.

SM4RTENANCE pilots focus on the following four key areas:

• 1.generative co-engineering and commissioning services for manufacturing assets,
• 2.collaborative net-zero operation services,
• 3.cooperative autonomous condition monitoring services, and
• 4.circular asset management services.

3.3.2. Implementation, benefits, and governance

The SM4RTENANCE project aims to deliver several benefits to factories, including improved data exchange capabilities within a trusted environment, enhanced asset resilience, reduced energy consumption, a lower CO₂ footprint, optimized Overall Equipment Effectiveness (OEE), and extended asset lifetime. The services and abilities that are being deployed to work on top of the data spaces for each different pilot, at the time of writing include predictive maintenance, DPPs, digital twins and general data sharing along the value chain to streamline production logistics and prevent or mitigate disruptions. Benefiting from the ability that data spaces grant their participants to offer their data, one pilot aims to pre-emptively comply with the traceability requirements from the upcoming EU DPP regulations, which will come in effect into textile manufacturing within the current decade. Also in the textile context, but from the weaving-machine production side, another pilot aims to streamline the OEM repair parts just-in-time delivery by conducting component-level condition monitoring.

Exploiting the data space connectivity, one trial aims to accelerate the sharing of asset information for enabling Collaborative Asset Monitoring (CAM) at a large scale via the existing data space SCSN and an upcoming maintenance-oriented data platform. From the predictive maintenance side, a pilot will improve their software-as-a-service failure-prognostics model for electric-vehicle batteries, organizing their data flow and clientele via the Catena-X data space. Another one will conduct predictive maintenance on the customer-owned milling machines by monitoring their condition and linking that information to the quality of the milled parts. Being the most represented topic, predictive maintenance is also the goal of an industrial-robot manufacturer, achieving diagnostics from electrical energy consumption and temperature monitoring.

The variety of pilots in SM4RTENANCE extends to digital twins as well: one pilot will develop and integrate a data space into the metrology value chain on top of which the digital twin of their coordinate measuring machines will be set up. The digital twin will become part of their coordinate measuring machine offers.

At the time of writing, the SM4RTENANCE pilot implementations are rounding up their proof-of-concept status and looking to start growing towards their full deployments. This involves minimalistic versions of the envisioned scope. In these initial-stage, data space service providers have set up the data spaces for each pilot and set up connectors for the designated players. Depending on the specific pilot, this involves off-consortium parties, although in other instances this means connector instances run internally by different pilot partners. Examples of tangible advancements are:

• •Integrating sensor input
• •Integrating with existing ERP (Enterprise Resource Planning) systems
• •Supply chain data collection (DPP application)
• •Anomaly-detection AI models training
• •Data models development (simplified versions in some cases) and data products: a dataset arranged according to a data model, accompanied of metadata informing on the data properties, pricing and usage terms (see below).
• •Conducting data exchanges through the connectors

SM4RTENANCE dedicates project-wide efforts to streamline the preparation of up-to-standards datasets; it concentrates part of the work on data product preparation. Clear criteria of what constitutes a data product are being specified, continuing downstream from the DSSC guidelines [59,60] and the CEN—CENELEC Trusted Data Transactions Group baseline. For instance, the CEN/CENELEC definition considers data products as the object of data transactions, an integral unit that is being exchanged or accessed as part of the latter [61]. Besides the data itself, which has to be arranged in a predefined, sector-agreed shape, a data product contains a plethora of accompanying information, such as its description and purpose of use, data about its origin and sampling, and permissions, limitations and pricing, among other details. The work in SM4RTENANCE extends beyond the creation of precise definitions, as it also encompasses the development or adaption of tools to enable participants to create SM4RTENANCE-compliant data products, i.e., data-transaction units that other parties can readily interpret and make use of. These efforts are crucial to enrich the data pools that will be hosted by data space participants.

Although data space technology is a great enabler to accelerate data exchange among manufacturers, along with their network of suppliers and customers, data spaces are typically deployed around a use case or sector specific goal. This facilitates the definition of common grounds regarding data exchange policy and standards but is otherwise mostly confined to individual sectors. Each data space, whether focused on automotive, machine tools, metrology, textiles, or aviation, provides a trusted environment for participants in that sector to share and consume data under agreed governance rules. On their own, these spaces would risk creating de-facto new silos, even if wider than the original single-company ones. SM4RTENANCE is meant to have the data exchange transcend sectorial boundaries, defining the concept of data space Federation. The Federation aims to be an overarching framework that connects multiple sector-specific data spaces into a coherent whole. The Federation will provide the mechanisms to facilitate the interlink between participants from different data spaces, ensuring that data flows and services can cross organizational and sectoral boundaries. In turn, this will entice the participation of new stakeholders, all while reducing the reliance on big tech companies that are headquartered outside the EU, reducing the power asymmetry between the few large-scale providers and SMEs. Two key aspects currently being addressed are the cross-verifiable identities and federation-wide shared catalogues, both of which are under development in the project. The SM4RTENANCE Federation seeks to have multiple avenues of granting data-space interoperability, i.e., not limiting itself to a single approach or vendor-specific service, which could disadvantage SMEs.

Data space governance is a concept to be developed in full once the data space expects to scale up in participants. At the time of writing, each pilot is completing the deployment of their proof-of-concept stage, and therefore the governance structure is set to be defined during the transition into a full-implementation stage. In parallel, on the larger cross-data-space scale there is current ongoing work dedicated to defining the SM4RTENANCE Federation governance structure, and the roles and responsibilities each participant may take once the Federation is in place and operational. This covers, for instance, data space specialist companies that deploy part of or an entire data space. Another example is the role of the infrastructure providers, who administer the underlying hardware and software where a data space is deployed. Many other roles, like the more regular data providers or data consumers, along with their business obligations and expectations, are being defined for the Federation. In this pursuit, especial attention is paid to existing governance and business schemes from consolidated data spaces, to not reinvent the wheel and make use of real-world-tested concepts [62,63].

4. Conclusions

The EU aims to create a single market for data, where trusted data sharing will be the basis of the digital transformation. Through a holistic strategy that includes regulations and initiatives such as the DGA, the Data Act, and the establishment of Common European Data Spaces (CEDS), the EU is investing in and promoting frameworks that ensure sovereignty, interoperability, and security in data exchange across strategic sectors. One of these sectors is manufacturing, where all engaged parties in the industrial ecosystem (suppliers, operators, technology providers, and SMEs) could discover and exchange data under clearly defined legal and governance terms. Industry players can find innovative solutions tailored to their needs or reduce operational costs. SMEs will have access a variety of industrial data that they would otherwise not have or would need to spend significant resources to obtain. However, the effort to establish a manufacturing data space faces several challenges on both the technical and business sides. The two EU-funded projects, UNDERPIN and SM4RTENANCE, address specific points to contribute to the wider adoption of CEDS, especially for SMEs.

The UNDERPIN addresses technical barriers by introducing semantic interoperability through knowledge graphs and ontology-based frameworks while enhancing trust with blockchain-enabled smart contracts and standardized legal agreements. The project introduces a semantic layer based on the IDS reference architecture and uses OBDA technologies to standardize metadata and facilitate seamless, machine-readable data exchange. This ensures that heterogeneous manufacturing systems can communicate effectively, supporting the EU’s ambition for standardized data ecosystems. The use of federated vocabulary hubs further strengthens scalability and governance, creating the basis for seamless cross-sectoral integration. The project also advances the EU’s objectives for secure and trusted data spaces through mechanisms that ensure legal compliance, data sovereignty, and transparent governance. Blockchain-based smart contracts combined with GDPR-aligned legal frameworks create trust in data transactions. Through pilots in oil refineries and wind farms, the project demonstrates how predictive maintenance models and secure data-sharing frameworks can deliver long-term value. The development of DPPs further prepares stakeholders for future regulatory requirements, ensuring that data spaces remain viable and adaptable beyond the project’s duration.

The SM4RTENANCE project is addressing the limited use of industrial data. In order for data to be put to value, it is key that it reaches the hands of the interested party, who must first be aware that such data exists to enter the market for it. Such visibility is afforded by the data space feature of a catalogue that describes all the data assets and services available within a data space. At the same time, a requirement to engage in data exchange within data space ecosystems is that the data itself must be adequately prepared: data should be organized in a commonly understood model. In other words, proper data arrangement and annotation prior to exchange need to become part of the data culture.

SM4RTENANCE aims to make this transition seamless for manufacturers or supply-chain participants through the development and adaptation of data-product creation tools. These data products can encapsulate raw machine-sensor records, production data from ERP systems, or data from any other source into forms that can be reliably exchanged and consumed. The SM4RTENANCE project has seeded the deployment of pilot-scale data spaces, each built around a specific use case and involving, at their inception, a small number of business-related companies. This will stimulate sovereign data sharing, with the data owner maintaining sovereign control over the data -i.e., being entitled to condition how the data will be acceptably used. On top of this, SM4RTENANCE is building a federated structure, seeking to enable cross-sector collaboration as needed (e.g., machine tools, metrology, textiles, automotive), ensuring that interoperability and access across spaces is possible, and securing new pathways for opening today’s data silos.

The successful implementation of data spaces for manufacturing is essential for realizing the EU’s vision of a unified data market. Projects like UNDERPIN and SM4RTENANCE demonstrate how targeted technical solutions and innovations, as well as cross-sector collaboration within the industrial ecosystem, can overcome existing barriers. By doing so, they support the European vision for innovation through data sharing and establish best practices, templates and standards for others who intend to follow in the same processes.

Ethics Statement

The authors have read and follow the ethical requirements for publication in Data in Brief and confirming that the current work does not involve human subjects, animal experiments, or any data collected from social media platforms.

CRediT Author Statement

Aristotelis Ntafalias: Conceptualization, Visualisation, Writing-original draft, Writing -review &editing, Supervision; Marcelo J. Ambrosio: Conceptualization, Methodology, Visualisation, Writing-original draft, Writing -review &editing; Alexandros Nizamis: Methodology, Writing-original draft; Panos Papadopoulos: Conceptualization, Supervision, Writing -review &editing; Oscar Lázaro: Conceptualization, Writing -review &editing; Carmen Polcaro: Writing -review &editing.