1 Introduction

Much has been said about the value of data, such as “data is the new oil,” originally coined in 2006 by Clive Humby, the co-creator of British multinational retailing giant Tesco’s Clubcard loyalty program and founder of a successful data analytics agency (see [1]). But few companies beyond Amazon, Facebook, and Google seem to have mastered refining and distilling crude data into value. Their market capitalization overshadows all others (see [2]). If data savviness correlates with market capitalization, then the stock market is proof that somehow being truly data-driven and monetizing data is so much easier said than done. One solution to “democratize” value creation with data is the concept of a data space. Yes, yet another concept born in an academic research lab, in this case in the labs of the Fraunhofer Society, Europe’s largest application-oriented research organization, which began creating International Data Spaces (IDS) technology in 2015. On the one hand, we love the rigor that comes with scientific methods, such as validating the efficacy of a vaccine (are you sure it will help … rather than kill me?). On the other hand, however, the business community has learned that not every idea from science translates into economic value and profit. How fitting, then, that experts and engineers at Deutsche Telekom, a for-profit entity, have put data spaces and IDS (International Data Spaces) technology to the test. “A data space is a shared, trusted space for transactions with data […] based on common standards (or values, technologies, interfaces) that allow and promote transactions with data” [3, p. 109]. Deutsche Telekom has been counting on IDS technology and became a pioneer participant in the International Data Space Association (IDSA) created to promote IDS. Deutsche Telekom is deeply involved at multiple levels, at the Board level as well as in the trenches, such as the launch coalition. Contributions include real-world IDS prototyping in the National Platform Future of Mobility’s (NPM) RealLab Hamburg (“Mobility with IDS: Adding the ‘N’ in NPM to RealLab HH,” link), “Rulebook” writing, and formulizing certification procedures. Based on our IDS work and empirical findings, we conclude rather provocatively that data space is a law. And we explain why, using examples from two verticals—mobility and industry—to substantiate and illustrate our findings. Lastly, we will abstract from these examples to share with you the commonalities across use cases and verticals that have been observed with our IDS-based data space applications to present a conclusion as a type of blueprint or recipe for business success with a data space.

1.1 Data Is Broken … and Data Space Is a Law

Date spaces are shaping up to be a kind of law like Moore’s law (the doubling of computer power every 2 years), which is not a law of nature but rather a trend that follows or is enabled by fundamental developments. In the case of Moore’s law, it is miniaturization and scaling in manufacturing; with a data space, it is economics of specialization and modularization. The latter is not new at all but has been spreading throughout the physical world (think container shipping, [4]) as well as in software. In software, modularization has spread across the stack from the transport layer (example of the Internet’s packetized messaging, the Transmission Control Protocol/Internet Protocol or TCP/IP) to the application layer with web services in the early 2000s (such as Microsoft’s .Net; [5]) and Dockerization yesterday (“Dockerize an application,” link). Today, attention and innovation have shifted to the data layer. The reflex has been to hoard and pool data on hard drives (it’s mine, I want to keep it), in databases and data warehouses since Michael Porter’s seminal Harvard Business Review article on IoT, and in data lakes [6]. Today, it is clear that this approach is broken.

  • First, it is costly to store data; it is money spent without an immediate benefit.

  • Second, additional money is required to keep track of it. Without keeping track, the data may as well not exist in the first place. Tossed on a pile, pumped into a lake … good luck finding anything in this fishing expedition: A structured search space is required for the search to be successful, such as using keywords and tagging.

  • Third, these keywords must be carefully chosen and hierarchically sorted for anybody to find anything without searching forever. For experts, a library classification system such as the Dewey Decimal Classification (DDC) comes to mind. For the rest of us, the food analogy is helpful here: We all buy food. We do not slaughter any animals in our bedrooms or grow vegetables in our living rooms; instead, we buy food portioned, packaged, and labeled. We do this going to a “food space” or supermarket, where we consult the store directory, walk to the appropriate aisle, find the right shelf, and compare items on it based on labels. This is what happens pretty much anywhere in the world. With data, there is no supermarket, no directory, no aisles, no shelves, no labels (the Telekom Data Intelligence hub provides a hosted data store with a directory, complete with a self-service process to put data products on the shelf; link). Even basics such as portioning and packaging are a mystery: We know we buy eggs by the dozen, butter by the pound, and milk by the liter. But how do we consistently measure data quantity? By terabyte or length of a time series, or frequency of time series values, etc. (see “Data: Quantity or quality”; link)? And here it is: we dream of monetizing data but have not even agreed on how to measure any orders.

  • Fourth, beyond the measurement problem, with data, it seems that most of us do slaughter animals in our bedrooms and grow vegetables in our living rooms. Data tends to be home-made and made to order, just like cars before Henry Ford turned them into a product [7] and economized on production using factory automation [1]. This explains why data scientists spend more than 80% of the time of a data analytics project on data: hunting it down, cleaning it, slicing and dicing it to understand the information contained in it, and finally preparing it for “dinner” (see our research on “Data is broken”; link).

  • Fifth, data is going to explode; look no further than the introduction of 5G and IoT, the Internet of Things, which turns every other thing into a website where everything is tracked. Any solution that has been developed to mitigate some of the aforementioned issues is a Band-Aid at best; it will not stop the bleeding. Estimates suggest that the amount of data will increase from 33 zettabytes (1 zettabyte = 1021 bytes) in 2018 to 2142 zettabyte by 2035 [8]. If the cost of data storage is already painful today, it will be a nightmare tomorrow. If your data scientists are already struggling to cook dinner for the family, wait for disaster to strike when you open the restaurant.

The natural solution is to modularize data. Applications pull data, and applications have gone from monolithic to service orchestration (see microservices, dockerization), so data could follow suit and become more federated and distributed, too. Just as elsewhere, economics could trigger a shift to just-in-time data and an n:n data infrastructure. Why has this not happened yet? A key problem so far has been data sovereignty, or, more precisely, the lack of it. Data sovereignty is the right to have control over one’s data. Once a file is sent, anything could happen to it. This where the IDS standard comes in to enable a setting that can ensure data sovereignty. It’s emergence as a data-sharing standard coincides with new regulation. As a key pillar of its data strategy, the European Commission is proposing the European Data Governance Act (DGA) to foster the availability of data for use by increasing trust in data intermediaries and by strengthening data-sharing mechanisms across the European Union.

1.2 Data Exchange and Trading for Better and New Business

No doubt about it: Data is becoming a fundamental raw material for the success of national economies. On the one hand, it is in the public interest to enable and even boost the availability, exchange, and use of data. On the other hand, it is in the individual interest of companies to benefit from the added value that greater use of more data promises. But there are some obstacles that are now well-known and widely discussed. There is a strong lack of:

  • Interoperability in technology and semantics

  • Data quality

  • Legal frameworks according to data

  • Security

  • Digital maturity of many companies and—most importantly—a fundamental mistrust between participants, often competitors in the markets

To overcome these obstacles, new principles, new technologies, and new business roles are introduced and will be further expanded and innovated in the future. Data sovereignty principles and services are at the heart of this. They ensure that control of data, especially sensitive data, is maintained, and are ready to bring more trust into the data business. Additional strong security technology is applicable, and legal frameworks and laws define the trusted environment that every thriving business needs. Last but not least, new roles to enable and maintain data sharing are emerging: data intermediates, data brokers, and data traders, often certified and neutral to economic sectors, because trust in competition means security and neutrality. You could compare them to banks, which care about money; only, they care about data. Both elements are fundamental. No economy will work without them in the future. The following use cases from the areas of mobility and industrial production show how the benefits of data exchange and data sharing already work today, applying existing technologies und generating advantages for stakeholders. In future business history books, they will be known as “early adopters.”

2 Transition to Data Space-Enabled Mobility

2.1 From Auto (Hardware) to Mobility (Service)

The automotive and transport sector is facing a structural reset and will change as profoundly as it did when Henry Ford industrialized the business in the 1910s with the introduction of the moving assembly line, which made individual motorized mobility affordable for the public. Today, radical new vehicle technology such as electrification and autonomous driving is emerging just as digitization is spreading from other industries like publishing and music [9, 10] to automotive with symptoms such as “connected car” services and direct online sales to consumers (DTC). And all of this coincides with climate change and the saturation of urban space with cars to create a perfect storm, which in turn cumulates into a sociocultural and political shift away from car-centric priorities (“Verkehrswende”; see, e.g., [11], link). As a consequence, and in terms of business model change, industry experts foresee a revenue shift from selling hardware to mobility services (see “Metamorphosis of auto into mobility,” link).

2.2 Traffic Is Broken: New Connected Mobility to the Rescue?

How do we get to our destination faster, in a more environmentally friendly way, but also more cheaply and hopefully more conveniently, especially in congested urban areas and cities? One solution is intermodal travel, linking different modes of transport for a seamless journey from point A to point B: start the first leg by car, switch to rail, and finish the last leg with a micromobility offer, such as an e-scooter or on-demand shuttle bus (see Fig. 30.1 for cascading intermodal scenarios). On the one hand, many major cities are experiencing traffic gridlock, more and longer traffic jams [12], and citizens are complaining about air and noise pollution (we published an analysis on “Stuck in Traffic,” link). On the other hand, new technologies are enabling mobility innovation. For one, the widespread use of smartphones and apps (mobile maps, turn-by-turn navigation, payment by smartphone) has enabled new connected mobility service offerings, such as ride-hailing, carpooling, and car-sharing services from new companies like Uber, Lyft, and Free Now. For another, better battery technology has driven innovation in two-wheelers such as e-bikes and e-scooters—in Germany, laws even had to be changed [13]. As a result, smartphone-based mobility platforms with multimodal or Mobility-as-a-Service (MaaS) offerings have emerged in major cities, such as Jelbi from Berlin’s main public transport company Berliner Verkehrsbetriebe (BVG) and Switch from Hamburger Hochbahn (HHA), Hamburg’s public transport operator. Streets have also become visibly more colorful, dotted with electric scooters or e-scooters—for example, from Bird (black), Lime (white), Tier (green), and Voi (red)—and shared bicycles from providers such as Donkey Republic (orange) and Next (silver). But despite many new transport options, especially for short distances, no intermodal service has yet emerged. Where is the problem? Wouldn’t micromobility make public transport, which is tied to rigid routes and time tables, more flexible and thus more attractive? What are we still waiting for? The mobility analytics experts at Deutsche Telekom are trying it out now, both with data analytics in simulation tests in our Berlin digital twin and in everyday life with funding from the Federal Ministry of Transport in the RealLab Hamburg.

Fig. 30.1
figure 1

A simple intermodal model and “smart” scenarios

2.3 Synthesis of Simulation and Real-Life Data Space Prototyping

Deutsche Telekom has a long, rich, and successful history supporting mobility and transportation. Its T-Systems business has been the leading information and communication technology (ICT) provider to the German automotive industry for the past decade (Automotive IT 2020 [14], pp. 18–20, link), from operating connected car back-end systems (“Globally connected vehicles,” link) to providing smart charging infrastructure (“Smart electric vehicle charging,” link). Deutsche Telekom is already actively participating in the automobility transformation: from contributing to the German government’s National Platform Future of Mobility (NPM) initiative to providing connectivity solutions for micromobility. Furthermore, Tier, a leading micromobility company, which received EUR 250 million in funding in 2020 on top of the more than EUR 130 million it had already raised, was launched at Hubraum in 2018 (“How Tier mobility are set to change transport forever,” link), Deutsche Telekom’s tech incubator with campuses in Berlin, Krakow, and Tel Aviv (“Our mission,” link). Now, mobility analytics experts at the Telekom Data Intelligence Hub (“Smart mobility,” link) at Deutsche Telekom IoT are working with T-Systems to combine simulation experiments (theory) with real-life prototyping of a mobility data space to enable new mobility offerings for their business clients (practice).

2.4 Simulation and Berlin Digital Twin: Benefits in Theory

End-user benefits are a necessary condition or conditio sine qua non for the success of intermodal travel. Without clear end-user benefits, everything else is a waste. But how can we find out? How can benefits be estimated if intermodal travel does not even exist yet? How can we test the impossible to prove the probable? Simulation can provide answers. It is a valued scientific tool (e.g., [15,16,17]) that has evolved in the field of economics from the foundations laid by Nobel Prize winners Simon [18] and Smith [19]. For mobility, there are already various tools available in Germany, for example, SUMO (Simulation of Urban Mobility), an open-source traffic simulation by the German Aerospace Center [20, 21], and the virtual measurement campaign (VMC) software by the Fraunhofer Institute for Industrial Mathematics ITWM [22]. For our purpose of a rough estimation of end-user benefits, a significantly simplified model and procedure was designed, coupled with a scientifically rigorous implementation and experimental strategy (for a detailed description, see “Simulating intermodal mobility,” link; also “Calculator powered by machine learning: Mobility-as-a-Service,” link). For our simulation in a digital representation of Berlin, our Berlin digital twin, the results speak a clear language: intermodal travel is faster (for a detailed description, see “Berlin digital twin and intermodal travel,” link). Experiments show that time savings can quickly exceed 10% and double with “smart” linking of transport options (see Fig. 30.2 with results for scenarios S2 and S3), for example, by recommending not just any parking space but the one with an e-scooter connection nearby to speed up the last leg (see scenario S3 of smart parking with smart e-scooter in Fig. 30.1). All in all, simulations help develop insights beyond the obvious. Particularly “smart” options reveal very specific data needs, such as parking occupancy prediction data, which turn into requirements for our real-life prototyping effort.

Fig. 30.2
figure 2

Speed advantage of intermodal travel

2.5 Data Spaces: From “Mine, Mine, Mine” to “Win-Win-Win”

What are we waiting for? Okay, consumer benefit is only the first requirement for success. And in this case, the time advantage was the only factor taken into account. It may be the most important, but it is only one decision variable for the customer. Other factors include cost and convenience, or more precisely, the right combination of all these parameters. And because of climate change, environmental protection must also be considered. Coming from the supply side, all companies involved must understand to what extent and in what way consumer benefits can be monetized, i.e., turned into money. This second prerequisite for success requires data sharing. “Smart” only works with data. Take Uber, for example: matching a passenger with a driver and vehicle depends on data, namely, the locations of passenger and driver, availability, and traffic conditions, each with the same timestamp. Without this data, Uber would not be able to orchestrate any transport service. “Uberization” requires that all stakeholders—passengers and drivers—share this data in near real time. The same applies to intermodal mobility: data sharing is key. The problem is obvious: some transport options are in direct competition with each other (public transport, cabs, ride-hailing, e-scooters, etc.). As a result, some service providers are competitors and do not trust each other. Everyone wants to “own” the customer or the customer connection and, therefore, keep the customer data to themselves. In addition, compliance with data protection regulations, namely, GDPR (General Data Protection Regulation of the European Commission, DSGVO in German), can serve as an excuse not to share important data. Many will remember the famous scene in the Pixar blockbuster movie “Finding Nemo,” where a flock of seagulls are foolishly chasing food without any coordination and with every seagull screaming: “mine, mine, mine.” It is not unlike what is happing with urban mobility in 2020. So far, everything is multimodal, and there is no coordination yet to orchestrate different modes and providers to create a seamless end-to-end-user journey. If intermodal worked, it would be a triple win outcome – “everyone’s a winner”: consumers (better travel), providers (more business), and cities (cleaner, safer, less noise).

2.6 From Simulation to Reality: IDS in NPM and RealLabHH

This dilemma could be remedied by technology that is designed to facilitate data sharing between parties that do not necessarily trust each other, such as competitors. This is where our work on the IDS standard comes into play (IDSA RAM 3.0, link). IDS is a DIN SPEC standard for data exchange while maintaining data sovereignty ([23, 24] and “IDS is officially a standard,” link). Or, to put it in simple terms: IDS technology enables parties that do not trust each other to trust a particular data transaction. And that is exactly what our involvement with the real laboratory Hamburg or RealLabHH is all about, determining whether such IDS-based solution is possible. The simulation has helped establish that there are clear benefits for the consumer, for the demand side of the business calculation. What about the other side, the supply side? With RealLabHH, it is now possible to look at the supply side, how IDS can support the orchestration of a service offering involving multiple transport companies. To this end, Telekom is working with the Telekom Data Intelligence Hub (DIH, link) team and the Urban Software Institute (link) to build a fully functional demonstrator that will be presented with RealLabHH at the 2021 ITS World Congress (link). Since IDS is a core technology of the European GAIA-X distributed data infrastructure (link), this will effectively be a first mini-GAIA-X mobility data space demonstrator with real companies involved, such as Hamburger Hochbahn (“Startschuss für das Reallabor Digitale Mobilität Hamburg—mit Bundesminister Scheuer,” link). More details on this project in our DIH story “DT and NPM” (link) and “Mobility with IDS” (link). The goal is to show how a federated data structure with sovereignty controls based on IDS can create benefits for all stakeholders (see Fig. 30.3): new, better travel options for citizens, such as intermodal travel, and new business opportunities for both established public transport companies (e.g., convenience offers combining rail and on-demand shuttles) and new micromobility providers (e.g., connecting e-scooters to public transit). Specifically, a planning app for a door-to-door service between Hamburg and Berlin is to be implemented as a demonstrator. With the initial IDS components, such as connectors, a broker, and identity management up and running, the project is already in the midst of fine-tuning and experimenting with data usage policies and enforcement options to facilitate automated machine-to-machine interaction that can ensure data sovereignty for all stakeholders. The timing could not have been better. Because at the same time, the German government is building the first infrastructure components for a German Mobility Data Space with its Acatech Datenraum Mobilität project (DRM; [25]). If these DRM components are installed in time, they could be used for the RealLabHH demonstrator to ensure compatibly with DRM, making it a first, truly operational GAIA-X use case.

Fig. 30.3
figure 3

An intermodal travel demonstrator with IDS-based data interoperability

2.7 Catena-X Automotive Network and Deutsche Telekom

Many industry observers were caught off guard when German Chancellor Angela Merkel proclaimed data spaces as a top priority for the German automotive business at the “Autogipfel” auto summit in the fall of 2020 [25]. What is a data space? Today, the term is used almost as a matter of course. Only a few months later in the spring of 2021, the German federal minister for economic affairs and energy, Mr. Altmaier, hosted the president of the German Association of the Automotive Industry (VDA), Ms. Müller, together with the CEOs of BMW and Daimler, Messrs. Zipse, and Källenius respectively, for the christening ceremony of Catena-X, a network for secure and cross-company data exchange in the automotive industry. And data spaces were right on center stage. One key enabler of the Catena-X network is the pan-European GAIA-X data infrastructure initiative, which in turn features data spaces built on data spaces technology, specifically the International Data Spaces (IDS) standard, DIN Spec 27070. Catena-X has evolved from first CEO-level talks. In December 2020, Bloomberg News announced that some of the biggest German companies had joined forces to build a German auto alliance [26]. The founders of the partner network include BMW, Deutsche Telekom, Robert Bosch, SAP, Siemens, and ZF Friedrichshafen. In the meantime, additional companies have joined the initiative including Mercedes-Benz AG, BASF SE, Henkel AG & Co. KGaA, Schaeffler AG, German Edge Cloud GmbH & Co. KG, ISTOS GmbH, SupplyOn AG, the German Aerospace Center (DLR), Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V, and ARENA2036 as well as several small- and medium-sized companies (SMEs; see [27]). Now, all participants have pledged to cooperate and collaborate jointly in an open network to advance and accelerate cross-company data exchange throughout the entire automotive value chain. Specific examples include continuously connected data chains to create digital twins of automobiles. Deutsche Telekom is an active participant with its experts on cloud computing, digital twins, GAIA-X and IDS, and, of course, mobility analytics from its Telekom Data Intelligence Hub team.

3 Industrial Data Spaces

For the manufacturing industry, there is tremendous value in data spaces; in shared data spaces, the value will become excessive. But to get there, there are some serious obstacles. Data is used and needed throughout the complete industrial production environment, and all processes converge on the shop floor: engineering, production, logistics, order management, service, supply chain, product life cycle, and life cycle of production machines themselves.

Very often, this data is sensitive, because it contains intellectual property (IP), unique selling propositions (USP), competition-critical information on products, production processes, or business models. For this reason, the use of data sharing or shared data spaces is not yet very well-known in this area of industrial production, mainly for competitive reasons.

On the other hand, there are great benefits in using and sharing more data. With respect to the mentioned sensitivity, data sovereignty can be seen as a critical enabler for more data availability.

This chapter focuses on examples and use cases that are close to the shop floor in terms of the production/manufacturing process and its machines. There is no doubt that there are other valuable use cases in all other mentioned processes.

3.1 Most Wanted Use Cases with Own Data: And with Shared Data

In production, there are currently seven most wanted use cases that are built on data and data sharing (Table 30.1). The challenge is that you cannot start at level 7; you have to start at level 1 and work your way up step by step. The good news is that you gain more business value at each step.

  1. 1.

    Dashboards: An information and assistance service for those responsible. The prerequisite is that machine data is available. Visibility through dashboards can be provided on various devices such as screens, computers, cell phones, or smartwatches.

  2. 2.

    Transparency: In addition to the dashboard, not only individual machines or processes are displayed, but all relevant ones in the context of interest. This step provides added value, and it provides an overview and enables dependencies to be identified. Even without another analytic algorithm, transparency is of very high value, e.g., getting an overview of a complete process or production line. The challenge is to make data from different sources available. However, since we are dealing with our own sources, this is quite easy. The added benefit: reaching the maturity of transparency for a defined context is the best starting point to go the next steps.

  3. 3.

    Condition monitoring: In addition to transparency, some limits and values are defined, which are monitored automatically. The degree of transparency and assistance is thus higher.

  4. 4.

    Collaborative condition monitoring: This is the first step that is not possible without data exchange. The basic idea is that limits and values in the operation of machines and productions lines will become much more precise, if as many real experience values of operational data can be used to define them. This can be a data exchange between machine operators, but it is most fruitful to involve machine builders, integrators, or component producers.

  5. 5.

    Anomaly detection: This is the first step where AI comes into play. Translation: First, large amounts of data are needed to train the anomaly algorithms, much more data than a company controls itself. Second, deep data is where high business value can be achieved, because problems can be identified first before humans are able to notice them. Therefore, good predictive maintenance is a prerequisite for the next step.

  6. 6.

    Predictive maintenance: The idea is simple and well-known; maintain early before something is broken, but not too early so as not to waste time and money. The value of predictive maintenance lies in the precision of the maintenance point. This can be derived first from precise limits and values for step 4, collaborative condition monitoring; second, from high-value anomaly detection; and third, from a lot of operational data exchanged with the company. The rule: better predictive maintenance with better data exchange.

  7. 7.

    New business models, the holy grail: Selling lifelong service by subscription instead of selling a machine once: air instead of pumps, holes instead of drills, mobility instead of cars, etc. The business case of such business models only holds if you can ensure that the service works when it needs to work, for example, 24/7. Translation: You have to manage maintenance before something breaks, i.e., you have to master level 6, predictive maintenance, enriched with excellent spare parts and service processes.

Key to climbing the seven value stages toward new business values is the right use and handling of data.

New core competencies around data are required: access, transport, and storage of data, quality management of data, exchange of data, analysis of data, and, last but not least, the creation of data-driven business models such as ecosystems from it. Security and legal issues are important in all steps and need to be managed carefully. All in all, it boils down to managing data as an asset, just as all other assets are managed in the company.

But the final and most mission-critical competence to open the door to high-value use cases is sharing and exchanging data to enrich one’s data base to a valuable extent (Table 30.1).

Table 30.1 Shared data needed for high-level use cases

3.2 The Umati Story

One example that enables new data-driven business models in the manufacturing industry and points the way is umati, the universal machine technology interface. This is a community of component and machine manufacturers and operators organized at the Verband Deutscher Maschinen- und Anlagenbau (VDMA), the German Engineering Federation. They provide an open platform communications (OPC) companion specification for cross-industry machine builders in all segments, which is to become a new world standard language for machines, a common semantics.

With reference to steps 1–7 of the most wanted use cases mentioned above, this is important from step 2, transparency, onward. For valuable information within a production line, e.g., about energy consumption, the information of all machines in the context must be available in the same semantics. It becomes mission-critical from step 4 onward, when it comes to data exchange. Collaborative condition monitoring requires collaborative semantics.

But it worked very well. The umati demonstrator from VDMA powered by T-Systems runs as a 365 × 24 × 7 testbed for all interested machine builders (see Fig. 30.4). Via self-onboarding, they can connect their machines and forward the data to the dedicated dashboards—worldwide. The demonstrator shows that data transparency and exchange across machine builders are possible.

Fig. 30.4
figure 4

Cross machine data sharing enabled by umati

The implementation of the umati concept in the new business models uses the principles of the demonstrator, expanded by business model-driven roles, governance, security, and services. Selling a machine service instead of the machine, as mentioned above, requires the machine builder to connect as many of its customers’ machines to its data-driven predictive maintenance and service platform. With the umati concept and the demonstrator principle, this is possible, worldwide. But does the customer want to do it? And why? There are two drivers for customers. First, own high-value benefits and, second, sufficient trust and control in the process. Point 1 must be served by the entire business model, so that no data provider and partner are forgotten on the benefit side. Point 2 can be ensured through the use of data sovereignty concepts.

3.3 Data Sovereignty for More Industrial Data Exchange

The last and most difficult obstacle to collaborative data use and data exchange is trust between partners. Moreover, a concept of usage control is helpful to maintain control over the data: Who does what and when how often with my data—this is the level of ambition. Only those who properly master the obstacle of trust and control can climb the steps to the high-value use cases with the most attractive business cases and benefits.

The data sovereignty concept of IDSA and GAIAX is a strong concept for sustainable data exchange. It takes care of all points of trust and control between the partners exchanging data.

The concept of digital sovereignty opens a world of new business opportunities and even completely new markets. And—moreover—it enables and accelerates existing ones with new business capabilities by adding data sovereignty services. This “add on” is valuable wherever a lack of data or a lack of trust, or both, present themselves as barriers for further growth.

The enablers for business models based on data sovereignty originate from the core services of digital sovereignty technologies and architectures such as IDS or GAIA-X.

What is needed for adding data sovereignty to accelerate data exchange in your business or even to organize and build data spaces to bring together data providers and data customers is the integration and use of the following services:

  • For bringing parties together, data intermediary and service intermediary, if necessary, neutral partners are proposed for competitive reasons

  • For semantic interoperability: vocabulary publisher and provider

  • For governance key services: identity authority, clearinghouse, and certification services

  • For support and integration: software developers with specialized data sovereignty know-how.

These roles and their related assets are specifically defined, e.g., in RAM, IDSA’s reference architecture. To summarize briefly, beyond the key source of data suppliers, seven key roles are relevant to running data sovereign data spaces (see “New Business Models for Dataspaces,” [28]).

Data suppliers produce and/or own data that can be made available in the dedicated data space. Depending on the business model and the operational model in place, the basic roles typically assumed by a data supplier are data creator, data owner, and data provider.

Data intermediary: The data intermediary acts as a trusted data broker and manages data exchange in data spaces. The data intermediary knows the stakeholders, can take care of common roles, routes data, stores data on demand, and manages metadata and the ecosystem of available data sources. An organization acting as a data intermediary can also assume other basic intermediary roles at the same time. Consequently, assuming additional basic roles means assuming additional tasks and duties for the data intermediary to execute. But one key rule is required: a data intermediary takes care of the data, but as a trusted intermediary, it never uses or analyzes the data itself.

Service provider: Services offer various functions, such as data analysis, data integration, data cleansing, or semantic enrichment of data. The service provider is a platform operator that provides services (i.e., an app, app store, including computing time as a trustee), metadata about services, or both.

Service intermediary: Entity that ensures the up-to-datedness and allocation of services on the data space via the management of metadata (yellow pages for services) about new and existing services. Provider of an interface for the data providers to provide metadata on their available services.

Vocabulary provider: Vocabularies can be used to annotate and describe data assets. A vocabulary intermediary technically manages and provides vocabularies (i.e., ontologies, reference data models, metadata elements). Vocabularies are owned and governed by the respective standardization organizations. A vocabulary intermediary typically assumes the basic roles of a vocabulary publisher and/or a vocabulary provider.

Clearinghouse: Functional instance that ensures the verification of financial/data-based transactions (both in terms of data exchange and monetary transactions) in the data space. Neutral role for the management of transactional metadata.

An identity authority provides a service to create, maintain, manage, monitor, and validate identity information from and for IDS participants. This is imperative for secure operation of the data space and to prevent unauthorized access to data. Each participant in a data space therefore inevitably has an identity (describing the respective participant) and uses an identity (for authentication). Without an identity authority, sovereign data exchange is not possible in the data space. It is an essential component of the data space that can be provided by a company to all participants of the ecosystem.

Certification establishes trust for all participants within the data space by ensuring a standardized level of security for all participants. Certification must be performed in two areas: certification of the operational environment and certification of components. This procedure is performed by two roles, the certification body and an evaluation facility. The evaluation facility performs the certification, while the certification body monitors the process, manages quality assurance, and provides guidance throughout the process. These roles ensure that only IDS-compliant organizations are granted access to the trusted business ecosystem. In this process, the certification body is responsible for setting up the certification scheme, which includes specifications of supported components and profiles, criteria catalogs, test specifications, and certifications processes. In addition, the certification body approves evaluation facilities and monitors their actions and decisions. At the time of publication, the IDSA headquarters acts as the initial certification body.

3.4 Make Your Choice

Of course, there are new business opportunities out there to offer some or even all of the above services for dedicated and upcoming industrial data spaces. On the other hand, it might be a good idea to buy these kinds of basic data sovereignty services from a trusted supplier and build a data space or collaborate with data spaces or with partners in data spaces producing the most wanted use cases.

Make your choice of which role you want to play and manage it properly. If you choose more than one role, pay attention to the different management goals.

  1. 1.

    Be a data provider: Look for the data spaces most relevant to you, where you can find the most customers for your data, and the greatest benefit for you

  2. 2.

    Be a service provider: Where can you get the best data you need for your service?

  3. 3.

    Be a data space platform provider: What makes this platform attractive to many data providers and service providers?

  4. 4.

    Be a data or service consumer: What data space platform gives me the best offers?

  5. 5.

    Be a basic service provider: Attract as many platforms as possible using your service as a standard

Make your choice—and start.

4 Conclusion

Deutsche Telekom has been actively involved in IDSA right from the start and is heavily involved at many levels from Board membership and “Launch Coalition” to “Rulebook” writing, because IDS technology is trying to do the right thing—data sharing with sovereignty—and looks promising. But Deutsche Telekom is a for-profit entity, and while all of us in business like the rigor of science, we have also learned that the proof is in the practice. That is why Deutsche Telekom also took the risk of using IDS technology. And we did not just start yesterday. Instead, we began early, in 2018, and applied it in more than one vertical to increase the sample size and broaden the empirical base, so to speak. Our pioneering applications of IDS technology in mobility and industry now give us insights into what it takes to turn IDS into a business success. We have abstracted and condensed our cross-industry lessons learned to share it with you.

  • First, joining the opportunity is easy. It is also affordable, because it is cloud-based and comes with platform-as-a-service (PaaS) offerings that let you “think big, start small, and fail and scale fast.” Find out for free: Test-drive the Telekom Data Intelligence Hub (link).

  • Second, the benefits of IDS are real, and the technology is ready for prime time. We have certainly started to put it to use. For example, GAIA-X, the pan-European hyper-cloud initiative of the German and French governments, relies on IDS as the core technology for its federation services layer (GXFS), and Deutsche Telekom is among the partners helping to implement it.

  • Third, the German government is building a first GAIA-X-compatible data space for mobility under the auspices of Acatech, the German Academy of Science and Engineering [25].

  • Fourth, both GAIA-X and the German government have chosen to rely on Deutsche Telekom, among others (link). And surely, Deutsche Telekom has proven that it is a trusted, neutral provider of critical infrastructure. It is also Europe’s largest telecommunications company, one of the most trusted brands, and #GoodMagenta and #GreenMagenta.