Smart Data – The stuff innovations are made of

Digital data play an essential role in the design, planning and control of industrial processes. As well as being needed as input, data are also generated and output in large quantities and can be utilized to develop new products, services, and business models. Fraunhofer collaborates with numerous companies in projects to develop concepts and methods that will enable the digital treasure trove to be utilized in a secure environment, thereby extracting more added value from industrial processes.

New business models

Björn Heinze is in a hurry as usual. He’s been called to a last-minute meeting in Cologne that starts in two hours, which leaves him very little time to get there from his engineering studio in Bonn. Heinze doesn’t own a car: he sold his last one a while ago, when parking became almost impossible in the downtown areas of the city. Instead he picks up his smartphone and books an autonomous electric vehicle via a car-sharing platform. In next to no time, his order has been processed and the car rolls up to his doorstep, guided there by the user profile he previously registered on the service provider’s website. The journey would normally be a test of patience, because the roads to Cologne are frequently congested, but the self-driving vehicle’s onboard navigation system quickly finds an alternative route that brings Heinze to his destination relaxed and on time.

Admittedly, this is a fictive scenario that doesn’t quite work out like this at present, but the technology on which it is based is very real. Many of the processes involved already exist – scheduling the meeting, booking a vehicle and paying the rental fee, geolocation, finding alternative routes, locating a charging station in Cologne to recharge the vehicle’s batteries for the homeward journey (because there might not be enough juice left to cover the whole distance), and invoicing the amount of electricity consumed. The protagonist in this story might not be aware of it, but huge amounts of data flow back and forth in the background, shuttling information between the car-sharing website, the parking facility operator, the utility company and its charging infrastructure, the traffic information service, the rental vehicle, and of course Björn Heinze himself.

The economy of data

The mobility revolution is in full swing. Auto manufacturers have woken up to the idea that the old business model of selling cars is becoming obsolete, and that they need to rebrand as mobility providers. People in the automotive industry are beginning to ask themselves whether the value they create by providing digital services might one day soon exceed the value they create by selling vehicles. Similar reflections are gaining ground in many other industries, for instance in medical and mechanical engineering, where the cost of devices and machines is falling steadily and an increasing share of value creation is accounted for by services such as predictive maintenance. The economy of things is gradually giving way to an economy of data – with far-reaching consequences. What these consequences are, and how industry can draw the greatest benefit from them, are questions currently being investigated by numerous Fraunhofer Institutes.

The rise of data-driven business models is forcing companies to completely rethink their corporate structures and the way they organize their internal and external supply chains. Whereas, in the past, systems were painstakingly configured by highly trained engineers, built-in sensors have now taken over their job – using systematically collected and recorded human know-how, which by rights ought to belong to the originator of that knowledge and, in many cases, would normally constitute an industrial secret. On the other side of the equation, online commerce has created a situation in which the open sharing and linking of data creates new assets. This is tantamount to a paradigm shift in the concept of data management:

  • Data have become a commodity that can be freely mined and traded, but also be manipulated or stolen.
  • Many market players share jointly generated data instead of keeping it to themselves.
  • Data ownership and usage rights are often undetermined, and can only be clarified after long negotiations between the parties concerned.
  • Platforms combine data from many different sources, and may include personal or other data that can be traced back to an individual person or group of persons.

Data are a valuable asset for companies because they increasingly serve as the basis for new business models, but at the same time represent a new and increasing cost factor due to the elevated consumption of resources, such as the energy required to supply the processing power for the high-performance servers and databases so essential in this age of big data. So what is the best way of attributing a monetary value to data? The Fraunhofer Institute for Software and Systems Engineering ISST in Dortmund has spent many years looking into this question.

Companies can measure the value of their data assets on the basis of:

  • The cost of registering ownership rights and updating proprietary databases.
  • The value in use of the data. Predictive maintenance is a pertinent example because it results in cost savings by detecting faults before they become acute and thereby avoiding downtime.
  • Market value. One way to determine this is by offering user data for sale and seeing how much buyers are willing to pay.

But what are the implications in the case of personal data, which are subject to strict privacy laws in Germany and other countries? How can a monetary value be placed on medical records held by hospitals and doctors’ surgeries, when people’s health is at stake? What special rules need to be applied in respect of official data gathered in the context of public administration processes? Can data concerning critical infrastructures such as the power grid or the Internet be treated as a saleable commodity without running the risk of triggering the collapse of entire market sectors or society as we know it? As the long list of questions shows, many different factors and legal issues can influence the value of data and the conditions under which data may or may not be exploited. It’s also sometimes necessary to look at certain issues from an international point of view, because differing interests may need to be reconciled.

Digital treasure: Find it, keep it safe, exploit it

Since the advent of Industrie 4.0, many industrial companies have adopted a data lake strategy, i.e. digging up all the data they can find within the company, tipping it into a vast repository or “lake,” and only afterward deciding what to do with it. But this type of strategy doesn’t necessarily produce the desired results because it leaves the field open for outsiders to step in and make money from data they didn’t even produce themselves. It’s easy to guess who these outsiders might be: first in line are the giant U.S.-based IT corporations, especially Google and Amazon. They are busily expanding into other sectors, backed by their expertise in the smart processing of large quantities of information. Google, for example, exploits freely available data and is avidly buying up companies that generate data. The best-known example is Nest, a manufacturer of smart heating control systems. Google’s interest doesn’t lie in the hardware but in gaining access to the data transmitted to and from the heating systems and the users’ smartphones, and other smart-home applications in the future, in order to sell digital services. All that without having to pay for the data on which the services are based.

Internet of Things (IoT)

German industry has similar opportunities for exploiting the considerable amount of data generated in use, for example, by German-made machines and vehicles, as well as data derived from communication with customers. New Internet of Things (IoT) technologies represent an abundant source of data. One example is the MIOTY® wireless IoT platform based on low power wide area (LPWA) technology developed by the Fraunhofer Institute for Integrated Circuits IIS, which enables a single receiver to download millions of data items collected by environmental sensors, machine control sensors, or traffic and in-vehicle sensors within a range of several kilometers. It is a very energy- and cost-efficient system. The sensor and location data can then be made available for new IoT applications after being processed using advanced data analytics methods developed by Fraunhofer IIS, including machine learning.

Adding value by means of new business models

So far, companies have been slow to grasp these opportunities. Fraunhofer has made it its mission to help them get more out of their unexploited data assets. Many institutes follow a traditional approach focusing on the technical aspects, where there is undoubtedly still a lot to do. But companies are increasingly looking for support in other areas such as data analysis, data protection, and the design of digital business models.

This is where Fraunhofer IIS and the associated Fraunhofer Center for Applied Research on Supply Chain Services SCS in Nürnberg can offer their expertise in service development and business model design. The biggest problem for companies is not a lack of data but that of knowing how much of their existing data is genuinely useful and what additional data they need to collect to fill in the gaps in their knowledge. In this digital era, one possibility is to make use of microelectronic devices integrated into physical objects, such as the containers used in production-line parts delivery systems, to create cyber-physical systems (CPS). A CPS is a distributed network of uniquely identifiable, interconnected embedded systems capable of communicating in real time. Production processes can be configured to take a CPS’s position into account. The CPS uses sensors to monitor the environmental parameters of physical processes. Each CPS processes its own data, enabling it to manage or regulate process parameters and exchange data with third parties. This results in a wealth of data that can be added to the company’s existing databases and used in other contexts.

The Fraunhofer Center for Applied Research on Supply Chain Services SCS has been developing data-based services and business models for companies for many years now. This not only requires an evaluation of the company’s own requirements with respect to its internal processes but also means identifying the customer’s needs so that they can be incorporated into new services and business models. In its many projects completed to date, the Fraunhofer Center has helped design end-to-end processes covering the collection, analysis and processing of relevant data and translated the results into business models enabling these data to be exploited in digital business models.

To specifically address the diverse business management issues associated with the design and implementation of business models in the digital world, Fraunhofer IIS set up the Research Center for Business Models in the Digital World ( in 2014. It is operated on behalf of Fraunhofer IIS by the Fraunhofer SCS research group in cooperation with the University of Bamberg. This Research Center proposes a six-stage, structured transformation process as a means of providing support to companies wishing to develop digital business models. Top priority is given to the question of how to make the best use of data, starting with the definition of the company’s own potential and followed by an analysis of the status quo, before finally arriving at a suitable scenario. Concrete use cases are studied with a view to creating the right transformation strategy for each company. These strategies can then be implemented and tested using Fraunhofer IIS’s own infrastructure, for example in the L.I.N.K. Test and Application Center for positioning, identification, navigation and communication technologies. An alternative is to make use of the open lab facilities offered by the JOSEPHS® service manufactory, a bricks-and-mortar storefront located in the heart of downtown Nürnberg where companies can invite potential users of new products and services to participate directly in their development and testing. But whatever the chosen route, it is essential that the transformed, data-driven business model generates added value for the end user. That is why Fraunhofer SCS places great importance on structured risk analysis and management processes that support data-driven business models.

Digital Business Engineering

Another institute working on methods for developing digital business models is Fraunhofer ISST. Its Digital Business Engineering method provides a step-by-step approach, not unlike a food recipe, that has already been tested with success in a number of use cases. The next step envisioned by the researchers is to roll out this method in many more applications and to develop a toolbox that companies can use to produce their own digital business model according to a proven set of basic principles. There is a need for fundamental guidelines because much of this territory is relatively unexplored. Nonetheless it will be a while before major companies start to include data assets in their balance sheet alongside the more traditional items that form part of their net asset position. And maybe one day budding entrepreneurs will be able to go their bank and obtain a startup loan using their data assets as security. Anyone who tries that today has little chance of a sympathetic response on the part of the bank manager. But that’s bound to change eventually.

The goal: Digital sovereignty

One important prerequisite for such scenarios is being able to prove who owns the data. This is fairly straightforward in the case of industrial applications such as predictive maintenance. The information generated by a machine in the course of its operation belongs to the user of that machine. But this doesn’t prevent the company that manufactured the machine from using the same data to develop algorithms for predictive maintenance and improve future versions of the product. Sales contracts normally exclude any other forms of commercial utilization.

Such data management puzzles place data owners in a quandary. On the one hand, their data increase in value the more they are utilized and shared, but this also augments the need for protective measures. In essence, this means it would be better to keep proprietary data under lock and key. Finding a way out of this dilemma is one of the biggest challenges of digital business. What companies want is the freedom to distribute their data without ever losing control over this valuable resource. The term commonly used to describe this issue is digital sovereignty.

Here’s an example: When online marketplaces for the procurement of parts and components first appeared at the end of the 1990s, suppliers willingly put their product catalogs online because they thought it would bring them more business. But when customers such as car manufacturers demanded that they should add customer-specific price information to the electronic data, the suppliers refused to do so. They didn’t want this information to be stored on a platform operated by a third party – in other words, they didn’t want to lose sovereignty over these particular data. Many other examples like this illustrate the point that some types of data are more sensitive than others and necessitate more protection.

Fraunhofer researchers are working on various solutions that allow data to be packaged, protected and enriched in a way that meets the requirement of preserving sovereignty. Their ideas include:

  • Providing on-demand data sharing, which allows users to opt out of data lake strategies requiring all data to be stored in a central repository,
  • Developing connectors within the Industrial Data Space (see below) as a means of organizing conditional data sharing,
  • Tagging data with restriction notes that specify the (limited) conditions under which their use is authorized,
  • Using distributed database technologies like blockchain to record data transactions,
  • Using standard vocabularies to improve interoperability and portability between different cloud platforms,
  • Developing transparency functions that notify users of transactions involving their data, giving them self-determined control over these data.

Digital sovereignty isn’t a permanent state. It should be imagined as a set of scales in which both sides have to be brought into balance. One pan contains the weight representing the data owner’s interests, namely privacy and security, and the value attributed to the data. The other pan contains the counterweight representing the user’s interests, which means the right to utilize, transfer, and share those data. Each case is balanced differently but the common denominator of all above-proposed solutions is that they allow the data owner to have the last word in any conflicts of interest. This is one of the keys to success in the data economy.

The solution: The Industrial Data Space

Until now, the missing link in the digitalization process was the availability of a technology that enables companies to develop a business model that permits data sharing without loss of sovereignty. It is up to the owner to decide whether data come with a price tag or are distributed for free. Data owners need reassurance that their sovereign rights are inviolable and that nobody can gain access to these data under a false identity. In short, this technology serves as the custodian of digital sovereignty for all parties concerned, while still leaving scope for creative partnerships.

It was precisely this line of thought that led to the creation of the Industrial Data Space. No less than twelve Fraunhofer Institutes are involved in the development of this secure data platform for German industry ( The collaborative research project was launched in October 2015 and is funded by the German Federal Ministry of Education and Research (BMBF). Its aim is to create a protected data space in which companies can share information using standardized, secure interfaces while at the same time retaining full sovereignty over their proprietary data. In parallel with the research project, a non-profit association was founded to represent the interests of Industrial Data Space users. It currently has more than 42 members from trade and industry, including thyssenkrupp, Bayer, Allianz and Rewe. Fraunhofer is also a member of this association. The association’s work includes defining a reference architecture model and piloting it in use cases.

The intelligent data infrastructure for business

The reason why so many companies have joined the association so soon after its creation is because the Industrial Data Space concept gives them confidence that the issue of data sovereignty ranks high in its order of priorities. German companies don’t want to run the risk of having their scope of action confined by dependency on big cloud service providers like Amazon and Google. Hence their insistence on the highest standards of control for essential services and data.

The software entities via which companies communicate with the Industrial Data Space are called connectors. They serve as an interface between the participating companies and support a wide range of protocols used to provide web services and connect with the Internet of Things. The connector correlates the company’s data and distributes it within the Industrial Data Space at the company’s request. In the opposite direction, the connector forwards inquiries from the Industrial Space to the company’s own systems. Connectors can also run various apps to filter or transform data and provide input for business processes. The Fraunhofer researchers working on the project intend to develop prototype connectors and apps for the reference architecture and demonstrate their use in typical use cases.

The industrial group thyssenkrupp AG has implemented a logistics pilot project in which a connector is linked on the internal side to the company’s freight entrance and loading dock management system. By consulting a chart, the controller can see immediately whether a particular truck carrying steel parts is running late or early, and adjust the schedule accordingly. The link to the real world of logistics is provided by the connector, which communicates via the Industrial Data Space either directly with the relevant apps on the truck driver’s smartphone or with the transport management system operated by the logistics service provider. The objective of this project is to save time and costs by optimizing the time window for loading and unloading.

The challenge: Data security

Despite all the talk about the value of data, data security, and sovereign ownership of data, it should not be forgotten that digitalization continues to be a major technological challenge. Numerous Fraunhofer Institutes are working closely together with industry to develop new solutions. One of them is the Fraunhofer Institute for Secure Information Technology SIT, which specializes in cybersecurity, i.e. the protection of data against theft and hacking. The Industrial Data Space is a good example showing how technologies like this go hand in hand with business models.

A similar approach could also be applied in the automotive context. Fraunhofer SIT, for example, is developing suitable data protection measures that can be integrated into vehicle control systems as early as the design stage. Without such measures, hackers would be able to disable the braking or other safety-critical systems in an autonomous car-sharing vehicle, posing a risk to life and limb of the vehicle’s occupants. Another type of risk arises from the analysis of vehicle data, which allows the owner’s movements to be tracked on the basis of journey or driving behavior profiles. Sometimes seemingly insignificant data can have unexpected consequences. A few vehicle insurers already offer special tariffs that reward good driving behavior. This requires the use of an in-vehicle data recorder that generates a speed profile. However, the evaluated speed data can also be used to create a profile of the vehicle’s movements and hence track the owner’s whereabouts. Fraunhofer SIT is therefore working on more transparent solutions that provide greater data protection by enabling users to see what use is being made of “their” data by third parties and give them greater control over the attribution of data access rights.

A crucial aspect of cybersecurity is making sure that systems are free of vulnerabilities. Fraunhofer SIT operates a test laboratory for the security analysis of IT systems ranging from discrete embedded systems to more complex systems such as a vehicle’s entire onboard network architecture, and even IT infrastructures and cloud services.

Smart innovations need 5G

Needs-based data collection requires a flexible transmission technology. The next-generation 5G wireless communication system will be one of the fundamental elements. The aim of 5G is to define a universal transmission standard for heterogeneous vertical markets such as Industrie 4.0, the automotive sector, the Internet of Things, and the smart grid. Key aspects of the 5G specifications include extremely low latency (less than one millisecond), high data rates, low power consumption, and the convergence of different network technologies. Researchers at the Fraunhofer Institutes for Integrated Circuits IIS, for Telecommunications, Heinrich-Hertz-Institut, HHI, and for Open Communication Systems FOKUS are working on appropriate solutions.

The new standard will be more than just another new generation of mobile technology. The aim is to create a universal wireless standard to serve as the basis for future-shaping topics such as the Internet of Things (IoT) and autonomous driving, and as a means of providing an advanced infrastructure for the development of telemedicine in the healthcare sector, smart living solutions for people with disabilities, and smart grid functions to support the energy transition.

This raises the question of where to store and process the data for these applications. In the cloud? That’s fine for companies that want to collect as much data as possible in a “data lake” structure, but quite the opposite of what is needed for Industrie 4.0, which is based on the concept of decentralized, largely autonomous, cyber-physical systems. Edge computing, in which data are filtered and processed by smart terminals on the edge of the network, is a better option in this case. But this option has the drawback of requiring the synchronization of data from many different sources in order to produce exploitable knowledge. The real answer probably lies halfway between the two. The Fraunhofer Institutes are therefore looking into hybrid concepts that combine the best of both approaches.


As the above examples show, digitalization is a complex subject that involves technological, organizational, economic and social changes. This new industrial revolution has the capacity to mark and change our lives like never before, and calls for the active participation of the business world. The Fraunhofer-Gesellschaft’s holistic and application-oriented approach to research makes it the ideal partner in the quest to shape these developments to their best effect, and create benefits for industry, society and end users alike.

Prof. Dr.-Ing. Albert Heuberger, director of the Fraunhofer Institute for Integrated Circuits IIS in Erlangen.

Prof. Dr. Boris Otto, director of the Fraunhofer Institute for Software and Systems Engineering ISST in Dortmund.

Prof. Dr. Michael Waidner, director of the Fraunhofer Institute for Secure Information Technology SIT in Darmstadt.