MachineDataMarketplace.NRW
How a data economy can transform the industry in NRW
Co-Authors: T. Bergs, Johannes Mayer, A Beckers,
Dieser Artikel ist auch auf Deutsch verfügbar.
Preamble
This use case was created within the project Blockchain Reallabor für das Rheinische Revier, funded by the Ministry of Economics, Innovation, Digitisation and Energy of the State of North Rhine-Westphalia, with the aim of forming a thematic community of interests. If you feel addressed, please contact us (mail@senseering.de) or contact the project directly via kontakt_realllabor@fit.fraunhofer.de or https://blockchain-reallabor.de/. The original publication by the Blockchain Reallabor can be found here (german).
Problem
Information from data and models is regarded as a driver of digital progress. The trend of industrial companies to extract monetary value from unique data they own or collect is considered necessary for survival in competition because other companies’ data is needed to optimize their own production and realize network effects [1].
The installed sensors on production machines collect large amounts of data during each production step. Currently, however, only the operators of such machines generate added value from the collected sensor data. Elementary production data (raw data) can be condensed into smart data with the help of artificial intelligence technologies. This data can be used to illustrate relationships regarding materials, machine states, intermediate products, qualities, consumables and environmental parameters, among other things. However, a robust prediction for process optimization using data-driven models requires data of unforeseeable and unexpected machine failures [2]. Such data is usually lacking, especially for small businesses. New data sets are significant even if the process framework conditions (quality characteristics such as processing speed or edge infeed) change, as the standard data are not sufficient to optimize the sequence and control of the processing [3]. In such cases, manufacturers or suppliers of materials, tools or machines try to generate data by means of test runs and attempts to improve their own products or detect errors. Such data usually originates from a test environment far away from regular operation.
B2B data trading usually fails because of the existing mistrust regarding the origin, integrity, quality and validity of production data as well as the underlying intention of the company [4]. Network effects between companies, such as the improvement of their own products, machines, processes or services through information from the upstream or downstream supply chain or through acquired technology data, are therefore not taken into account due to the lack of data exchange. If they were to be exchanged, data analysis and subsequent pattern derivation for learning effects for process optimization would require the expertise in the field of “Artificial Intelligence (AI)” of so-called data scientists. However, these specialists are rare [5].
Solution approach
The blockchain solution enables all market players (manufacturing companies, suppliers and data scientists) to make the cross-company handling of data goods transparent and comprehensible in order to make data usable in the sense of an economic good. In contrast to edge-based data storage options, which are managed centrally by one entity, blockchain technology is a geographically distributed, practically forgery and manipulation-proof database.
From the documentation of data exploitation, a data economy develops in which data function as digital goods. Process/product data or already existing technological models are transferred encrypted into the blockchain platform and can be used at any time against payment (e.g. pay-per-use or subscription models) for the optimization of the own production or for AI-based pattern recognition, learning, conclusions and derived self-corrections (training effect). Through the Data Scientists, the risks of worthless data and the binding of important capacities can be outsourced through data analysis. Confidence in the traded data is generated both through the gapless, traceable and unalterable storage of transactions including time and location stamps and through the comparison process within the network, in which selected participants confirm the correctness of transactions. The guarantee of data sovereignty is provided by the decentralized nature of the marketplace. All sensitive information remain with the data provider for the time being. Only after a successful purchase is the data transferred to the buyer (peer-to-peer). The data marketplace only has the data descriptions (metadata) and has no access to raw data at any time.
The MachineDataMarketplace.NRW offers a user-friendly platform that combines two rare components (data from outside the company, data scientists) to create added value for companies providing and analysing data. Own production and manufactured products can be raised to a new level (Smart Production) by linking own and purchased data. The potentials relate to plant optimization, product/process tracking, exchange of product properties and verification of the origin of products [2]. Data of the same machine type from different producers enables the improvement of maintenance measures in the sense of predictive maintenance by means of data-driven prediction models, so that machine uptime and product quality can be increased (plant optimization). The exchange of product- or process-relevant data, such as the location of products or the status of processes, enables an immediate reaction to errors. The digital representation of the process chain serves to clarify liability, predictive maintenance and increase efficiency by automating end-to-end processes (product/process tracking). The exchange of a digital twin (expansion of the database of a CAD model by the actual dimensions and qualities of a component) enables the individual synchronization and optimization of production processes. Continuous quality controls on receipt of goods become obsolete due to traceability (product properties). Data regarding the origin of selected goods serve as a basis of evidence in liability issues (verification of origin).
Based on this use case of the blockchain, new business models are created around tradable data products from smart data. For example, data generated in the course of a production process can now be transformed from “data garbage” into a valuable asset and profitably traded on the market. Data trading can be considered as a new source of income based on data that is often already available or easy to obtain. Workpieces and their properties (e.g. component geometries) are represented as digital twins and can be purchased. The marketplace enables direct information exchange, negotiations and automated payment processes between machines. It is open and easily accessible as well as fail-safe and tamper-proof.
Challenges
The foundation of this use case is, in addition to the existence of usable data, confidence in the mechanisms of data trading, including the integrity and validity of the traded data, as well as a possibility for the economic evaluation of the data on the market.
The availability of data for plant optimization requires a minimum of sensor equipment [2]. Interoperability of this measurement equipment is important to ensure reliable data acquisition. However, the mere existence of data is not sufficient. Their unadulterated initial state without assigned context makes it impossible to assign values, since transforming the data into information using analytical models requires a context such as the reference to the analysis period. Data must first be cleansed before trading. The situation is made more difficult by the fact that there is often a lack of trust in the data in the industrial environment. One possible approach to solving this problem is to provide company-owned data free of charge. Participation in the monetization added value is achieved through a participation model depending on the potential of the data.
The complete mapping of a digital twin requires intensive cooperation between the various partners in the value chain. Given the large number of stakeholders in the machine data marketplace described above, it must be taken into account how decisions are made between the different actors in the case of different types of blockchain technologies.
A further challenge in a network of different partners is the risk of wrong data sets to achieve competitive advantages [6]. The provision of incorrect data prevents an undisturbed production flow and leads to incorrectly set processes (e.g.: process forces). At present, companies are obliged to check the available information occasionally. However, this is not always possible without destroying the materials. The danger for the producer is the choice of a wrong product development strategy, caused by manipulated customer information.
Customers and employees are exposed to the risk of tracking and monitoring based on the information provided. Without anonymization or aggregation of data, entities along the supply chain can identify potential customers, view their personal preferences/characteristics and track employee actions. On the basis of the Data Protection Act (DSGVO), special requirements must be placed on the platform to protect personal data.
Stakeholder
There are six groups of potential stakeholders for this use case: data scientists, manufacturers, suppliers, collaborators (end point of the supply chain, interorganisational data exchange), customers and maintenance service providers. Regardless of the way they work together, the entities exchange different data with different purposes and security risks [6]. When exchanging data between producer, tool supplier, maintenance provider and collaborator, it must be noted that no comprehensive and direct production data is exchanged. The loss of intellectual property through inferences about the ongoing process and the benefits of the manufactured products can be caused by readable data access and is particularly critical in dynamic times in which partnership relationships are short-lived. For example, suppliers and maintenance personnel could possibly perform services for competitors as service providers and have an incentive to share data. The data streams to the customer contain little sensitive information. They serve to satisfy customer expectations and enable targeted support services.
The Industrial IOTA Lab Aachen opened a data marketplace for the process data of a fineblanking machine, which is operated by the Machine Tool Laboratory of the RWTH Aachen University.
The software development company senseering GmbH (headquartered in Cologne) offers a DLT data platform MyDataEconomy, which enables free data trading and edge computing. Data is traded and streamed without violating the data sovereignty of the owners or abusing the information. The data platform DatenAtlas enables this through the key technologies 5G, IOTA and Cloud-/Edge-AI.
Sources
[2] World Economic Forum in collaboration with Boston Consulting Group: Share to gain: Unlocking Data Value in Manufacturing; 2020
[3] https://www.it-production.com/hardware-und-infrastruktur/forschungsprojekt-iuno/
[4] Fraunhofer-Institut für Software- und Systemtechnik ISST: Datenmarktplätze — Plattfor-men für Datenaustausch und Datenmonetarisierung in der Data Economy
[5] https://www.vdi-nachrichten.com/karriere/datenanalysten-sind-rar/
[6] Pannekamp, J., Henze, M., Schmidt, S., Niemietz, P., Fey, M, Trauth, D., Bergs, T., Brecher, C., Wehrle, K.: Dataflow Challenges in an Internet of Production: A Security and Privacy Perspektive; 2019; Association for Computing Machinery
You have questions? Please contact us!
senseering GmbH
The senseering GmbH is a company founded in September 2018 that was awarded the RWTH Aachen University Spin-Off-Award. The core competence of senseering GmbH is the development and implementation of systems for the digitalization and networking of industrial and production facilities. Likewise, senseering GmbH advises on strategic corporate issues, in particular digital transformation, distributed-leger technologies, edge vs. cloud computing architectures for AI-based real-time control of industrial processes, digital business model innovation and the introduction of digital business processes such as home office, Azure or Microsoft365. Senseering is one of the winners of the first and largest AI innovation competition of the BMWi with the project www.spaicer.de.
Daniel Trauth (CEO) | www.senseering.de | E-Mail: mail@senseering.de
