
ASYS Automatisierungssysteme GmbH, Dornstadt
Specialist area: maintenance concepts; product-related proof of quality
The current standard practice of collecting and analysing data from production plants in the area of the Industrial Internet of Things (IIoT) allows companies to gather detailed information about their own processes and products. This information can then be used to optimise internal production.
The aim of the KOSMoS project, which is funded by the Federal Ministry of Research and Education, is to connect manufacturing companies with one another, thereby creating a secure, digital value network that transcends company boundaries. Within the consortium of nine project partners, inovex is the expert in Data Management and Analytics.
As users of data analyses and machine learning, KOSMoS gives industrial companies the opportunity to make their own production operations faster, more efficient, more flexible and thus more competitive. Machine learning is based on the recognition of patterns and uses recurring elements and relationships to generate new knowledge from empirical data and histories. By analysing the collected customer, log and sensor data, new products and services and data-based business models can be established, such as transparent maintenance concepts, dynamic leasing and quality certification of delivered products. Advantages may take the form of reduced and more effective maintenance work, a more efficient usage-based leasing rate or an optimised production process. inovex examines these analytical aspects in particular as part of the KOSMoS project.
The networking of machines, intelligent computers and other equipment in the industrial setting results in an ever greater amount of production-related data that can be collected, processed and made available. Besides the sensors and smart devices, standardised protocols such as OPC UA and MQTT are very important for ensuring that the exchange functions well between the various platforms. Although the generated data is often very highly frequent, highly dimensional and suited to the concept of big data, only a few data samples generally exist. Special methods and algorithms are needed in order to produce intelligent processes such as error monitoring and predictive maintenance from the collected data and thereby turn the vision of intelligent production into reality. inovex brings both its data engineering and algorithmics expertise to the KOSMoS project.
One of the core competencies of the IIoT is predictive maintenance. Predictive maintenance should be seen as an extension of the classic approach of performing maintenance at set intervals. Here the data collected by the IIoT is automatically analysed with corresponding data science and machine learning approaches before instructions and recommended courses of action are issued. The intelligent planning of maintenance tasks not only makes it possible for the company to prevent undesirable production downtime, but also operate the machines in their optimal state. Within the KOSMoS project, inovex supports the application partners by finding and implementing analytic solutions for their specific use cases.
A connected value network, in which cooperating firms share cross-company production and process data, can generate added value for every involved partner if there is sufficient trust. A blockchain is a data storage system that isn’t just managed by one company. Instead, it is decentralised, transparent and open to all business partners at the same time. As such, from a technical perspective, it is a network of distributed servers that are independent of one another. The linking of data sets and the agreement of all network nodes via one so-called consensus algorithm in relation to the correctness of statuses prevents the subsequent reversibility and manipulation of data.
Within the blockchain, so-called smart contracts can be used. Triggered by a data threshold value or another action, they initiate a process that automatically permits further processes within the value chain on the basis of manipulation-proof data that has been stored. In this way, contracts can be formed, processed and checked while cutting transaction costs and increasing the certainty of contracts.
Due to the large number of different components, particularly machine learning libraries and tools, various opinions exist even among the tech giants as to what represents the best approaches for pipelines in productive use. Within the context of the KOSMoS project, inovex is conducting research into architecture concepts that permit the secure and reliable operation of modular data processing and machine learning components not only in the cloud, but also ‘at the edge’ or ‘in the fog’ – in other words, much closer to the ‘shop floor’. At the same time, work is ongoing to establish where within the data management process the blockchain and smart contracts can be sensibly used.
Specialist area: maintenance concepts; product-related proof of quality
Specialist area: blockchain development; cross-company business models
Specialist area: data protection / security when collecting production data
Specialist area: process modelling; the linking of digital and physical components
Specialist area: edge solutions for communication between the shop floor and blockchain
Specialist area: cross-company, transparent maintenance concepts
Specialist area: audit-compliant models through dynamic leasing
The project work is being funded by the Federal Ministry of Research and Education as part of the research programme ‘Innovations for Production, Services and Work in the Future’ within the funding measures ‘Industry 4.0 – Collaborations in Dynamic Value Networks’ (lnKoWe) over a period of 36 months.
Adap works on the intersection of machine learning and distributed computing to build a new class of intelligent systems.
The non-profit organization wants to develop Stuttgart into a globally leading region for blockchain technology.
Flower takes a unified approach to driving new approaches in federated learning.
Get in touch!