I4S+ content for subscribers only

Companies, especially SMEs, face the challenge of introducing digital technologies efficiently and as smoothly as possible. Using the introduction of a digital assistance system in technical service as an example, this article shows which challenges and problem areas arise, how they can be overcome and which factors promote a successful introduction process. In the process it is worked out how SMEs with few resources can generate a high degree of participation and acceptance. The basis is a socio-technical understanding that takes a holistic view of the overall system of people, technology and organization in the introduction process of digital technologies.

Data plays an important role in our world - including production technology. Businesses are faced with rising customer demands and competitive pressure. Furthermore, the trend towards smaller batch sizes and increasing variant diversity requires quick reactivity and agility. In order to make the right decisions under these circumstances, data must be generated and analyzed to derive insights. AI technologies are suitable to address the growing uncertainty and complexity. In the following, methods are described that are vital to master data processes for high-quality AI technologies.

The increasing shortage of IT specialists requires lower-skilled employees to be empowered to perform tasks that previously required the involvement of experts. Industry 4.0’s emerging technologies for human-technology interaction and for the digital twin allow the design of intuitive user interfaces, system-independent communication interfaces, and user-specific assistance functionalities to meet this challenge. This paper presents a framework for configuring and monitoring of process flows for different production and logistics systems. By reviewing existing programming approaches, the paper derives requirements for the framework, describes its general architecture and the technical realization of the modular interaction interface. A prototypical implementation validates the presented concept on the example of a cellular conveyor system and a collaborative robot system.

Deep learning-based error detection methods outperform traditional methods because of the continuously increasing complexity of technical systems and inherent flexibility and scalability of Deep Learning techniques. This article introduces the KrakenBox – an autonomous Deep Learning-based error detector for industrial Cyber-Physical Systems (CPS). It exploits a lightweight, Long Short-Term Memory (LSTM) network capable of online error detection that can be deployed on an embedded platform such as NVIDIA Jetson AGX Xavier or even Google Coral Edge TPU. This article describes the architecture of the KrakenBox and demonstrates its application with two case studies.

In the project ARAS (Advanced Robot Assistance Solution) a robot programming assistant was developed, which allows for automated generation of robot programs for assembly processes. By using a multimodal approach for human-machine-interaction, assembly steps are recognized with machine learning algorithms while a worker is showing the robot how an assembly process is performed. Afterwards, a robot program is generated automatically. This way, new robot programs are created within minutes without the user having any knowledge about programming or robotics.

Einer der Erfolgsfaktoren für die Realisierung von Cloud-Manufacturing-Plattformen ist die Geschwindigkeit, mit der Angebote generiert werden können. Um aus dem Kundenauftrag möglichst automatisiert die Planungsprozesse der Plattform zu unterstützen extrahieren wir Bearbeitungsprozesse aus technischen Zeichnungen mittels eines künstlichen neuronalen Netzes. Dabei werden die Probleme der Multi-Label-Bildklassifikation und eines unsymmetrischen Datensatzes behandelt. Abschließend wird die Leistungsfähigkeit des Datensatzes an den Testdaten demonstriert und ein rekursives Verfahren zur Verbesserung des Systems im laufenden Betrieb beschrieben.

The digital transformation, especially the shift of maintenance to smart maintenance, contains a number of challenges. The management of data plays a significant role in this [1]. The use of sensor technology and devices for mobile data acquisition offers a variety of possibilities for quickly capturing large amounts of data, even in real time. In this context, however, there is also a need for tools that enable efficient data exchange and, at the same time, structured storage of large amounts of data. One of these tools is the digital lifecycle record, which has enormous potential for data-driven services thanks to its special features.

In recent years, the potential of cloud manufacturing in the B2B business fields has evolved enormously due to the increasing utilization of new digital technologies. Digital business processes confront manufacturing companies with new opportunities and challenges. For the digital transformation of these companies, the structured analysis of corresponding platform concepts is essential. In the first step, this can be supported by a framework that analyses basic insights and framework conditions for the development of dynamic cloud manufacturing platforms for the production sector. A corresponding framework is developed and presented in this article.

With a world population of ten billion people by the middle of the 21st century, natural resources must be used sustainably to prevent environmental disasters and wars. Traditional concepts such as recycling alone are no longer sufficient. Instead, we must think in terms of material cycles and transform the traditional production economy into a circular economy. To achieve this, a close link between production and recycling must be established, which is hardly conceivable without digitalization. This article begins with an overview of steps of industrial development towards more sustainability, finally showing how the digital transformation can facilitate the realization of a circular economy.

To be able to use artificial intelligence, a company does not necessarily need a qualified specialist. The Fraunhofer study “Cloud-based AI Platforms - Opportunities and Limits of Services for Machine Learning as a Service” shows how small and medium-sized companies can proceed instead. This article summarizes the arguments and results of the study, citations from it are therefore not marked accordingly.