Fraunhofer Institute for Mechatronic Systems Design

Data-driven methods are increasingly utilized in metal forming processes for monitoring and quality optimization. An adapted modeling notation DSL4DPiFS for forming processes is presented to model hardware, software, and data flow aspects to support the design and analysis of data-driven methods. DSL4DPiFS enables metal forming and automation experts to model field-level information as data sources, and the data sinks for data analysis. The notation was adapted to the requirements of selected metal forming processes and evaluated in three case studies.

Große Sprachmodelle wie GPT-4 bieten erhebliche Potenziale für das Systems Engineering. Prompt-Engineering ermöglicht einen flexiblen Einsatz im Anforderungsmanagement, Systementwurf und in der Integration, Verifikation und Validierung ohne aufwendiges Modelltraining. Die Formulierung von Prompts und die Anwendung fortschrittlicher Techniken erfordern jedoch tiefes Domänenwissen. Der Beitrag zeigt Potenziale und Herausforderungen dieser Technik auf und illustriert praktische Anwendungsbeispiele

The increasing complexity of modern technical systems necessitates innovative approaches such as Model-Based Systems Engineering (MBSE). In this context, using Artificial Intelligence (AI) emerges as a key enabler for practical application and efficiency improvement. This article introduces a maturity model for AI-based assistance systems in MBSE. It helps companies assess their current automation level in MBSE activities, providing a foundation for strategic planning of process improvements.

Sustainability and digitalization represent two megatrends that profoundly impact companies, particularly those in manufacturing industries, shaping their competitive advantage and long-term market success. While there is a need for companies to converge both megatrends strategically, practitioners and scholars lack a comprehensive framework so far. This study grasps this urgent call by addressing the question of how sustainability and digitalization can be strategically integrated into the dual transformation of manufacturing corporations. Therefore, we adopt the Design Science Research (DSR) approach and collect data from a literature review and semi-structured interviews with expert practitioners. Grounded in a holistic understanding of sustainability, we derive a reference model for the dual transformation permeating all facets of business operations that includes the interdependence and strategic integration of sustainable digitalization and sustainability by digitalization. By leveraging this model, scholars and practitioners can navigate the complexities of dual transformation.

The security of software systems remains a critical sociotechnical challenge despite existing tools and processes. The articles in this special issue address aspects of security that go beyond code, offering ways to empower developers, provide trust and assurance, and address planning and regulation requirements.

In punch-bending, products such as brackets, electronic contacts or spring elements are produced from wire-shaped semi-finished products using separation processes and several successive forming processes. Within the multi-stage straightening and bending processes, cross-stage and quantity-dependent effects have a significant influence on the quality of the end product. In order to optimize the punch-bending process with regard to the resulting component deviations and waste rate, this article presents the concept of a digital twin for an innovative hybrid model of a multi-stage punch-bending process. To ensure efficient development and implementation of the digital twin, the graphical modeling notation DSL4DPiFS is used for additional support. It makes it possible to derive the required interfaces of the Asset Administration Shell of the hybrid data-driven model.

The rapid advancements in digital transformation have led to the emergence of dataspaces as a pivotal element for industry-wide and cross-industry data integration and interoperability across various businesses. Despite their potential, the adoption and effective utilization of dataspaces by business stakeholders remain challenging. This paper aims to address this gap by developing a comprehensive learning environment tailored for business stakeholders. The proposed environment includes a demonstrator and a training concept designed to enhance stakeholders' understanding and capabilities in managing and leveraging dataspaces. Through an interview study and an analysis of the current state of research, we identify problem fields and derive key requirements for the development of the learning environment. Our findings contribute to the body of knowledge by providing practical guidance through learning environments for the deployment of dataspaces in business contexts and highlighting areas for future research.

The inline measurement of process parameters describing the separation process of disk stack separators is expensive and complex. An alternative method for assessing the current process parameters employing more accessible data is required. In this work, the separation efficiency during the operation is estimated using vibration measurements by determining characteristic vibration patterns due to the increasing load of the separation bowl. The disk stack separator used is an industry-scale laboratory model of a disk stack separator, equipped with three accelerometers for vibration measurements. The influence of different bowl geometries and particle systems on the vibrationpatterns and the separation efficiency has been investigated. The gathered data and knowledge about the inlet conditions (volume flow and solids concentration) are used to create a soft-sensor for the separation efficiency. A deviation of approx. 2 % between direct measurements of the seperation efficiency and the results of the soft-sensor is reached.

Assistive wearable devices can significantly enhance the quality of life for individuals with movement impairments, aid the rehabilitation process, and augment movement abilities of healthy users. However, personalizing the assistance to individual preferences and needs remains a challenge. Brain-Computer Interface (BCI) offers a promising solution for this personalization problem. The overarching goal of this study is to investigate the feasibility of utilizing passive BCI technology to personalize the assistance provided by a knee exoskeleton. Participants performed seated knee flexion-extension tasks while wearing a one-degree-of-freedom knee exoskeleton with varying levels of applied force. Their brain activities were recorded throughout the movements using electroencephalography (EEG). EEG spectral bands from several brain regions were compared between the conditions with the lowest and highest exoskeleton forces to identify statistically significant changes. A Naive Bayes classifier was trained on these spectral features to distinguish between the two conditions. Statistical analysis revealed significant increases in δ and θ activity and decreases in α and β activity in the frontal, motor, and occipital cortices. These changes suggest heightened attention, concentration, and motor engagement when the task became more difficult. The trained Naive Bayes classifier achieved an average accuracy of approximately 72% in distinguishing between the two conditions. The outcomes of our study demonstrate the potential of passive BCI in personalizing assistance provided by wearable devices. Future research should further explore integrating passive BCI into assistive wearable devices to enhance user experience.

This paper deals with micromagnetic measurements for online detection of strain‐induced α′‐martensite during plastic deformation of metastable austenitic steel AISI 304L. The operating principles of the sensors are magnetic Barkhausen noise (MBN) and eddy currents (EC), which are suitable for detection of microstructure evolution due to formation of ferromagnetic phases. The focus of this study was put on the qualification of different micromagnetic techniques and different measurement systems under conditions similar to the real ones during production, which is crucial for implementation of a property‐controlled flow forming process. The investigation was carried out on tubular specimens produced by flow forming, which have different content of α′‐martensite. To characterize the sensitivity of the sensors, different contact conditions between sensors and workpieces were reproduced. MBN sensors are suitable for detecting amount of α′‐martensite, but the measurements are affected by the surface roughness. This entails that the calibration models for MBN sensors must take account of these effects. EC sensors show a closer match with the amount of α′‐martensite without having major affectation by other effects.

The digital twin in manufacturing, also called digital factory twin (DFT), is not only a recent technology but also a polymorphic concept which generates a profusion of state-of-the-art work from industry and research without real coherence. The work relates to very different realities including various definitions, shapes, models and applications. A wide variety of architectures are proposed and there is a significant lack of information on how to deploy a digital twin efficiently and how to define an appropriate architecture. The DFT is often implemented use-case-driven and in various ways concerning IT architectures. The paper proposes a reference architecture based on enterprise architecture modelling language ArchiMate and TOGAF principles, which allows to model business architectures and IT architectures. We make use of that and model a DFT business architecture based on the lifecycle of a factory and a DFT IT architecture based on IT applications and IT functions.

Product management (PM) spans the entire product lifecycle and requires effective navigation through evolving customer demands, volatile markets, and uncertain competitive landscapes. In the face of these challenges, industrial companies must adopt new strategies. Data-driven product management has emerged as a critical solution, enabling companies to transform vast amounts of data, such as customer feedback and usage data, into actionable insights for informed decision-making. However, a clear framework for identifying and categorizing relevant data sources is still lacking. This paper develops a comprehensive taxonomy to address this gap by systematically characterizing data types relevant to data-driven product management. Using the empirical-to-conceptual and conceptual-to-empirical approach based on NICKERSON ET AL., we identified and classified 47 distinct data types, grouped into 10 data clusters through cluster analysis. These clusters provide a structured and actionable overview of the data landscape, making it easier for organizations to operationalize data-driven strategies. We also analyzed typical software categories, such as ERP systems, where these data types are stored and managed. Our findings are critical for organizations seeking to improve their data-driven product management capabilities by providing a systematic framework for structuring data and an understanding of crucial software tools. This research fills a significant gap in academic literature and practical implementation, providing organizations with a clear path to use data in product management effectively.