Steering Representations—Towards a Critical Understanding of Digital Twins

Digital Twins are conceptualised in the academic technical discourse as real-time realistic digital representations of physical entities. Originating from product engineering, the Digital Twin quickly advanced into other fields, including the life sciences and earth sciences. Digital Twins are seen by the tech sector as the new promising tool for efficiency and optimisation, while governmental agencies see it as a fruitful means for improving decision-making to meet sustainability goals. A striking example of the latter is the European Commission who wishes to delegate a significant role to Digital Twins in addressing climate change and supporting Green Deal policy. As Digital Twins give rise to high expectations, ambitions, and are being entrusted important societal roles, it is crucial to critically reflect on the nature of Digital Twins. In this article, we therefore philosophically reflect on Digital Twins by critically analysing dominant conceptualisations, the assumptions underlying them, and their normative implications. We dissect the concept and argue that a Digital Twin does not merely fulfil the role of being a representation, but is in fact a steering technique used to direct a physical entity towards certain goals by means of multiple representations. Currently, this steering seems mainly fuelled by a reductionist approach focused on efficiency and optimisation. However, this is not the only direction from which a Digital Twin can be thought and, consequently, designed and deployed. We therefore set an agenda based on a critical understanding of Digital Twins that helps to draw out their beneficial potential, while addressing their potential issues.

Machine Learning in Chemical Product Engineering: The State of the Art and a Guide for Newcomers

Chemical Product Engineering (CPE) is marked by numerous challenges, such as the complexity of the properties–structure–ingredients–process relationship of the different products and the necessity to discover and develop constantly and quickly new molecules and materials with tailor-made properties. In recent years, artificial intelligence (AI) and machine learning (ML) methods have gained increasing attention due to their performance in tackling particularly complex problems in various areas, such as computer vision and natural language processing. As such, they present a specific interest in addressing the complex challenges of CPE. This article provides an updated review of the state of the art regarding the implementation of ML techniques in different types of CPE problems with a particular focus on four specific domains, namely the design and discovery of new molecules and materials, the modeling of processes, the prediction of chemical reactions/retrosynthesis and the support for sensorial analysis. This review is further completed by general guidelines for the selection of an appropriate ML technique given the characteristics of each problem and by a critical discussion of several key issues associated with the development of ML modeling approaches. Accordingly, this paper may serve both the experienced researcher in the field as well as the newcomer.

An Approach for the Transfer of Real Surfaces in Finite Element Simulations

Virtual simulations are a relevant element in product engineering processes and facilitate engineers to test different concepts during early phases of the development. However, in tribological product engineering, simulations are hardly used because input data such as material behavior are often missing. Besides the material behavior, the surface roughness of the contacting elements is relevant for tribological systems. To expand the capabilities of the virtual engineering of tribological components such as bearings or brakes, the hereby presented approach allows for the depiction of real rough surfaces in finite element simulations. Rough surfaces are scanned by a white-light interferometer (WLI) and further processed by removing the outliers and replacing non-measured samples. Next, a spline generation creates a solid body, which is imported to CAD software and afterwards meshed with triangle and quadrilateral elements in different sizes. The results comprise the evaluation of six differently manufactured (turned, coated, and pressed) real surfaces. The surfaces are compared by the deviations of the roughness values after measuring with the WLI and after meshing them. Furthermore, the elements’ aspect ratios and skewness describe the mesh quality. The results show that the transfer is dependent upon deep cliffs and large Sz values in comparison to the lateral expansion.

Method for localisation of sound sources and aggregation to an acoustic centre

Preliminary work by the IPEK - Institute of Product Engineering at KIT has shown that the simulated pass-by measurement for exterior noise homologation of vehicles has relevant optimization potential: the measurement can be carried out in smaller halls and with a smaller measurement setup than required by the norm and thus with less construction cost and effort. A prerequisite for this however is the scaling of the entire setup. For the scaling in turn, the sound sources of the vehicle must be combined to a single point sound source - the acoustic centre. Previous approaches for conventional drives assume a static centre in the front part of the vehicle. For complex drive topologies, e.g. hybrid drives, and unsteady driving conditions, however, this assumption is not valid anymore. Therefore, with the help of an acoustic camera, a method for localizing the dominant sound sources of the vehicle and a software-based application for summarizing them to an acoustic centre were developed. The method is able to take into account stationary, unsteady and sudden events in the calculation of the acoustic centre, which is moved as a result. Using substitute sound sources and two vehicles, the method and the used measurement technology were examined and verified for their applicability.

Change impact analysis of complex product using an improved three-parameter interval grey relation model

Change impact evaluation of complex product plays an important role in controlling change cost and improving change efficiency of engineering change enterprises. In order to improve the accuracy of engineering change impact evaluation, this paper introduces three-parameter interval grey number to evaluate complex products according to the data characteristics. The linear combination of BWM and Gini coefficient method is used to improve the three-parameter interval grey number correlation model. It is applied to the impact evaluation of complex product engineering change. This paper firstly constructs a multi-stage complex network for complex product engineering change. Then the engineering change impact evaluation index system is determined. Finally, a case analysis was carried out with the permanent magnet synchronous centrifugal compressor in a large permanent magnet synchronous centrifugal unit to verify the effectiveness of the proposed method.

Computational intelligence in product design engineering: review and trends

Product design engineering is undergoing a transformation from informal and largely experience-based discipline to a science-based domain. Computational intelligence offers models and algorithms that can contribute greatly to design formalization and automation. This paper surveys computational intelligence concepts and approaches applicable to product design engineering. Taxonomy of the surveyed literature is presented according to the generally recognized areas in both product design engineering and computational intelligence. Some research issues that arise from the broad perspective presented in the paper have been signaled but not fully pursued. No survey of such a broad field can be complete, however, the material presented in the paper is a summary of state-of-the-art computational intelligence concepts and approaches in product design engineering. Keywords: Computational intelligence, engineering design, product engineering, decision making, design automation

Computational intelligence in product design engineering: review and trends

Product design engineering is undergoing a transformation from informal and largely experience-based discipline to a science-based domain. Computational intelligence offers models and algorithms that can contribute greatly to design formalization and automation. This paper surveys computational intelligence concepts and approaches applicable to product design engineering. Taxonomy of the surveyed literature is presented according to the generally recognized areas in both product design engineering and computational intelligence. Some research issues that arise from the broad perspective presented in the paper have been signaled but not fully pursued. No survey of such a broad field can be complete, however, the material presented in the paper is a summary of state-of-the-art computational intelligence concepts and approaches in product design engineering. Keywords: Computational intelligence, engineering design, product engineering, decision making, design automation