Integration concept of Digital Shadows and BIM for a long-term optimal decision support in factory planning

Industrial objects nowadays rapidly transform due to the development of digital technologies. The concept of the Factory of the Future (FoF) involves digitization of all parts of the factory. In this paper two technologies are motivated and considered as the basic technologies that should be used in FoF: Digital Shadow (DS) and Building Information Modelling (BIM). Basic theory on these issues is given and potentials of BIM and DS integration is formulated. Based on the ability of digital technologies, their integration and convergence to generate value, definition Digital Asset is introduced from the economic point of view as a digital resource which brings economic benefit. A concept of integrating BIM and DS technologies for decision support in factory planning is formulated, including Life Cycle Assessment (LCA) and semantic modeling. The concept includes a description of the aggregate of technologies and their interconnections as a Digital Asset of the FoF. Further research objectives are focused on integration of BIM and DS which requires their interoperability ensured by an Ontology-Based Data Access (OBDA) approach, based on the Semantic web.

Achieving parametric transparency in model-based factory planning

The planning of new factories, as well as the re-planning of existing factories, has become more frequent due to increasingly changing business requirements, as for example shorter product life cycles and Industry 4.0. A higher number of involved planners and the resulting high amount of planning information strongly require coordination. In this context, the importance of Building Information Modeling (BIM) in factory planning rises as it provides a method of integrated building planning and planning validation by means of 3D software and object-oriented modelling. However, despite the use of BIM, there are still major interface problems in factory planning that cannot be solved by the still manual plausibility checks of non-geometrical planning information. To enable automatic checking of planning results, thereby improving the BIM-based factory planning process, machine-readable explication of the parametric dependencies are required between different planning fields such as production planning and building planning. The goal of this paper is to show parametric and thus non-geometric dependencies that exist between the sub-models of BIM-based factory planning in such a way that software agents can automatically evaluate this design information. Within the planning interface between production planning and building planning, the paper focusses on the particular exchange between the planning of the manufacturing system and the planning of a cutting fluid pump. With the involvement of domain experts from factory planning, systems engineering and production engineering, we as the authors have managed to develop a coherent system of block diagrams, constraint diagrams and parametric diagrams that explicate the focused interface in a machine-readable manner. We believe our accomplishments are an essential element for completely automated planning validation in BIM-based factory planning and general object-oriented modelling in the future.

Uncertainty Estimation in Deep Neural Networks for Point Cloud Segmentation in Factory Planning

The digital factory provides undoubtedly great potential for future production systems in terms of efficiency and effectivity. A key aspect on the way to realize the digital copy of a real factory is the understanding of complex indoor environments on the basis of three-dimensional (3D) data. In order to generate an accurate factory model including the major components, i.e., building parts, product assets, and process details, the 3D data that are collected during digitalization can be processed with advanced methods of deep learning. For instance, the semantic segmentation of a point cloud enables the identification of relevant objects within the environment. In this work, we propose a fully Bayesian and an approximate Bayesian neural network for point cloud segmentation. Both of the networks are used within a workflow in order to generate an environment model on the basis of raw point clouds. The Bayesian and approximate Bayesian networks allow us to analyse how different ways of estimating uncertainty in these networks improve segmentation results on raw point clouds. We achieve superior model performance for both, the Bayesian and the approximate Bayesian model compared to the frequentist one. This performance difference becomes even more striking when incorporating the networks’ uncertainty in their predictions. For evaluation, we use the scientific data set S3DIS as well as a data set, which was collected by the authors at a German automotive production plant. The methods proposed in this work lead to more accurate segmentation results and the incorporation of uncertainty information also makes this approach especially applicable to safety critical applications aside from our factory planning use case.

Kollaborative Fabrikplanung/Collaborative factory planning

Unternehmen des produzierenden Gewerbes sind mit einem zunehmend dynamischen Marktumfeld konfrontiert, weshalb fabrikplanerische Anpassungen immer schneller erfolgen müssen. Der größte Zeitanteil in Fabrikplanungsprojekten wird jedoch für die Datensuche, -aufbereitung und den anschließenden Informationsaustausch zwischen den unterschiedlichen Beteiligten aufgewendet. Um diesen Aufwand zu reduzieren, wird ein Konzept zur kollaborativen Fabrikplanung mit Building Information Modeling vorgestellt.   The production industry is facing increasingly dynamic market conditions. Therefore, the adaption of production systems must happen ever more quickly. However, data acquisition, processing and the subsequent exchange of information takes up a lot of time in factory planning projects. To reduce this effort, the paper introduces a method for collaborative factory planning based on building information modelling.