DigiTwin: An Approach for Production Process Optimization in a Built Environment

As a recently coined buzzword without a clear definition, Digital Twin relies on the high-fidelity digital representation of the physical product and the continuously accumulated data and real-time presentation of the collected data to simultaneously update and modify with its physical counterpart. In research and practice, many types and expressions of Digital Twin take place for plethora of use cases along the product lifecycle. The simulation of production processes using a Digital Twin can be utilised for prospective planning, analysis of existing systems or process-parallel monitoring. In all cases, the Digital Twin provides manufacturing companies various benefits for improvement in production and logistics processes leading to cost savings and higher flexibility. The generation of the Digital Twin in a built environment poses a huge challenge—in particular for small and medium-sized enterprises. For this market segment, appropriate concepts and solutions need to be elaborated and developed, including modern IT techniques such as Machine Learning. In this introductory chapter, the approach used to pave the way for this book is presented. The idea of Digital Twin, the origins, the goals and the expected audience of this book are roughly described. Finally, we give the first insight in the content of this book and the mutual interdependence of the chapters. This book explores the way to generate and commercialize the Digital Twin in manufacturing in order to provide a convincing offering as the outcome of a public-founded research project.

Production planning and control can be supported by Digital Factory methods to optimize production processes and workflows. A central component of the Digital Factory is simulation, which is used to represent real objects and processes in a virtual environment. This virtual environment is suitable for performing analyses and planning processes so that understanding about the real system can be gained. Accordingly, planning processes such as factory planning, investment planning, capacity planning, bottleneck analyses, inventory planning and internal material transport can benefit from simulation by gaining more valid and far-reaching insights. However, simulations must be designed for specific use cases in order to be able to process them. Therefore, the corresponding parameters for the use of the simulation must be determined. The approach presented here focuses on several use cases to create a framework that is not only valid for a single use case, but allows for an arbitrary application which is as comprehensive as possible. This leads to a Digital Twin, which, in turn, can handle several use cases and is not focused on one use case like the simulation. The following chapter deals with the mentioned applications, primarily focusing on the requirements of the use cases for the simulation framework by identifying and specifying the required parameters. Accordingly, a comprehensive list of parameters and their exact properties is presented to support production planning and control. With this understanding, an efficient determination of these parameters can be carried out in the further course of this book, from which the generation of a Digital Twin is enabled.

Over the last few years, a concept called Digital Twin has evolved rapidly as a new key approach in the field of Product Lifecycle Management (PLM). Briefly, a Digital Twin is a digital representation of an active unique product or unique product-service-system with its selected characteristics within dedicated lifecycle phases. This concept has experienced a tremendous impact by IoT technology, which has drastically reduced the costs. It builds the foundation not only for connected products and services but also for entirely new offerings and business models. Three main characteristics of Digital Twin were identified: representation of a physical system, bidirectional data exchange, and the connection along the entire lifecycle. For a better understanding, three subtypes of Digital Twin are presented, namely: The Digital Master, the Digital Manufacturing Twin, and the Digital Instance Twin which refer to the different phases of the product lifecycle: design, production and operation. Therefore, this chapter formulates a consistent and detailed definition of Digital Twins and gives insight in dedicated research direction. Finally, based on the Digital Twin characteristics, an approach for generation of Digital Twin in manufacturing is shown.

A three-dimensional digital representation, as the Digital Twin of machines and objects within production plants, is becoming increasingly important for efficient planning and documentation of production. During the acquisition of the images, the two areas of accuracy and detail are key. The accuracy describes the precision in which the digital twin represents the real objects. For example, how well the dimensions of a machine within a reconstructed model matches its counterpart in reality. Detail refers to the level of detail in the digital model. Should the door of a machine tool be distinguishable from the machine or not. Two basic technologies have been established for scanning. The photogrammetry method creates a dense point cloud based on images using a sophisticated toolchain. The laser method firstly measures an accurate model of the environment using a laser scanner. In a second step, the color information is assigned to the measured points via the evaluation of photos. For certain applications, models resulting in highest accuracy and detail are not necessary; in these cases, speed and practicability of the acquisition is the primary concern. Due to recent advancements photogrammetry these methods are best suited in the case of generating a digital twin for production facilities. An overview of the methods available as well es the underlying principles is presented. Practical considerations and examples show the feasibility and results of different photogrammetry approaches, resulting in the presentation of a photogrammetric system well suited to the task of creating a digital twin of a production facility.

Artificial intelligence (AI) is one of the core drivers of industrial development and a critical factor in promoting the integration of emerging technologies, in the new generation of big data and Industry 4.0. In particular, machine learning (ML) becomes increasingly more frequently applicable in manufacturing applications. This chapter presents a systematic overview of today’s applications of ML techniques and approaches for their usage in the manufacturing environment. The utilization of ML methods is related to manufacturing process planning and control, predictive maintenance, quality control, in situ process control and optimization, logistics, robotics, assistance and learning systems for shopfloor employees. Exhaustive fundamental and problem concept describes how to select an appropriate ML approach including the combination of multiple approaches. Supervised methods dominate the state of the art with reinforcement learning methods gaining more attention in recent years. Subsequently, the gains of ML such as derivation of model configuration based on the data, generation of behavioral models through training, easy validation of the model and optimization of the real system based on the model are illustrated. This illustration is surrounded by two use cases from plant engineering. Data analysis points out how these concepts can be analyzed and with behavior model can be achieved regarding the implemented ML method. Finally, conclusions and outlook reflect the benefits and drawbacks of ML and draw the way for future development.

Recognition of an object from a point cloud, image or video is an important task in computer vision which plays a crucial role in many real-world applications. The challenges involved in object recognition, aiming at locating object instances from a large number of predefined categories in collections (images, video or, model library), are multi-model, multi-pose, complicated background, occlusion, and depth variations. In the past few years numerous methods were developed to tackle these challenges and have reported remarkable progress for 3D objects. However, suitable methods of object recognition are needed to achieve added value in built environment. Suitable acquisition methods are also necessary to compensate the impact of darkness, dirt, and occlusion. This chapter provides a comprehensive overview of the recent advances in 3D object recognition of indoor objects using Convolutional Neural Networks (CNN). Methodology for object recognition, approaches for point cloud generation, and test bases are presented. The comparison of main recognition methods based on methods of geometric shape descriptor and supervised learning and their strengths and weakness are also included. The focus lies on the specific requirements and constrains in an industrial environment like tight assembly, light, dirt, occlusion, or incomplete data sets. Finally, a recommendation for use of existing CNN framework for implementation of an automatic object recognition procedure is given.

The challenge of enhancing and generalizing interoperability as an important pre-requisite for Digital Twin is often hindered by the fact that data quality requisites depend on the purpose for which the data will be used and on the subjectivity of the data consumer. Data quality is getting as important as product quality in manufacturing process. In this chapter we present how to systematically handle the data quality requirements and support a comprehensive data quality management. After the definition of Digital Thread as a data highway, the classification of data quality is discussed based on data quality dimensions and standards related to field of manufacturing. The data quality metrics is discovered upon the guidelines developed by national and international harmonization bodies from global automotive industry. Three practical examples from design and manufacturing as well as data migration in industrial context give insight in practical challenges and achievements in the field of data quality as well as future directives. The discussion section emphasizes the high importance of data quality for the generation of Digital Twin.

In order to prepare the point cloud for a feature-based process of planning and simulation, the objects it contains must first be recognized and a CAD-based structure must be created. The great challenge in every process chain is to implement the desired function with minimal effort and losses. As one of the alternatives, the object recognition is selected to implement the layout planning of a factory in 3D. In this chapter, it is demonstrated how to embed the object recognition in the process of virtual 3D layout planning in a built environment as well as which findings and results can be expected. The aim of this chapter is to investigate and evaluate the usefulness of a realistic 3D virtual factory model in factory layout planning primarily for Digital Twin based on object recognition. This is addressed by a practical study of how existing methods for data acquisition and processing can be concatenated and, subsequently, applied under real industrial constraints and conditions. During this study, realistic 3D layout models are created using point clouds acquired by commercial terrestrial laser scanner and prepared for object recognition with Convolutional Neural Networks considering the strict data quality requirements. The selection of models was discussed and the results were evaluated in industrial workshops with engineers involved in the layout planning and machine operators that will work within the production system. Seamless, robust, (semi-)automatic workflow of primarily standard, modular components with low user assistance was of particular interest. This chapter is concluded with the discussion of the achieved results, the solution alternatives, and the present approaches how to extend the utilization, improve, and simplify the entire process.

Setting up a virtual model of a production system is time-consuming and requires expertise in the use of simulation software as well as knowledge of the operations in production systems. For this reason, the creation of virtual models as a basis for the Digital Twin of a production system is hardly feasible for many companies. This chapter presents an approach for the efficient generation of simulation models. For this purpose, production systems are first systematically described using an ontology. On the one hand, this description serves to develop a generic creation method. This is necessary to be able to convert any production system into a virtual representation. On the other hand, the ontology provides a structured data model for subsequent modelling of the Digital Twin. In this chapter, an approach is introduced for generating a simulation model of the shop floor, including its functional behaviour as well as a visual representation. Since most steps of the approach are automated, the simulation model can be generated highly efficient. Also, less expertise in handling simulation programmes is needed. The remaining manual steps require either no special IT knowledge or can be carried out via services.

Product and service innovations are the key driver of the competitiveness in the manufacturing industry. For long time, partnerships between manufacturers, suppliers and service providers have been established to improve overall performance of entire supply chain. Most members in such a supply chain try to expand their business by developing platforms and building their ecosystems. This also occurs in the IT-related industries, in particular when services are provided for large companies. In this chapter, the commercialization of the DigiTwin solution is presented as an extension of a well-running ecosystem OpenDESC.com. The background to the emergence of OpenDESC.com, the customer base and the corresponding offering are described. The method how to systematically plan changes, improvements and extensions by enterprise architecture integration is discussed. The extended offering for new customer categories based on the outcome of DigiTwin is classified and described. The method for assessment of this ecosystem is proposed. Finally, conclusions and outlook summarize the insights and give future directions.

The concept of Digital Twin is almost twenty years old, posing an enhancement of well-known concepts like Computer-Integrated Manufacturing (CIM) and Product Lifecycle Management (PLM), with expected benefits through ubiquitous simulation, real-time analysis and synchronous processing. The theoretical framework and practical implementations of Digital Twins still do not follow this vision rather discover some specific characteristics of the Digital Twin. Although many successful implementations exist, a Digital Twin is still offered and sold as a project for a company’s specific purpose. While sufficient implementation details are not publicly available, it is difficult to assess effectiveness of different solutions and make comparisons in a structured manner. A recent taxonomy from literature is used to explain the state of development of Digital Twin. At first, the way for the further development of the presented approach is discussed. At second, current trends and challenges related to the Digital Twin and its research domains are explained in a comprehensive overview. The assessment is based on eight dimensions and 18 characteristics which serve as an initial part of a future universal reference framework. The Digital Twin applications in different domains are classified and the possible Digital Twin evolution is discussed. Novel research trends and challenges are identified, advancing the theory and practice of Digital Twins. The results show an evolution of Digital Twin's role from an enabler of cyber-physical systems to a product lifecycle data integration and processing platform.