Advances in Automotive Production Technology – Theory and Application

The ongoing transition towards electro-mobility requires an increased reactivity and reconfigurability in automotive assembly. However, the traditional line assembly, which is characterized by rigid cycle times and linear product flow, has already been pushed to its flexibility limits. Drivers are the increase of product changes, variants and derivatives within assembly lines. To further increase reactivity and reconfigurability, matrix structured assembly configurations are a possible solution. Several studies highlight the theoretical advantages, but it has not been applied and validated in industrial use-cases, due to the high transformational gap between line and matrix configurations. In contrast, segment-wise line-less structures show a high potential for this.A use-case oriented approach improves reactivity and reconfigurability by implementing an agile hybrid assembly system that combines the advantages of line and matrix structured (also referred to as line-less) assembly systems and offers a lower investment threshold. Three fields of action are presented: The first consists of flexible planning and control software modules. Within the planning phase, an automated scenario analysis is performed for optimization by applying simulations. During the production phase, the simulated model is re-used for the operation of a dynamical multi-agent manufacturing execution system with online scheduling algorithms. The second field of action deals with reconfigurable infrastructures, which comprises short-term dispatching intralogistics and a flexible layout, facilitated by AGV transport routes and reconfigurable self-adaptive workstations. The third field of action comprises a system model that is an underlying fully integrated digital twin. Control interfaces integrate the infrastructure into the manufacturing execution system to enable rapid system changes.The presented hybrid system contributes to the design of future assembly systems by showing which aspects of line and matrix configuration can be combined to have a beneficial impact on a broad spectrum of production scenarios. By considering all relevant fields of action in a holistic way and by analyzing a hybrid configuration, the arising challenges for producing companies are addressed in a practical and functional manner.

An increasingly uncertain market environment, high product variety and shortened product life cycles lead to an increased demand for adaptable production systems. Due to higher initial investment costs, it becomes more difficult to assess the profitability of such production systems with conventional methods, since the advantages of adaptable production systems are not considered sufficiently. This article presents an approach allowing to determine the economic feasibility of highly adaptable production systems which are repeatedly undergoing reconfiguration processes to adapt to products, processes and technologies that are unknown during planning and launch. In contrast to others, this approach considers a preferably high level of adaptability enabling the production system to change extensively and quickly. To test the method a scenario from the publicly funded project Fluid Production is used.

Since the introduction of the assembly line in production around 100 years ago, the principal of mass and series production has not changed much. However, in the last decades, more individualized products lead to higher product variants, which challenge rigidly linked assembly lines. To provide higher adaptability to changing product variants and volumes, in manufacturing as well as in assembly, the concept of a production system structured as a matrix is developed (abbreviated as MMS). Here, the equipment of the production system is composed of various process modules providing the needed functions. Depending on the needed functions, the work pieces literally search by and by their way through production. The process modules themselves consist of one or more stations providing process functionalities. Assuming that these stations can be distributed to the various process modules in short time, this production structure offers a high changeability during operation. It can be used to reconfigure the system continuously to changing production programs. Through the high degrees of freedom of a matrix production system, finding this optimal configuration of the equipment can be seen as a complex task. For the initial planning of the system, several approaches exist. However, so far, there is no method for reconfiguring the system to changed requirements, mainly to changes of the composition of the production programs, during the operation of the production system. This paper gives an overview of the task and sketches an approach of how to indicate that a reconfiguration is beneficial and subsequently to find and realize an optimized one. Therefore, the feedback control technique is introduced and it is shown, how it can be applied in continuous change processes of production entities. Then, the technique is adapted to apply it to the reconfiguration problem of MMS. Finally, the needed research to realize that approach is outlined.

Digitalization offers new opportunities to improve the quality of planning, adaptation and optimization of assembly processes. The approach presented in this paper allows for the (re-)planning of the assembly process with alternative assembly paths. Since usually multiple valid assembly sequences exist for one and the same product, an automated assembly path analysis is realized for the identification of a time-optimal sequence. The implementation is realized using a combination of Petri nets and graph-based design languages. This framework allows an easy integration of the assembly planning process into the digital design process and the automatic evaluation of possible assembly sequences.

Volatile market demands, stronger regionalization of markets, ever-shortening product and innovation cycles as well as an ongoing demand for individualized products increase the need for adaptable production systems. More than a century after the start of mass production, alternative production systems are required to go beyond the current state of the art concerning adaptability, flexibility and reconfigurability to market requirements and demands. Fluid Manufacturing Systems (FLMS) describe such new production system concepts. The basic idea is to dynamically adapt and change all logistics and production processes, based on the comprehensive application of cyber-physical production systems (CPPS), thus enabling ongoing change in setup, configuration and product scope. CPPS provide a high degree of changeability, thus allowing for fast adaptions of the system to the changing requirements. Therefore the processes are continuously assessed, benchmarked and reconfigured to match the functional capabilities of production and logistic resources to the actual requirements originating from products and external influencing factors. Within this paper, conventional production systems such as Dedicated Manufacturing Lines (DML), Matrix Manufacturing System (MMS) and Flexible Manufacturing Systems (FMS) are described and characterized using defined criteria. The paper closes with a description of the Fluid Manufacturing System (FLMS), the core hypotheses and the advantages of the presented concept compared to conventional productoion systems.

Due to growing public awareness and rising requirements of legislation and customers’ expectations in the field of sustainability, it is increasingly important for enterprises to assess and subsequently reduce their environmental impact. However, the acquisition of environmental data in enterprises still causes considerable effort, due to the necessary manual acquisition.A unified asset administration shell (AAS) potentially provides data transparency and environmental data interoperability along the value chain and thus a more detailed (real-time capable) accounting of the environmental impact of products and services. Hence, this paper presents an approach for an automated environmental impact assessments (EIA) of products and production sites.Thereby, a first application of the AAS in the context of automated EIAs was implemented at the ARENA2036 research factory. The AAS automatically collects energy and emission data throughout a production process and thus allows the allocation of actual emissions to product and equipment (environmental wallet). The results reveal a first starting point for automated EIA, facilitating individual EIAs to address increasing product variety.

Higher individualization levels, shorter product life cycles and fluctuating market demand influence manufacturing environments considerably, leading for example to a greater planning complexity and more frequent reconfiguration of production equipment and systems. These factors generate high non-value-added production costs that reduce factory adaptability.In this context, automobile manufacturers demand a reduction in installation and integration activities that allow for an increase in the versatility of their production facilities. Business model innovations are hence required for manufacturing equipment suppliers to meet their customers’ demands and stay competitive in the long term.This paper identifies promising approaches for manufacturing business model innovation, including preliminary assessment within the research campus ARENA2036. The approaches lead to alternative business models that represent a possible way out of the conventional business dynamics between automobile manufacturers and their equipment suppliers, currently inhibiting innovation towards more flexible production systems.

Ensuring high availability despite the growing frequency of changes in production requirements leads to an increased reconfiguration demand in the domain of industrial automation systems, which will be dominated by Cyber-Physical Production Systems (CPPS) in the future. Therefore, a concept, covering a methodology containing four steps, is utilized to answer the research question how to enhance CPPS with a self-organized reconfiguration management. This contribution focuses on the first two steps of this concept: the identification of reconfiguration demand and the generation of alternative configurations. A condensed presentation of the related work reveals that an interface-oriented formalized process description depicts an appropriate conceptual basis. Building upon this, the presented concept contains a capability model of the CPPS and the reasoning to determine a potential reconfiguration demand, as well as a procedure for the generation of alternative configurations. To evaluate the concept, an agent-based implementation is given, which uses an OPC-UA controlled simulated modular production system as a substitute CPPS in discrete manufacturing.

In many manual assembly systems, there is great potential for optimization, especially when products in small quantities, high variants or with high complexity are produced. The more often the assembly is changed, the greater is the potential.The main reason for the optimization potential is the still high effort required for an assembly planning. Especially in today’s challenging and volatile environment, classic assembly planning often reaches its limits. As a result, assembly systems are often not planned in sufficient detail. The consequence is a lack of transparency: Workers in assembly do not get clear work instructions and planners do not get feedback from the assembly.There are approaches to reduce the effort required for assembly planning meeting the challenge from two sides: On the one hand, there are approaches to further integrate assembly planning with previous processes, such as product development. On the other hand, there are approaches that optimize the processes from an assembly perspective.This paper focuses on a method to optimize assembly sequence planning based on actual data. Data is collected, for example, via sensors in the assembly area. Afterwards, different runs of the assembly process are analyzed. Then, an algorithm derives the best practice to assemble the product. Best practice describes the assembly sequence that leads to the fastest assembly. The method fits into a methodology to transfer benchmarking to manual assembly and can be used for a one-time optimization project as well as for continuous optimization. The results generated in the algorithm are then made available to workers and planners.

Today’s productions are driven by increasing variants, uncertainty of variants’ distribution as well as volume and shorter innovation cycles.A matrix manufacturing system aims to tackle these challenges. This new system concept consists of independent and flexibly linked process modules, which have no uniform cycle time and no fixed product order sequence in the system.However, new challenges arise from this system setup. The changes in the structure have an impact on logistics. In research, there are only a few investigations regarding the consequences to logistics, especially to material supply. Common and new innovative supply strategies are used without knowing their suitability to the new system. The applicability of kanban, single product supply or kitting basket supply differs to the usage in a line assembly. A systematic derivation of suitability in the new context is missing.The paper fills the gap of research in the outlined field. Firstly, changes and characteristics through the new structure which occur as challenges to the material supply are investigated. These are e.g. the flexibility of order and process sequences. In a second step, the material supply strategies are evaluated according to strategic requirements. As a result, each material supply strategy’s suitability is evaluated for usage in matrix manufacturing systems.The paper concludes with a derivation of guidelines for the planning and selection of a material supply strategy.

Smart Factories are designed by Universities [e.g. 7, 1] and Research Institutes (e.g. IPA, IFF, IOSB, IFF, DFKI). Also, automotive companies are developing their ‘own’ Smart Factories [e.g. 3, 6]. The focus is on manufacturing processes, technical equipment, and control of technical processes but not on business and order processes [4, 5]. To run a Smart Factory also a concept for a “Digital Order Twin” (DOT) is needed to ensure that for each final product the right components are available in the right quantity at the right time and at the right place, especially if the product is configurated by a customer or dealer. But existing IT-Systems for Enterprise-Resource-Planning (ERPS), Material-Requirement-Planning (MRP) and Manufacturing Execution (MES) are not able to exploit the huge amount of acquired data ‘in-time’, because they are separated IT-Modules which are connected by batch-oriented interfaces only, the concept and algorithm base on the “Water-Fall-Model” without recursion to a preceding IT-System or IT-Module. In an autonomously controlled Smart Factory the right components must be referenced to each single final product instantly and in a more flexible way than the concept of pearl chain can do, instead of a unique DOT uses instantly data of RFID, QR-Code, Smart Devices and Cyber-Physical Objects and can interact in time with Systems and Modules.

Flexibility provides a way to address technological progress and changing user needs in a vehicle’s architecture. Yet, this has serious implications on the development and manufacturing process. It necessitates additional investments in long-lasting interfaces and changeable production design. Within this paper, we address the research question “which components of a vehicle should be designed flexible and which parts can remain static to streamline investments in flexibility?”. We develop a methodology to support this decision by combining technology intelligence methods with comprehensive user acceptance research. Our approach builds upon a certain notion of function, which connects technological progress with customer utility. To this end, we analyze patents to evaluate technology dynamics and a user survey to detect functional needs. The research project FlexCAR designs a flexible vehicle architecture and was used for evaluation. Thereby, the methodology showed significant and sound results in its application and revealed flexible components with little effort. To the best of our knowledge, it is the first methodology of its kind.

Automotive production systems face the challenge to produce models and brands with different drive concepts and individually configured equipment variants in a highly efficient way.Studies on modular assembly systems in automotive industry have demonstrated potential for productivity gains through the implementation of an alternative, value-add-oriented process organization. The rigid concatenation of mechanical production processes is the limiting factor; firstly, for an efficient implementation of product individualization and secondly, for a highly available robust production which can optimize the overall factory production flow.Rising degrees of freedom in material flow control associated with more flexi-ble production flow increases the complexity of the overall production system. Decision support for humans by planning systems with integrated control logic is thus a decisive factor for mastering complexity. Currently, the overall perfor-mance of modular manufacturing processes is not sufficiently supported by the IT-architecture on factory level. The individually operating subsystems are not capable of supporting reactive manufacturing control.As a basis for reactive manufacturing control, the information requirements towards modular manufacturing processes across different domains are defined in this paper. Furthermore, a cross-system information and communication matrix is proposed that structures information processing between individually operating subsystems. The application of broker-technology could subsequently enable holistic information-based control on factory level to support human experience-based decision making.

For the distribution of consumer and industrial goods, every company is obliged to provide assembly instructions. For the consumer goods market, for example, the German Civil Code defines that defective assembly instructions must be declared as a material defect. However, the creation of comprehensible and defect-free assembly instructions is still a very time-consuming manual process, which must be determined in an extremely time-consuming procedure. Nonetheless, assembly instructions are more than just obligatory documents. They are also required in places where they are not prescribed. For example, assembly instructions are needed in production to pass on assembly knowledge to the assembly operators. Here, it often turns out that this knowledge is either not available or can only be used to a limited extent.The two key elements of an assembly instruction are the assembly sequence and the visual illustrations. Currently, the assembly sequence is determined manually by the designers based on their personal experience, whereas illustrations are generated with costly software tools which are not even able to check the feasibility of the planned instruction. This work presents a novel approach to generate assembly instructions directly and automatically from CAD models of the designers. For this purpose, the commercial CAD software SolidWorks was extended by a macro Tool. All necessary data to generate an assembly instruction are extracted from the CAD model. The extracted data are assembly features, stability and geometric restrictions, subassemblies and assembly directions. Based on these data, the assembly operations are evaluated with a fitness function which includes the attributes like tool changing costs or distances of assembly paths. The whole assembly sequence optimization process was modeled as a Travelling Salesman Problem. After the ideal assembly sequence was found by the macro Tool, this tool also generated matching visualizations of the assembly operations based on the CAD model. The approach was validated by three different models, an assembly benchmark, a single-cylinder engine and a gear box.

Selective Assembly is used to reduce the impact of geometrical variation of parts on the quality of their assembly. Therefore, matching rules for appropriate part combinations are determined in order to optimize the assembly quality. In this work, the geometrical features which are considered for selection are derived from the Virtual Assembly of the measured individual parts. Here the measurements are performed with a Computed Tomography (CT) system. In this new approach, the complete population of parts in a batch is optimized instead of a serial optimization procedure. This minimizes or even avoids the number of waste parts. In order to apply this new assembly strategy and to evaluate the achieved results, the assembly of the cover and the housing of a screen washing nozzle is studied, which is joined by laser welding. By means of global optimization using a genetic algorithm, the overlap in the welding seam, which determines the quality of the final product, is optimized. Deviations to the nominal welding seam are reduced up to 45% compared to a worst case assembly.

One of the biggest challenges with the growing amount of renewable energy generation is the fluctuation in energy supply. In the case of Germany, renewable and volatile energy resources (wind and solar) are expanded. Industrial demand-side management, therefore, plays an important role for the automotive industry in Germany with its high energy demand. For a sustainable production manufacturing processes need to be more energy efficient and adaptable to volatile supply. The main goal is to synchronize manufacturing processes and their energy consumption with energy supply. While there are holistic concepts and ideas for a service-oriented platform for energy flexibility, a defined workflow for implementing energy flexibility signals at the shop-floor level is still missing. Our work proposes a method to adapt manufacturing processes with consideration to energy flexibility. The presented method aggregates data from a manufacturing process. Therefore, sensor data interoperates with flexibility signals on the shop-floor level where an intelligent controller can set process parameters by communicating through an OPC UA Server with the PLC. As an important cross-sectoral technology in the automotive industry, flexible compressed air is used for the validation of the presented method.

Digital twins represent physical systems through dynamic adaptive digital replicas. These replicas are virtual images of the functionality and interactions of the physical system and its components. Digital twin provides real-time monitoring and decision-making support. These are essential pillars in the Industry 4.0 paradigm. On the system level, multiple architectures are established for digital twin concepts in the literature. However, the roadmap for deploying a functional digital twin has not been fully recognized thus far. In this paper, we propose a framework for the deployment of a digital twin in production systems. The framework covers both levels of virtualization; digital shadowing, and digital twining. It utilizes present technologies in network communication, data management, and knowledge extraction. We describe the main components of the digital twin, and their relation to the management schemes currently implemented in production systems. Additionally, we devise an approach for integration of available simulation and data analytic tools for dynamic modeling and system performance evaluation. Through our proposed framework, the current technologies and tools are capable of deploying a digital twin. Consequently, learning algorithms and monitoring procedures are exploited for dynamic performance improvement.

ARENA2036 is a joint research campus incorporating production assets from different industrial and academic partners. To allow the implementation of cross-partner value streams and workflows, a common middleware for online date exchange and asset operation is mandatory. We implemented a generic and lightweight middleware which follows the concept of Asset Administration Shells as specified by the Platform I4.0. However, to allow for easy adaption and setup by a diverse range of partners we simplified modeling requirements and complexity of the actual data exchange. The result is a specification for the assets’ self-descriptions in form of submodels (For many people, the term “submodel" implies the existence of a “model", which does not exist here. In order to have a consistent terminology with the standard of the Platform I4.0, we will use the term “submodel" here. For the future, we propose the term “aspect model" instead.) and a convention on how to map this onto MQTT. Additionally, we integrated means for online discovery and state monitoring of all connected assets.

AutomationML is an open neutral XML based data exchange format used in automation systems. It has come into the public for more than 10 years and is being used in many different areas in all kinds of manufacturing applications, such as digital twin, reconfigurable manufacturing systems, heterogeneous data exchange, etc. However, no comprehensive literature review on the research and application progress of AutomationML has been found since the initiation of AutomationML. Based on the study and analysis of AutomationML related publications, this paper gives a detailed illustration on the state-of-the-art of AutomationML. Firstly, the background and terminologies related to AutomationML are introduced. Secondly, the paper applies a methodology to collect AutomationML related publications, on which an analysis based on a multidimensional literature classification is conducted. Thirdly, according to the analysis results, current research status and whether AutomationML can meet the requirements for industry 4.0 environment are discussed. Finally, conclusion and outlook are illustrated in the end.

Modularity, adaptability and integration of new technologies like Human-Robot Cooperation (HRC) help in facing the major challenges posed by the increased product variants with shortened life cycles and fluctuating market conditions of the automotive industry. However, utilizing them requires strong software support and complicates the already demanding planning and implementation of an assembly system. The strong dependency on software creates a new void in the planning and implementation processes. Usually, the programmer, not the process owner, fills this void based on his knowledge. This results in frequent and resource-intensive adaptations during commissioning due to implicit knowledge and requirements during the development process. This paper presents a lean approach for implementing an adaptable assembly system. Our approach combines an abstract process description, a virtual model of assembly system and a standardized control system which enables the realization of an assembly system. Our modeling technique helps a process owner to develop robust assembly systems. Also, it enables the design of a process and supports in obtaining the corresponding boilerplate code needed to execute the process on standard hardware utilized by the industry. This is demonstrated and tested by means of a HRC underbody assembly process in vehicle assembly under realistic conditions in a demonstrator factory.

An ambitious reduction of wire harness design time and cost by at least 50% each can only be achieved by automation. In the ITEA2-project IDEaliSM, it was demonstrated that the automation of 3D wire harness generation can be achieved using graph-based design languages and the VEC (vehicle electrical container) as an open data standard for wire harnesses. All required data and process steps for generating the wire harness are described, as well as a wire harness assembly simulation.

This paper presents a middle data layer that is designed and implemented based on the ANSI/ISA-95 industrial standard. The proposed middle data layer extracts key information on manufacturing operations from the control systems as well as the business systems. The proposed model enhances interoperability in Industry 4.0 by using a standardized and formalized data structure. We expand the above idea into three directions. Firstly, we analyze all the layers of the traditional automation hierarchy model of ISA-95 to get an overview of objects and activities that are needed for building a middle data layer and re-structure the traditional hierarchy levels by introducing the ISA-95 based middle data layer. Secondly, we design the data architectures by categorizing the data source and explain the formalized and standardized data models. Finally, we use a Smart Manual Station of a production setup to show how to apply the proposed ISA-95 middle data layer to the real world case. The results indicate that the middle data layer enhances the interoperability of manufacturing systems and creates a universal standardized data structures for systems integration for factory automation.

The Internet of Things (IoT) is coined by many different standards, protocols, and data formats that are often not compatible to each other. Thus, the integration of different heterogeneous (IoT) components into a uniform IoT setup can be a time-consuming manual task. This lacking interoperability between IoT components has been addressed with different approaches in the past. However, only very few of these approaches rely on Machine Learning techniques. In this work, we present a new way towards IoT interoperability based on Deep Reinforcement Learning (DRL). In detail, we demonstrate that DRL algorithms, which use network architectures inspired by Natural Language Processing (NLP), can be applied to learn to control an environment by merely taking raw JSON or XML structures, which reflect the current state of the environment, as input. Applied to IoT setups, where the current state of a component is often reflected by features embedded into JSON or XML structures and exchanged via messages, our NLP DRL approach eliminates the need for feature engineering and manually written code for pre-processing of data, feature extraction, and decision making.

One of the biggest challenges currently facing the industry is to make systems more flexible. Technological advances are enabling new application scenarios such as safe and efficient human-robot collaboration, which is based on the real-time availability of required data at the location where it is needed at that point in time. The crucial building block for this modernization of factories is the communication network between the automation components as it is the one enabler for constant and reliable exchange of information. In the future, an essential and ever increasing part of such a communication network will be wireless communication. It plays a major role in expanding the mobility and agility of sensors and actuators and in the retrofitting of rigid legacy structures. For smart production in particular, wireless technologies that are reliable and support deterministic cycle times in the range of less than 1 ms are required, which is barely addressed by current wireless developments based on IEEE 802.11 and 5G URLLC. This paper presents a novel real-time radio technology – Ultra Reliable Wireless Industrial Network (UWIN), developed by Fraunhofer IIS, which is currently in a preliminary pre-product development stage. In addition to explaining the UWIN concept, this paper provides a short overview on state-of-the-art real-time radio technologies.

Industrial energy efficiency is widely acknowledged as an important and effective way to mitigate industrial greenhouse gas emissions. However, so far, implementation rates of energy efficiency measures are disappointing. Industry 4.0 has been described as a way to increase energy efficiency and, therefore, barriers to energy efficiency. Here, we share the concept of a digital service providing assistance for investments in the replacement of energy consuming physical assets (Hardware) by continuously calculation the life cycle costs of the underlying assets and more energy efficient replacement versions (Upgrade). We present the specification for this digital after-sales service and a basic implementation in the context of a use case applied in the context of a compressed air demonstrator.

In this paper we present a novel method for performing in site real time quality inspection (QI) and consequently, digitalization of manual processes performed by human workers. It complements and improves our previous work in this area, which makes use of telemetry gathered from a smartwatch to classify manual actions as successful or unsuccessful. This new methodology provides the worker with a real time capable, robust and more accurate quality inspector. This work enhances the existing system through the elimination of input from the user by making use of a BIOX bracelet that detects gestures. The signal processing and classification methods are simplified and optimised by using assembled neural networks thus merging together the data gathered from multiple signal sources. Consequently, the overall QI system is improved with around 70%, thus furthering the necessary development needed to have a system ready to be used on a production environment.

Robotic arm performance varies due to normal and abnormal events. Normal events may include degradation of equipment, motors, mechanical system joints, and gears, while abnormal events may occur such as faulty episodes. In this paper, we address positional performance degradation that can be stopped and redressed if suitable required action is achieved. The Tool Center Point (TCP) position measurement devices are expensive, hence unavailable to every robot. Some industrial processes are critically sensitive to target tool position such as assembly, pin and past, and material handling. We propose a data driven artificial intelligence tool to detect anomalies and degradation of the robotic arm for a positional health assessment without the need for special advanced sensors. TCP deviation is predicted using deep machine learning models that train on time series of historical data of the robot’s performance. Statistical thresholds are calculated to detect the robotic arm’s degradation and anomalies by performing residual analysis. An alarm system is built by applying the proposed monitoring tool online.

Entanglements can cause robots to pick multiple parts within random bin picking applications. Previous approaches cope with this problem by shaking the gripped workpiece above the bin. However, these methods increase the cycle time and may decrease the robustness of the application. Therefore we propose a new method to separate entangled workpiece situations by using deep supervised learning. To generate annotated training data for a convolutional neural network we set up a simulation scene. In this scene, bins are filled with different amounts of sorted workpieces in several entangled situations. Each workpiece is then moved into different directions to path poses which are evenly distributed along the surface of a hemisphere. The emerging dataset consists of cropped depth images of entangled workpiece situations and several path poses. A serial connection of convolutional neural networks is trained on this dataset and proposes a sequence of poses yielding the general departure path. Finally, the performance of this method is validated on simulated data. To the best of our knowledge, our proposed method is the first systematic approach to find the best extraction strategy to separate entangled workpieces in a pile while decreasing the effective cycle time for gripping entangled workpieces and increasing the robustness significantly.

In random bin picking, grasps on a workpiece are often defined manually, which requires extensive time and expert knowledge. In this paper, we propose a method that generates and prioritizes grasps for vacuum and magnetic grippers by analyzing the CAD model of a workpiece and gripper geometry. Using projections of these models, heatmaps such as the overlap of gripper and workpiece, the center of gravity, and the surface smoothness are generated. To get a combined heatmap, which estimates the probability for a successful grip, all individual heatmaps are fused by means of a weighted sum. Grid-based sampling generates prioritized grasps and suggests the most suitable gripper automatically. This approach increases the autonomy of bin picking significantly.

Flat knitted data gloves can be used as an assistance system for the detection of touches and gripping forces in logistics and production processes. By being able to determine, whether an action was carried out in accordance with the specifications, errors can be reduced, efficiency increased, and process documentation facilitated. This work focuses on the development and characterization of knitted sensor structures, which can be integrated during the production process of a completely knitted work glove. Compared to the integration of conventional rigid sensors into textiles, a fully textile sensor has advantages in terms of elasticity, breathability, draping properties, and flexible design. As it is a single step production process there is a high potential for very low production costs. Gripping forces, biomechanical requirements as well as knitting requirements were defined. Yarn material was fabricated for both the electrodes and the sensory functional layer. Different sensors were then manufactured and subsequently evaluated in regards to electromechanical properties such as sensor sensitivity and reproducibility of the sensor signal. Since the glove is stretched differently when worn, the behavior towards stretching was investigated. When subjected to dynamic forces while pre-stressed by different tensile forces, a clear distinction between loaded and unloaded states is possible.

The current market trends demand a flexible production technology, but today’s control systems are monolithic and therefore not flexible. The monolithism can be broken by using container-based virtualization and a microservice architecture. In a microservice architecture, messaging is used to achieve independence of services. Furthermore, deterministic real-time communication is necessary due to the control environment. Since messaging emerged from the IT world, real-time capability is not given. Therefore, in this work, a concept for real-time messaging is developed, implemented and validated on a test system.

In the present investigation a conception of an adjustable hemming die for the assembly of a roller hemmed engine bonnet in automotive industry is presented potentially permitting time and cost savings in serial production. Numerical simulations are performed revealing that the fitting accuracy of the roller hemmed part can, in principle, be drastically improved. Based on the results, a mechanical construction of the hemming die is developed, a realized prototype of which is experimentally explored. Tests demonstrate that the fitting accuracy of the assembled part can be manipulated significantly.

Large sheet metal parts with low drawing depths usually reveal only low work hardening levels in the central area of the components. As a result, such parts often do not meet specified property requirements. Short-cycle stretch technology (SCS) represents a novel forming process for combined stretching and deep drawing, which allows high-quality production of such large sheet metal parts with improved mechanical properties. Due to a consistently simple tool design with changeable die inserts, SCS technology additionally offers a cost-effective and relatively flexible alternative to currently known processes for combined stretch forming and deep drawing. This paper defines the basic principle and presents close-to-production tool concepts of the newly developed SCS process. Furthermore, the high energy saving potential of the novel forming technology is illustrated in comparison to conventional deep drawing.

A central success factor for the digitalization of production processes is the provision of a consistent database across domains and life cycles. The current situation in the automotive industry is characterized by a multitude of engineering tools with diverse, proprietary data formats, which require complex conversion processes and show dramatic deficits in the consistency and accessibility of the data. In recent years, graph-based design languages have been refined and their range of application expanded to an extent that they represent an interesting approach to addressing these problems. The focus of this paper is on the automated generation of assembly processes and assembly resources (e.g. type-related production equipment such as clamping devices) using the example of automotive body parts (front flap and B-pillar). Within the presented engineering framework of automated production planning, clamping concepts and assembly cells (with automated wiring) are generated using graph-based design languages. Based on the manufacturing concept, a joining sequence and fastener planning is carried out, from which a product-specific clamping and fixing concept is derived. The central advantage of this procedure is, in addition to the high degree of automation, the possibility of providing a consistent database. This database allows the automatic derivation for the various specialized engineering tools.

The increasing diversity of products and variants requires a flexible and fast path generation in robot-based painting processes. In the state of the art, path generation in the painting industry is a time consuming and cost-intensive iteration process in which the generated paths are evaluated and optimized via painting trials. In this paper, we present a novel concept for a self-programming painting cell, which is based on the key technologies 3D-scanning, multi-physics painting simulations, and a contactless film thickness measurement using terahertz technology. The core element of this cyber-physical painting system is a unique combination of numerical painting simulations with a gradient-based multi-objective optimization method, to virtually compute painting paths that produce a homogeneous thickness on the painted object. In order to drastically reduce the time and computationally intensive numerical fluid dynamic simulations, a step-by-step coupling of an offline and online simulation was implemented. In a final step, a collision free robot motion without singularities is generated automatically from the computed painting path. The concept was validated under pilot plant conditions by the painting of a fender using an electrostatically assisted high-speed rotary bell atomizer. The paint film thickness, measured with terahertz technology was used as the target and validation criterion, as it shows a strong correlation to other quality values. The results show that the achieved film thickness was within the process specification, although deviations between simulated and measured film thicknesses were found in the edge zones of the workpiece.

Decisions on the implementation of innovative concepts and technologies into automotive pre-treatment lines are regularly marked by uncertainty regarding trade-offs between economic, ecological and technical aspects. Large amounts of water are consumed during the production in car body painting plants. It is hardly possible to avoid that certain microorganisms (MOs) proliferate in process water and pre-treatment bath tanks. If the bacterial load increases too much, the quality of the paint finish is likely to be impaired. Therefore, chemical biocides are regularly used in pre-treatment and dip coating plants. However, repeated use of the same biocides can lead to resistance of some MOs strains. In addition, stricter legislation is gradually withdrawing certain biocides from the market, making it more difficult to obtain approval for newly designed active substances. Hence, conventional decontamination methods might no longer work in the future. The Pulsed Electric Field (PEF)-Treatment is an innovative technology within the field of pre-treatment lines. By applying high voltage pulses (kV range, µs duration), a high field strength is generated in the process fluid, across the cell membrane of the MOs, which permeabilises the cell membrane. As a result, the MOs lose their cell interior (cytoplasm), and most likely die.The results of the project show that PEF-treatment has a high market potential, several advantages but currently also higher cost compared to conventional biocide treatment.

Reliability and trust in the new technologies are essential if the autonomous, connected and electrified automobile is to become established in society. The more the performance of an electrical/ electronic (E/E) system increases, the more sensitive it becomes to particulate contamination. The risk of a short circuit increases exponentially rather than linearly. Integrating technical cleanliness into product design and production is essential to reduce the risk of a short circuit. But what is the right level of cleanliness? This question can soon be answered with the newly introduced fault injection test procedure. The method can contribute to increasing road safety in areas where normative or legislative requirements have been lacking up to now.

In most current battery electrical vehicles the battery modules are positioned in the underfloor area. For sporty-emotional cars a lower seating position might make it necessary to use the package in the rear end. In addition, lightweight design solutions are necessary. This paper deals with the question of how a new innovative and integrative rear end for a battery electric vehicle can be designed. Therefore, different solutions are designed and assessed, whereby a stack of five battery modules and a specially devolved axle are designed. With regard to separate topics, parts are manufactured and different tests are applied. With this approach it is possible to reach ambitious goals by means of functional integration and close interdepartmental collaboration.

The focus of this work is capturing the generation of instabilities in bondlines of uncured adhesives using the finite element method. Two geometries have been studied, a cylindrical probe and a specimen representative of a possible adhesive line used in the automotive production. Such geometries have been analyzed under tensile loading considering different width/thickness aspect ratios. A Neo-Hookian material model was chosen to describe the behavior of the material and the formation of instabilities during the process. A discretization and a parametric study have been performed to show the validity of the model and the results have been found to agree with results found in literature.

For the production of automotive interior parts, there is a high demand regarding the surface quality combined with large production rates. These requirements can potentially be achieved by using Stereolithography (SL).This work concerns the dimensional limitations of depth of cure during UV-post-curing of parts made of urethane acrylates by Digital Light Processing (DLP). Due to absorption of the UV-light in the photopolymer, the penetration depth is limited. A better understanding of the depth of cure is necessary to define optimal post-curing parameters and potential limits of the wall thickness. To investigate the depth of cure, green-state specimens are irradiated unidirectionally by UV-light with a wavelength of 365 nm and 405 nm. Subsequent hardness measurements according to Shore D are used to determine the corresponding depth profiles. Thereby the influence of photoinitiator concentration, irradiation dose and wavelength on depth of cure is analysed. Additional transmission measurements during post-curing contribute to a better understanding of the penetration depth of the UV-light for optimal production times.For a non-pigmented photopolymer, the measured depth of cure is between 0.59 mm and 8.58 mm. The highest depth of cure is obtained by a low photoinitiator concentration of 0.75%wt, long wavelength of 405 nm and high irradiation dose of 70 J/cm $$^2$$ 2 .

Composite materials have brought new development and sizing possibilities for structural components in transportation systems. Their high specific material properties are enabling weight reduction while increasing structural performance. On the downside, composite materials are generally related to high material and manufacturing costs and increased characterization efforts. Through the braiding technique, profiled structures can be manufactured in a highly automated and reproducible process. Moreover, braided composites can absorb more energy compared to their unidirectional or woven counterparts ( Falzon P. J., Herszberg I., Bannister M. K., Leong K. H.: Compression and Compression-after-impact Properties of 2-D Braided Carbon/Epoxy Composites. Proceedings of the First Australasian Congress on Applied Mechanics: ACAM-96, pp. 297 (1996).).In this paper, we describe the development and validation of a simulation framework as sustainable alternative to material- and cost-intensive experimental testing. Our work aims at considering the influence of manufacturing effects and textile architecture on the material properties and therefore at increasing the reliability of structure sizing. As validation basis, flat specimens of biaxial and triaxial braided composites are first manufactured and tested under quasi-static loading. We then develop a digital twin of the braiding process and its material characterisation. Within this framework, the braid’s textile architecture is predicted with multiple finite-element simulations at the mesoscopic scale.The numerical predictions show the strong influence of braiding angle and braiding core diameter on the textile architecture and consequently on the material properties. More particularly, crucial effects with negative impact on the mechanical properties (presence of gaps or yarn locking) are highlighted. On a pure numerical basis, we finally calculate the process window for braided structures, which links the process parameters to the resulting material properties. The present approach is a crucial step toward the reduction of experimental investigations in early development.

Materials with higher specific and buckling strength are increasingly used in order to reduce the weight of car bodies while maintaining or increasing passenger safety. Examples for such materials are ultra-high-strength aluminum alloys and press-hardened steels. A major challenge now is to combine the ultra-high-strength aluminum alloys with each other and, if necessary, with other metals without losing their outstanding strength properties.High-strength aluminum alloys with virtually no loss of strength properties can be welded with the so-called friction stir welding process since the 1990s. In this process, a rotating tool is pressed into the joint gap of the joining partners and moved along the joint gap to produce the weld seam. An anvil adapted to the component is required on the opposite side of the weld seam in order to absorb the high process forces.A joining process for the body-in-white assembly of aluminum and hybrid car bodies should have the following specifications: Joining high-strength aluminum alloys without hot cracks, high-strength joints, sufficient service life of the joining device, robot-supported execution, no complex and expensive component-adapted anvil and an accessibility similar to resistance spot welding guns.Based on these requirements, a process concept for a new type of robot-guided friction stir welding gun for the production of short, merging weld seams was developed at the Materials Testing Institute (MPA, University of Stuttgart) and a patent application was filed. A first functional model of the welding gun was built and the underlying kinematics were successfully tested in initial tests. The concept differs from other known solutions because it compensates process forces through the weld gun completely. Moreover, the concept allows curved weld seams which are common in the automotive industry. These curved weld seams are realized as polygon courses in a robust manner and without component-specific anvils.

For the automotive industry, automated composite manufacturing via state-of-the-art Automated Fibre Placement (AFP) or Automated Tape Laying (ATL) is of limited interest due to moderate cycle times and limitations concerning applicable reinforcement materials. The novel Advanced Ply Placement (APP) on the other hand offers a high degree of freedom concerning fabric utilization. A unique multi-robotic handling process allows the APP to process dry fibers as well as pre-impregnated fibers with a large variety of fiber areal weights and fabric types. With the APP approach, automated, wrinkle free placement of wide, unidirectional fabrics onto complex, double curved geometries is possible.Offline robot path planning of the APP process is realized via FlexiCAM, a CAE interface for geometry based trajectory generation that allows databased parameter adaption and process simulation.The present paper will explain the working principles of the multi-robotic Advanced Ply Placement and distinguish between established automated manufacturing processes. Results on geometry based path generation via FlexiCAM and their applicability in the automotive industry will be discussed and the benefits of knowledge-based offline path planning will be highlighted.

Manufacturing processes of carbon fiber reinforced plastics (CFRP) components are often characterized by a high fraction of manual activities especially in terms of handling, cleaning and edge sealing. The innovative machining process introduced in the project CFKComplete comprises a fully integrated and automated workflow reducing unnecessary handling and transportation activities. The concept has been realized on a 5-axis machine tool, providing enough workspace for large structural workpieces and the developed technology units. The units themselves are defined by their modularity, scalability and independency of the guiding machine. Key aspect of the innovative concept are four purpose built and developed interchangeable technology units, which are driven by the base machinery. Thus, the machine’s working scope is being enhanced from machining to a fully integrated workflow including quality inspection, cleaning, sealing and curing.

The laser is the only tool that can address all six main manufacturing groups of the German standard DIN 8580 simply by applying different processing parameters. Given the present state of technology, however, different dedicated machine concepts are still being used for the respective applications. To enable digital manufacturing in its full consistency, novel, fully reconfigurable machines need to be developed. We therefore outline the further developments that are required to combine all presently known laser-based manufacturing processes on one and the same machine. As both the laser devices and the knowledge about the fundamentals of laser materials processing are already very advanced, research must now be intensified on system engineering. The vision is an intelligent machine, which is fed with CAD data, semi-finished products, or sub-components and that is capable to autonomously produce the desired components with a 100% quality guarantee at a batch size of 1 – “first time right” – and at the costs of comparable mass-produced items. The machine independently selects the best production strategy, i.e. the best combination and sequence of different manufacturing processes. Such a fully flexible and autonomous laser machine will not only boost the implementation of digital manufacturing on a broad scale and provide an approach to resolve the “polylemma of production” but will also enable the relocalization of value creation and manufacturing back into high-wage countries and by this potentially disrupt today’s globalized value creating networks.

This work describes why additive manufacturing is a key technology for efficient design iterations and rapid production ramp-up with large-scale manufacturing technologies. Laser cutting, injection moulding and folding were used to increase the production capacity of face shields for health care workers during the COVID-19 pandemic. We applied systematic learnings from the iterative processes used for additive manufacturing to these large-scale manufacturing technologies and the respective face shield designs. In cooperation with medical experts, structural and functional design requirements of face shields were identified and are described in detail in this work. The regulatory design requirements according to EN 166 are introduced, which were considered to receive a CE certification for three of the presented designs. The employed manufacturing techniques are specified and the respective implications on the design solutions are discussed. The paper concludes with a summary of the production initiative at the research campus ARENA2036 with a total output of over 13 000 face shields from April to June 2020, which were distributed internationally.

Autonomous robots in logistics are a promising approach towards a fully automated material flow. In order to use their full potential however, they must be able to extract semantic information from logistics environments. In contrast to other application areas of autonomous robots (e.g. autonomous driving, service robotics) the logistics domain lacks a common dataset and benchmark suite covering multiple sensor modalities and perception tasks. This paper conceptualizes a framework for artificial perception research in logistics that aims to close this gap in a sustainable, data-driven way. Our framework consists of three components: (1) A foundation, based on logistics-specific standards, concepts and requirements. (2) An open dataset, covering multiple sensor modalities and perception tasks and (3) a standardized benchmark suite. As shown in other research areas, a common and open platform for data-driven research facilitates novel developments and makes results comparable and traceable over time.

This paper presents a concept of a safety-related sensor system that is intended to ensure the detection of people in the vicinity of automated guided vehicles (AGV). This becomes necessary due to the increased usage of these vehicles in coexistence and in collaboration with people in industrial environment. The normative requirements will also change in a way that in the future people will have to be explicitly detected by safety sensors. Until now, only very simple test specimens have been used in the current standards to represent people. In this contribution, the efforts to expand these test specimens are shown. The experimental setup of the safety-related sensor system consists of three sensors from the sector of consumer electronics and is based on the open source framework Robot Operating System (ROS).

Providing a flexible logistics is an important contribution to retaining and expanding the convertibility of productions. It can be achieved by reducing storage capacities, whereby the goods movement through production is performed ideally in a continuous flow. In combination with the intelligent floor, the novel autonomous guided vehicle provides a cost-efficient and highly automated flow of goods for small-sized payloads up to 50 kg with short idle times. This solution is compared with a manually operated logistics and an in-plant milkrun system for a car interior parts production.

Automated Guided Vehicles (AGVs) have come up to a quite familiar picture of nowadays manufacturing sites and in general they follow some rigidly given maps of the shopfloor. Future-oriented concepts require a much more agile approach that allows AGVs to continue its logistic tasks at the shopfloor securely and safely though being confronted with recently newly aligned facilities, new barriers and boundaries without necessary reprogramming the AGV`s environment in front.The actual paper deals with solutions which were worked out in the research and learning factory “smartfactory@tugraz” where all relevant facilities for these research goals have been available. Additional insights and results derive from the research facility of the Center Connected Industry and its industrial application eco system. So numerous Mobile Working Stations (MWS), the latest version of a Real-Time Locating System (RTLS) and a commercially available AGV belong to the core equipment of the presented research.As a result there could be found, that the specified requirements and new applications of future production and logistic concepts can be achieved via dynamic targets. With this development, targets can be changed locally and the AGV automatically moves to the current target position, without necessary changes in the control software.

The automotive industry, including their whole parts and component manufacturers, are facing challenges in a dimension not known yet.The strong differentiation of the product portfolio in connection with specific production lines holds a high vulnerability to capacity fluctuations in certain lines due to volatile demand within a manufacturer’s product spectrum. From the present point of view, the production of automobiles with fundamentally varying technical specifications requires a flexible and scalable production, which enables an efficient production of batch size one. The key elements of a new concept, developed at the IFT are collaborative intralogistics components. One such system called “Mobile Supermarket” specializes on the supply characteristics of frequently needed parts with high variety and works according to the goods-to-man principle, which requires a highly sophisticated safety and interaction concept.

We present a general approach for realizing safe collaborative behavior and high-accuracy operations in industrial robots. For example, collaborative behavior is required in human robot collaboration, whereas high-accuracy is needed in robotic machining. The operation sequence of a robot with corresponding often-contradictory demands on safety functions, control mode and accuracy is specified in a task-oriented path planner. The task planner generates optimized reference trajectories based on an environmental model, kinematic and task-specific constraints. The resulting reference path is fed into an online operation management, which monitors the workspace, ensures safety, and realizes different online control strategies especially for the dynamic tasks such as human robot collaboration or high-precision operations. For the high-accuracy tasks, we implemented a model predictive controller using a mechatronic model for predicting future configurations of the robot based on the measured present configuration. The model-predictive control is demonstrated for the compensation of deviations due to external forces that acts on the end effector of the robot.

The function of logistics is to move objects from place A to place B Therefore it is an important task to determine the position of objects along the supply chain. Up to now, different technologies for indoor localization have been developed. During this technological evolution, two important aspects have been left behind, the communication between the systems and the central management of the data provided. Within the research project ‘Industrial Indoor Localization’, an open source software standard for environment modeling called ‘Reference Architecture Indoor Localization’ (RAIL) has been designed and tested. This protocol enables Location Based Services to allow querying location and object information quickly. Within the scope of the project, four services are being developed and exemplary described in this paper. At the beginning, a location-dependent order allocation algorithm for order picking is presented. This algorithm reduces the waiting time in narrow-aisle warehouses through forecasting. Secondly, an order orchestration service based on location data for picking control is evaluated. Furthermore, a functional area recognition to support picking and a service for finding points of interest is demonstrated, which includes navigation based on a routing algorithm. Using indoor localization, these services improve the intralogistical processes. These include increasing picking performance or reducing order throughput time by eliminating the need to scan the barcode. Finally, these services increase work safety, due to consideration of functional areas, and improve the transparency of the location of points of interest.

Part complexity increase the ratio of manual tasks on machine tools, i.e. feeding, to maintain high productivity. Mobile robot cells automate these tasks but require cost-intensive retrofittings for additional interfaces. This paper focuses on a new mobile robot cell without any mechanical or electrical connection to the machine tool. Therefore, the solution uses a camera system for wireless machine tool localization and signal transmission.

Flexibility for all different models within one body shop production line could be achieved without changing jigs, grippers as well as joining tools. A comprehensive analysis of the process chain formed the basis for the necessary development steps. In addition, planning methods based on mathematical optimization algorithms were developed and applied to cope with the complexity of the task. All the individual aspects are obtained, and the newly developed components were combined into a complete plant concept and its economic efficiency has been proven by Volkswagen planning department.

A key differentiator in customer satisfaction in the automotive industry is offering a choice of high-dimensional possibilities to customize an individual vehicle. In the luxury segment this equals more than one billion variants. Innovative data-driven processes are necessary for the planning and handling of vehicles in production and distribution in order to guarantee indispensable factors such as stability, flexibility and transparency across the entire supply chain and to deliver the right vehicle to the right place at the promised time. Highly complex business environments, multi-variant products, trends with effects on distribution networks and future mobility concepts confront manufacturers with new challenges. This paper provides a survey of currently used methods and technologies to handle the previously mentioned challenges in the area of the customer order management of an automotive manufacturer. A specific field of research is the concept of planned orders which are derived from early anticipated customer requirements and their utilization throughout the entire planning and order management process. In addition to the generic approaches for achieving integrated planning of sales and production programs as well as the resulting material requirements, artificial intelligence methods are investigated with regard to the concept of the planned orders. For this purpose, various existing approaches to anticipate planned orders and to effectively manage the supply chain are examined. In conclusion, this paper provides a survey of state of the art methods regarding artificial intelligence linked to current research on agile production systems. Despite existing uncertainty in the upcoming years the implementation of stable data-driven processes is crucial. In this context, the possibility to prepare an automotive manufacturer for the upcoming challenges in digitalization and globalization is evaluated.