Automotive Production Line Research Articles

Cytotoxicity and biocompatibility of a material based in recycled polyvinyl butyral PVB and high-density polyethylene HDPE determined in human peripheral leukocytes

Abstract Polyvinyl butyral (PVB) is a polymer resin byproduct of the automotive industry. Separation and recycling of PVB from automotive windshield production lines solve waste management issues and might be used in biomedicine implants, particularly when blended with high-density polyethylene (HDPE). Materials designed for biological applications must undergo pre-clinical safety evaluation, including in vitro biocompatibility and cytotoxicity testing. This study evaluated in vitro the biocompatibility and cytotoxicity of thick composite films made of recycled PVB (rPVB), HDPE, and composites made of concentrations of both polymers using peripheral blood mononuclear cells (PBMCs) from healthy donors. The microstructure and elemental analysis revealed no potential concerns regarding mechanical damage to the cell cultures. After culturing the PBMCs in the presence of the polymers and their blends, we observed no acute release of the proinflammatory cytokines TNF-α and IL1-β, nor evidence of cell death, measured by the release of nucleosomes to the extracellular medium at 24 h nor at 7 days of culture. We also found no damage to the cell monolayer. The stimulation with lipopolysaccharide (LPS), used here as proinflammatory control, induced the significant release of both cytokines and caused damage to the cell monolayer. In conclusion, the pure recycled PVB, pure HDPE, and their composites are safe and biocompatible.

An offset-transformer hierarchical model for point cloud-based resistance spot welding quality classification

Resistance spot welding (RSW) is a widely used welding technology in automotive manufacturing, and weld nugget quality is closely related to the quality of the vehicle body. Offline random checks are largely relied on the quality inspection of weld nuggets, but they have low efficiency and high cost. To address this issue, this paper proposes a deep learning model for RSW weld nugget classification, named the offset-transformer hierarchical model (OFTFHC), which is based on the point cloud data of its appearance shape. OFTFHC uses a hierarchical network structure to gradually expand the receptive field. A local feature module is introduced to extract local features from the point cloud, effectively enabling the recognition of the fine structural features of the resistance spot weld point cloud. A residual ratio module, which is based on MLP_MA and uses max and average functions for feature enhancement, is designed to adapt to the complex spatial structure of the point cloud. The offset-transformer structure is used to learn global context features, thereby enhancing the global feature extraction capability. Through classification experiments on RSW weld nuggets across 5 categories with a total of 1050 samples, OFTFHC achieved an average accuracy of 80.6 %, outperforming existing models. This demonstrates the effectiveness and superiority of the method, making it highly suitable for weld nugget quality control in automotive automation production lines.

Recycling Automotive Plastic Waste: Residual Polyvinyl Butyral (rPVB) as Solid Lubricant in Polyoxymethylene (POM) Blends

Polyoxymethylene (POM) is an engineering thermoplastic commonly used in tribological applications due to its low friction and auto-lubrication properties. To improve its performance further and broaden its applications to higher sliding speeds and loads, it is commonly reinforced with different particles, fibers, or solid lubricants. In this study, the effect of adding residual polyvinyl butyral (rPVB) from automotive windshields production lines on the tribological performance of POM during pin-on-disk wear tests at different loads (12, 15, and 18 N) and sliding speeds (0.25, 0.50, and 0.75 m/s) is investigated. It was found that rPVB acts as a solid lubricant, reducing the coefficient of friction (CoF) up to 50%. It is suggested that the formation of a protective layer of adhered material on the wear track is the mechanism behind CoF reduction. Melt flow index, Shore D hardness and elastic properties decrease linearly with rPVB content. However, the reduction in tensile strength, and thus shear strength of the blends, seems to be more significant on their CoF behavior than the reduction in hardness. No trend was found for wear resistance. Under certain sliding conditions rPVB decreased volume loss, but in other cases it increased it. It is suggested that, for the conditions analyzed in this study, adding 10 wt% of rPVB to a POM matrix results in the best balance of tribological performance, with an overall CoF reduction of 36%, and an average volume loss increase of only 1%.

NON-QUALITY RISK EVALUATION FROM TOOL WEAR ASSESSMENT

The development of part quality virtual sensors requires knowledge and observability of cutting conditions and in particular tool wear as tool are consumables. This paper presents an unsupervised anomalies detection approach to assess tool wear from standard machine load sensors in order to evaluate a non-quality risk metric. The developed methodology combines physics and business rules with density estimators to analyse the behaviour of axes and spindle loads. Industrial data from an automotive production line are used to illustrate the methodology application.

Liquid Metal Embrittlement Cracking in Uncoated Transformation-Induced Plasticity Steel during Consecutive Resistance Spot Welding

In the automotive production line, a single pair of electrodes is employed to produce hundreds of consecutive welds before undergoing dressing or replacement. In consecutive resistance spot welding (RSW) involving Zn-coated steels, the electrodes undergo metallurgical degradation, characterized by Cu-Zn alloying, which impacts the susceptibility to liquid metal embrittlement (LME) cracking. In the present investigation, the possibility of LME crack formation in uncoated TRIP steel joints during consecutive RSW (involving 400 welds in galvannealed and uncoated TRIP steels) was investigated. The results have shown that different Cu-Zn phases were formed on the electrode surface because of its contamination with Zn from the galvannealed coating. Therefore, during the welding of the uncoated TRIP steel, the heat generated at the electrode/sheet interface would result in the melting of the Cu-Zn phases, thereby exposing the uncoated steel surface to molten Zn and Cu, leading to LME cracking. The cracks exhibited a maximum length of approximately 30 µm at Location A (weld center) and 50 µm at Location B (shoulder of the weld). The occurrence and characteristics of the cracks differed depending on the location as the number of welds increased due to the variation in Zn content. Type A cracks did not form when the number of welds was less than 280. Several cracks with a total length of approximately 30 μm were suddenly formed between 280 and 400 welds. On the other hand, type B cracks began to appear after 40 welds. However, the number and size of these exhibited inconsistency as the number of welds increased. Overall, the results have shown that small LME cracks can form even in uncoated steels during consecutive welding of Zn-coated and uncoated steel joints.

Development of data-driven PHM solutions for robot hemming in automotive production lines

Robotic roller hemming is currently one of the most used solution for joining metal sheets in automotive industry, especially for those production lines which need to favor flexibility with respect to raw productivity and is mostly employed to assemble car doors. Hemming is a fairly delicate process since it does not only suffice a technical requirement – to join two panels together – but also an aesthetical one, since the joint panels are an integral part of the vehicle design and as such an important selling point. An unpredicted rupture or an advanced degradation condition of the system would lead to a significant loss in the quality of the final product or to a sudden stoppage of the production line. The development of a PHM system for hemming devices would hence provide a significant advantage, especially if designed to work for both new and legacy equipment. In this paper, we provide the results of a preliminary analysis of a new PHM framework for robotic roller hemming studied to work without having access to PLC data, the employed data-driven methodology is detailed and applied to the case of increasing wear in the head finger roll. Results from different prognostics routines are hence presented and compared.

IMPACT OF COVID-19 PANDEMIC ON THE EFFECTIVENESS OF AUTOMOTIVE PRODUCTION LINES-CASE STUDY

Since its discovery, the COVID-19 virus spread all over the world and caused millions of deaths, this paper focuses on studying the impact of the pandemic on the connected and non-connected automotive production lines. This study is developed on two production lines in an automotive manufacturing factory that assembles 700 cars per day and the study is elaborated following two main steps: firstly, studying the impact of the virus spreading on the OEE “Overall Equipment Effectiveness” of the production lines, which is a quantitative metric used for the evaluation of the line effectiveness based on availability, performance and quality, and secondly analyzing the relationship between these factors and the OEE using the Design of Experiments method.

Applying declarative analysis to industrial automotive software product line models

Program analysis of automotive software has several unique challenges, including that the code base is ultra large, comprising over a hundred million lines of code running on a single vehicle; the code is structured as a software product line (SPL) for managing a family of related software products from a common set of artifacts; and the analysis results (despite being numerous and despite being variable) need to be presented to the engineer in a way that is manageable. In previous work, we reported on lifting declarative analyses to apply to a software product line, rather than to an individual product variant. This paper reports on milestone results from applying lifted declarative analyses (behaviour alteration, recursion analysis, simplifiable global variable analysis, and two of their variants) to automotive software product lines from General Motors and assessing the scalability of the analyses and the effectiveness of reporting to engineers conditional analysis results (i.e., results conditioned on SPL program variants). We also reflect on some of the lessons learned throughout this project.

Spot Welding Parameter Tuning for Weld Defect Prevention in Automotive Production Lines: An ML-Based Approach

Spot welding is a critical joining process which presents specific challenges in early defect detection, has high rework costs, and consumes excessive amounts of materials, hindering effective, sustainable production. Especially in automotive manufacturing, the welding source’s quality needs to be controlled to increase the efficiency and sustainable performance of the production lines. Using data analytics, manufacturing companies can control and predict the welding parameters causing problems related to resource quality and process performance. In this study, we aimed to define the root cause of welding defects and solve the welding input value range problem using machine learning algorithms. In an automotive production line application, we analyzed real-time IoT data and created variables regarding the best working range of welding input parameters required in the inference analysis for expulsion reduction. The results will help to provide guidelines and parameter selection approaches to model ML-based solutions for the optimization problems associated with welding.

APPLICATION OF REBA AND KARAKURI KAIZEN TECHNIQUES TO REDUCE ERGONOMIC RISK LEVELS IN A WORKPLACE

Repetitive works and the works done with inappropriate body postures cause musculoskeletal disorders and workforce losses. Design, technology, and humans must come together in ergonomic conditions. In this study, the ergonomic risk levels that employees in automotive production lines being exposed to were determined by the rapid entire body assessment (REBA) method and by considering anthropometric measurements and ergonomic body postures at a line with high risk level, and a shelf system was designed in accordance with the Karakuri working principle. Through the computer-aided design, the system operation was simulated and solutions for ergonomic risks could be provided before the production. The ergonomic risk level, which was “high” in the current situation, could be reduced with the newly designed mechanism, and an ergonomic workspace could be provided for employees. In the new situation, the ergonomic risk level is “low.” In addition to ergonomic improvements, as the manual transportation works of the employee were eliminated, the cycle time of the line, which was 120 s in the first state, was reduced to 100 s after the use of the Karakuri mechanism and the efficiency of the line increases by 17%.

Local Mixer with Prior Position for Cars’ Type Recognition

Deep learning has attracted attention widely as the successful application of deep learning for vision tasks, such as image classification, object detection and so on. Due to the robustness and universality of deep learning, automotive manufacturing, a crucial part of national economy, needs deep learning to make production lines more intelligent and improve efficiency. However, some superior generally deep learning models, such as ViT, TNT, and Swin transformer, cannot meet automotive manufacturing requirements with high accuracy on a specific scene. As for automotive production lines, engineers usually adopt some smart designs, which can provide prior knowledge for designing deep learning models. Specifically, in an image, the position of target is usually fixed. Therefore, in order to take advantage of prior position, this paper designs a local mixer with prior position to capture local feature. Its main idea is that dividing the whole feature map into window feature maps and connecting window feature maps along channel dimension in order to make convolution kernel parameters for each window feature map are independent from others. Besides, MLP is adopted as global mixer to capture global feature and the pyramidal architecture with CNN is adopted. Comprehensive results demonstrate the effectiveness of proposed model on cars’ type recognition. In particular, the proposed model achieves 97.938% accuracy on our data set, surpassing some transformer-like models.

Continuous Improvement for Cost Savings in the Automotive Industry

Today’s manufacturing environment is characterized by rising production costs. Automotive companies are faced with a challenge of meeting the ever-changing customer needs at the minimum possible cost. Reducing the production costs is of paramount importance for a company to remain competitive in a turbulent environment. How can companies in the automotive industry increase cost savings? This paper focuses on continuous improvement for cost savings in the automotive industry. This was conducted at an automotive production line using the work-study technique. Time and method studies were conducted on the production line of one of the products. This was done under these objectives: to identify excess raw material input and non-value adding activities in the product line, to determine current factory capacity, to determine factory capacity utilization, and to implement improvements on the product line so as to increase productivity. The daily target at the company was eight complete units per shift, and the research explored the potentials of increasing that daily target with little to no increase in the number of operators and workstations. The results demonstrate an increase in productivity by 50% by adding one more workstation and one more operator at the surface finishing station. This was successfully implemented and the company is now producing 12 units per shift.

A Digital Twin Case Study on Automotive Production Line.

The manufacturing sector is one of the areas where the advantages of digital twin technology can benefit mostly. The product development, including its software, electronics, mechanics, and physical behavior, is included in the digital twin of the product. Furthermore, simultaneous data capturing from the sensors and data processing are also available in the digital twin. This enables each phase of the development cycle to be simulated, processed, and validated to discover the potential problems before the production of real components. In this study, the use of digital twin technology in the commercial production phase of the automotive production line with a case study is introduced. This study is one of the most comprehensive studies in the literature related to automotive production; therefore, it puts forth the power of using digital twin technology in that area. As the result of this case study, a 6.01% increase in the commercial production line efficiency and an 87.56% gain for downtime are achieved.

Managed Evolution of Automotive Software Product Line Architectures: A Systematic Literature Study

The rapidly growing number of software-based features in the automotive domain as well as the special requirements in this domain ask for dedicated engineering approaches, models, and processes. Nowadays, software development in the automotive sector is generally developed as product line development, in which major parts of the software are kept adaptable in order to enable reusability of the software in different vehicle variants. In addition, reuse also plays an important role in the development of new vehicle generations in order to reduce development costs. Today, a high number of methods and techniques exist to support the product line driven development of software in the automotive sector. However, these approaches generally consider only partial aspects of development. In this paper, we present an in-depth literature study based on a conceptual model of artifacts and activities for the managed evolution of automotive software product line architectures. We are interested in the coverage of the particular aspects of the conceptual model and, thus, the fields covered in current research and research gaps, respectively. Furthermore, we aim to identify the methods and techniques used to implement automotive software product lines in general, and their usage scope in particular. As a result, this in-depth review reveals that none of the studies represent a holistic approach for the managed evolution of automotive software product lines. In addition, approaches from agile software development are of growing interest in this field.

Online quality inspection of resistance spot welding for automotive production lines

Reliable quality control of resistance spot welding (RSW) is a long-standing challenge, due to random disturbance on automotive production lines. In this paper, a quality evaluation framework is proposed based on dynamic resistance (DR) signals, aiming to accurately predict welding quality. The proposed framework integrates welding process stability with deep learning models. Given the uniform variation pattern of each weld with the same schedule, process stability can be determined based on the reference curve constructed by the low-rank and sparse decomposition method. Subsequently, a one-dimensional convolutional neural network (1DCNN) with channel attention mechanism is developed to further predict welding quality, which can perform channel-wise feature recalibration to enhance the classification performance. Extensive experiments substantiate that the proposed network yields a remarkable classification performance compared with typical algorithms on several RSW datasets collected on an actual production line. This study provides a valuable reference to achieve an intelligent online quality inspection system in the automotive manufacturing industry.

Identification and Reconstruction of Impact Load for Lightweight Design of Production Equipment

This paper proposes a method for determining externally applied impact loads on complex structures using strain analysis. An impact load transducer was developed to determine impact loads. Using this transducer (which incorporates strain gauges), the relationship between the measured strains and applied impact load was studied, and a model for conversion from strain analysis to impact load was developed. The reconstructed impact curve that characterizes the impact peak force, impact duration, and load in the steady state after impact was employed as an input load curve in finite element analysis. The reconstructed impact load was validated by comparing the structural strain measured on the specimen in the experiments and the strain calculated by the simulations. The results show that the maximum difference between experimentally and numerically determined structural peak strains is 3.2 με. Moreover, the method was validated by predicting the impact load of a descending vehicle chassis on the production equipment in an automotive production line. It demonstrated high efficiency and accuracy. The reconstructed load curve obtained using the developed method provides high efficiency in addition to high accuracy. Furthermore, it circumvents the complexities of modeling dynamic impact simulation, including complex impactor shape, interface, and friction conditions. Thus, the developed method provides scholars with an efficient approach for an extensive study of the responses of complex structures in various fields such as stress strain analysis, fatigue analysis, and topology optimization for lightweight design of production equipment.

Development of dynamic wireless power transfer system for vehicle logistics robot

AbstractLaborsaving is a major issue in production. In an automotive production line, a yard that is used as a temporary storage for vehicles before shipping. The vehicle‐loading robot that operates autonomously has been proposed to automate the alignment of vehicles in the yard instead of human driving. In this study, the dynamic wireless power transfer (DWPT) system that includes the structure of the roadside for the vehicle‐loaded robot is proposed. It is proven that the transmitter coil has enough durability for the load of the robot weight using simulation and actual measurement. It is also shown that the reinforcement bars of the road structure cause considerable eddy current loss even with stainless steel. This is similar to loss by coil resistance. This system achieves a 1.8 kW DWPT with an automated coil detection system and a frequency control system, under factory conditions.

Development of Dynamic Wireless Power Transfer System for Vehicle Logistics Robot

Laborsaving is a major issue in production. In an automotive production line, a yard that is used as a temporary storage for vehicles before shipping. The vehicle-loading robot that operates autonomously has been proposed to automate the alignment of vehicles in the yard instead of human driving. In this study, the dynamic wireless power transfer system that includes the structure of the roadside for the vehicle-loaded robot is proposed. It is proven that the transmitter coil has enough durability for the load of the robot weight using simulation and actual measurement. It is also shown that the reinforcement bars of the road structure cause considerable eddy current loss even with stainless steel. This is similar to loss by coil resistance. This system achieves a 1.8kW dynamic wireless power transfer with an automated coil detection system and a frequency control system, under factory conditions.

Digital Twin-Based Integrated Assessment of Flexible and Reconfigurable Automotive Part Production Lines

The manufacturing industry has witnessed rapid changes, including unpredictable product demand, diverse customer requirements, and increased pressure to launch new products. To deal with such changes, the reconfigurable manufacturing system has been proposed as one of the advanced manufacturing systems that is close to the realisation of smart manufacturing since it is able to reconfigure its hardware, software, and system structures in a much quicker manner. Conventional simulation technologies lack convergence with physical manufacturing systems, and reconfigurable manufacturing lines require the manual construction of production line models for each reconfiguration. This study presents a digital twin-based integrated reconfiguration assessment application that synchronises with real-time manufacturing data and provides accurate, automated simulation functionality to build and analyse a manufacturing system. The paper discusses the architectural design and implementation of the application, an information model, and an assessment model that enable quantitatively assessment on reconfigurations of manufacturing systems from various aspects. The effectiveness of the proposed application is verified via application to an automotive parts production line to assess the reconfiguration indicators of the manufacturing system under different scenarios. The results reveal that the proposed application provides faster and more accurate reconfiguration assessments compared to existing methods. The findings of this study are expected to facilitate accurate and consistent decision making for evaluating the various indicators of production line performance.

A parallel strategy for predicting the quality of welded joints in automotive bodies based on machine learning

Due to the complexity of the resistance spot welding process, it is still a challenge to accurately know the operating status of the welding robot under the current parameter settings and to assess the welding quality of electrode caps under different types of plates in real time with large data sizes. To solve this problem, this paper classifies the overall data set and proposes a parallel strategy method for predicting the quality of weld joints using machine learning for subsets of the data with different distribution patterns. Firstly, the PCA dimensionality reduction model was used to set the number of principal components to reduce the dimensionality of the welding process feature value dataset and reduce the difficulty of classifying the data subgroups, and the elbow method was used to set the number of clustering centers to complete the classification of the sub-datasets by applying the k-means model on the basis of the dimensionality reduction data. Finally, the feature parameters of each sub-dataset are used as input for machine learning, and a parallel prediction strategy for weld joint quality is developed based on the data distribution characteristics of each sub-dataset. The test results show that the model in this paper outperforms the static BP neural network in predicting the quality of all types of welded joints, the machine learning parallel strategy tailored to the characteristics of the data population works well with more complexly distributed welded big data. This paper provides accurate and effective estimation of body resistance welding condition, which can provide some guidance for online inspection of body resistance spot welding quality in automotive production lines.