Automobile Assembly Research Articles

Deep reinforcement learning with light-weight vision model for sequential robotic object sorting

Training an agent to generate cooperative joint learning policies for executing object sorting in cluttered scenes is complex. This paper presents a comparative model-free deep reinforcement learning (DRL) system built from variants of end-to-end lightweight deep neural networks that jointly complement three primitive policies (pushing, grasping, and placing) that operate from visual observations to actions. The end goal of the proposed DRL systems is to perform object sorting in very dynamic and complex environmental scenarios of cluttered and random objects (irregular and regular blocks) existing in different color variations. For this, we investigated DRL methods built from 12 custom instances of Pixel-wise Q-valued Critic Networks (PQCN) from four main backbone networks (DenseNet121, DenseNet169, MobileNetV3, and SqueezeNet) individually combined with custom fully convolutional neural networks (FCN) for learning the image-based observation space through affordance mapping paradigm while considering both dual and single transfer learning schemes. Additionally, we explored three categories of gradient-based optimization methods and considered custom reward functions with varying discount factors. The results show that the PQCN constructed by integrating DenseNet121 or MobileNetV3 with FCN and trained by a dual transfer learning scheme (with a discount factor of γ=0.70) yielded the best performance in most categories of our evaluation in the training phases. Overall, the PQCN-DenseNet121, when trained with a dual transfer learning process, yielded the best generalization (sorting success rate) in most of the evaluation metrics across all object categories in the testing phase. Furthermore, our study provides a new benchmark for validating deep reinforcement learning algorithms while sorting object blocks under varying degrees of complexity. This research finds practical relevance in industrial fields such as manufacturing and packaging, building construction, plant sorting, automobile assembly lines, and waste sorting. Here is the article’s source code: https://github.com/Emmanuel-Okafor/DRL_for_Object_Sorting.git.

Wear Reduction in Heavily Loaded Units of Transport Vehicles

Introduction. The pivot assembly provides connection between the rotating and non-rotating parts of machines and mechanisms such as cranes, excavators, trailers, railway rolling stocks. In relation to rolling stock, it connects the load-carrying part of the car with the bogie and is one of the most critical and wear-out friction units. Its technical condition affects the intensity and form of wear of the surfaces themselves, the amount of resistance to rotation of the bogie when the car moves in curved sections of the track, the amplitude of the lateral rocking of the car, the intensity of wear of the wheel flanges and, as a consequence, the safety of operation of the rolling stock as a whole. Until now, periodic grease is used for this unit, which, even with a short mileage, manages to be squeezed out of the contact zones and, thereby, creates conditions for the predominance of dry friction. Until now, various proposals to solve this problem have not found application in mass production for a number of reasons, and therefore the search for ways to reduce wear in the pivot assembly of cars is still relevant. This study proposes a solution to this problem that does not require structural changes to the pivot assembly itself. The aim of this work was to develop a technologically advanced anti-friction coating with good adhesion, which will be applied to the surface of a replaceable disk installed between the rubbing surfaces of the pivot assembly during scheduled repairs of the car chassis. This approach will reduce the friction force and wear rate in the unit when lubrication shortage occurs due to squeezing out of the grease. Materials and Methods. Antifriction characteristics of the developed coating was observed on a friction machine providing specific loads on the test sample up to 5000 N and a sliding speed from 0.13 m/s. The samples were examined by scanning electron microscopy (FEI Quanta 200 microscope). SEM images were acquired in a back-scattered electron (BSE) mode using a semiconductor detector. To analyze the elemental composition of beam samples, an energy dispersive spectrometer (EDAX Element EDS system) was used. Results. A three-layer functional phosphorus-containing composite coating of the surfaces of the unit was developed, which made it possible to significantly reduce the coefficient of friction and, as a consequence, the intensity of wear of the pivot unit surface during dry friction. The optimal conditions for obtaining composite coating layers were determined. The influence of the thickness of each layer and the conditions for its production on its functional characteristics was studied. Discussion and Conclusion. The proposed solution is manufacturable and, with appropriate adaptation, can be used to reduce wear in any open pivot assembly without radically changing its design. The methods for producing coating layers are accessible and technologically advanced for serial use.

Solute Enrichment in the Fusion Zone during Resistance Spot Welding of a Third Generation Advanced High Strength Steel

Resistance spot welding is a critical joining technique in automobile assembly. The load carrying properties of spot welds are generally accepted to correlate with weld diameter, which increases with increasing weld current or duration. The formation of a softened layer, or weld halo, surrounding the fusion zone in a spot-welded third generation (Gen3) advanced high strength steel (AHSS) was recently reported in the literature. To optimize weld performance by schedule design, it is necessary to understand the halo formation characteristics and potential impacts. Accordingly, welding of a Gen3 AHSS was performed using weld times between 130 – 1300 ms. Microhardness mapping characterized weld microhardness and the evolution of the halo during welding. Electron probe microanalysis and timeof-flight secondary ion mass spectrometry enabled measurement of solute distributions through the weld halo, while scanning electron microscopy was used for microstructural characterization. The solidified structure was examined using light-optical microscopy, and with the microhardness and compositional data, used to infer the mechanism by which the halo forms during welding. It was found that the halo develops due to solute rejection from a cellular solidification front that advances towards the center of the fusion zone while weld current is applied. Extended weld times increase the size of the weld halo and the solute content of the inner fusion zone. The decrease in weld halo microhardness and the increase in inner fusion zone microhardness is largely explained by the changes in local carbon content associated with halo formation.

A Physical Fatigue Evaluation Method for Automotive Manual Assembly: An Experiment of Cerebral Oxygenation with ARE Platform

Due to the complexity of the automobile manufacturing process, some flexible and delicate assembly work relies on manual operations. However, high-frequency and high-load repetitive operations make assembly workers prone to physical fatigue. This study proposes a method for evaluating human physical fatigue for the manual assembly of automobiles with methods: NIOSH (National Institute for Occupational Safety and Health), OWAS (Ovako Working Posture Analysis System) and RULA (Rapid Upper Limb Assessment). The cerebral oxygenation signal is selected as an objective physiological index reflecting the human fatigue level to verify the proposed physical fatigue evaluation method. Taking auto seat assembly and automobile manual assembly as an example, 18 group experiments were carried out with the ARE platform (Augmented Reality-based Ergonomic Platform). Furthermore, predictions of metabolic energy expenditure were performed for experiments in Tecnomatix Jack. Finally, it is concluded that the proposed physical fatigue evaluation method can reflect the human physical fatigue level and is more accurate than the evaluation of metabolic energy consumption in Tecnomatix Jack because of the immersion that comes with the AR devices and the precision that comes with motion capture devices.

The facility design and operation effect of control vehicle emissions hazard in test section of automobile assembly workshop

The hazard of vehicle emissions mainly come from the four wheel positioning, drum test and vehicle emissions test sections in automobile assembly workshop, which can lead to abnormal hemoglobin and hepatic insufficiency in workers. We researched on preventing toxic gases technologies for the vehicle emissions generated by these three sections, designed the ventilation facilities, and then detected and evaluated the operation effect, thereby improving the working environment, ensuring the occupational health of workers, and providing scientific basis for the control of vehicle emissions hazards.

Hierarchical learning of robotic contact policies

The paper addresses the issue of learning tasks where a robot maintains permanent contact with the environment. We propose a new methodology based on a hierarchical learning scheme coupled with task representation through directed graphs. These graphs are constituted of nodes and branches that correspond to the states and robotic actions, respectively. The upper level of the hierarchy essentially operates as a decision-making algorithm. It leverages reinforcement learning (RL) techniques to facilitate optimal decision-making. The actions are generated by a constraint-space following (CSF) controller that autonomously identifies feasible directions for motion. The controller generates robot motion by adjusting its stiffness in the direction defined by the Frenet–Serret frame, which is aligned with the robot path. The proposed framework was experimentally verified through a series of challenging robotic tasks such as maze learning, door opening, learning to shift the manual car gear, and learning car license plate light assembly by disassembly.

Industrialization in an Age of China Rising: A Tale of Two Chinese Automobile Companies in Malaysia

Abstract: This article aims to contribute to the exploration of the ways through which Malaysia—the fourth largest economy in Southeast Asia—has engaged Chinese investment to pursue industrialization. It does so by comparing and contrasting the different approaches by which two of the most important Chinese-financed automobile projects, Chery Automobile's assembly operations in 2008 and Zhejiang Geely's partial acquisition of local producer Proton in 2017, have been embedded into the Malaysian political economy. While Chery Automobile ended up with a conundrum of extremely low local sales, Zhejiang Geely has fostered mutually beneficial partnerships with Proton and its cohort of suppliers that are shaped by the industry's ethnopolitical directives. More recently, Zhejiang Geely has exported an increasingly larger proportion of vehicles from its Malaysian operations. The contrasting fate of these projects illustrates that while they are driven by Chinese money and technological know-how, they are not necessarily Chinese-dominated. It is domestic politics—the establishment of a functional politico-commercial coalition in particular—that matters in shaping the outcome and impact of Chinese investment in Malaysia. More importantly, the paper highlights how foreign direct investment and the supposed (positive and negative) externalities are intimately tied to local factors such as political institutions and ethnocentric directives.

Deep Learning Based Online Nondestructive Defect Detection for Self-Piercing Riveted Joints in Automotive Body Manufacturing

Self Piercing Riveting (SPR) is widely used for joining lightweight and dissimilar materials in automotive body manufacturing, the quality of which directly affects the safety of vehicles. However, there is still no reliable method that can be used for SPR quality control without destructive test and manual intervention. This paper presents an online non-destructive SPR defect detection method based on deep learning. By learning the temporal dependencies of punch force varying with rivet displacement under different joint combinations, the proposed method can provide real-time defect alarms and avoid the enormous cost of joint dissection. We develop an SPR parameter selection mechanism to rule out the irrelevant parameters, which enhances the learning performance. For the problem of model overfitting caused by the savage imbalance of SPR data, we design a conditional generative adversarial network based data generation model. In order to accommodate the difference in defect patterns between factory and laboratory, we devise a transfer learning based model migration method, which substantially reduces the amount of labeled factory data for model training. The evaluations on real SPR data collected from two car assembly lines of Audi and NIO verify that the proposed method achieves a high detection accuracy and a low missing rate in SPR defect detection.

Factors influencing the participation of Vietnamese enterprises in global value chains: insights from automobile production and assembly, electronics and textile industries

This study aims to examine factors influencing global value chain position of enterprises from main industry sectors of Vietnam. In order to attain such objectives, the panel data analysis was conducted based on 208 observations from 52 enterprises in the automobile production and assembly, electronics and textile industries during the 2016 to 2020 period. The multiple linear regression model is utilized to analyse the factors influencing the participation of such industry sectors in the global value chain. The results indicate that technological capability has the most significant positive influence on the participation of Vietnamese enterprises in the global value chain, followed by FDI and business size. In contrast, the factor of business age does not seem to have a discernible effect. These findings bring some policy implications on the improvement of the country's position in the global value chain, focusing on selected industries including automotive, electronics and textile.

Analisis Finite Element Aluminium Alloy 6063-T5 Pada Jig Positioning Untuk Perakitan Rooftop Interior Mobil

Tools in the production process are needed in the production process in order to accelerate and improve the quality of production, especially in the automotive industry. In this study, a jig positioning machine design will be designed which is used when assembly on the roof panel of a car so that it can help areas that are difficult to reach when doing assembly or design on the roof panel of the car. After designing the required design for the car interior roof panel assembly, it is necessary to test the design of the machine so that when the material on the machine ready to be made is already in the safe category. Testing on the machine is carried out using a material with the type of Aluminum Alloy 6035-t5 with the addition of different force factors, namely 34.3N, 39.2N, and 44.1N. The addition of the force factor is based on the maximum load given when doing the roof panel assembly on the car. The test is carried out by performing simulations using CAD Software on the arm of the jig positioning machine which will be the support when applying the load. The simulation carried out will analyze the value of material suitability factors including the value of deformation, strain-stress, and factor of safety on the jig positioning machine. The results of the simulation show that there is a change in the value of the material suitability factor due to the difference in the applied force. The suitability factor of the material on the designed jig positioning arm can accommodate the weight of the load with a maximum force of 44.1N with a maximum displacement value of 1.062e+02 mm, a stress value of 1.087e+08N/m^2, a FOS value of 1.33e+00, and a Strain value of 1 ,19e-03. Based on the simulation results of the suitability factor of the material, the jig positioning arm is within safe limits.

Technological methods of improving the quality and productivity of processing car engine parts

Problem. The most high-precision parts of the car engine are camshaft and crankshaft, parts that require two-sided processing of the end surfaces: piston pins, valve springs. Traditional processing methods are used in their manufacture and repair. Therefore, taking into account the large volumes of car production, the improvement of existing and development of new processing methods is very important. Goal. The purpose of this work is the analysis and improvement of existing methods and the implementation for new ones of processing the specified parts to increase the productivity and accuracy of their processing. Methodology. The processing of support necks and cams is expedient to be carried out in one installation during deep high-speed grinding with crossed axes of the circle and camshaft due to stabilization of the cutting depth and feed along the contour with uneven rotation of the part. The method of high-speed milling of camshaft cams with crossed axes of the tool and parts on CNC machines is also promising. Root and connecting rod necks of crankshafts can be processed in one installation by deep high-speed grinding with crossed axes of the circle and the part due to stabilization of the depth of cutting and feed along the contour with uniform rotation of the part. For end surfaces, it is possible to introduce single-pass processing on double-sided end grinding machines with oriented grinding wheels with calibration areas. Results. The obtained results indicate a 25-30% increase in productivity and processing accuracy of the specified parts. Originality. It is used when processing the effect of the intersection of the axes of the tool and the part and carrying out a special correction of the tool, which allows to obtain a calibration area. Practical value. The proposed methods improve the parameters of accuracy, roughness and productivity of processing, which is an important scientific and practical task, whose solution will allow to increase the service life of car components and assemblies, reduce the cost of their production, and raise the level of competitiveness of products and services.

Computer modeling and software research of car and engine parts

Problem. The development of automotive technology for various purposes and the improvement of existing car models, requires a fast and flexible design process. Spatial models of details and nodes are the starting points for design, and the necessary design documentation is performed on their basis. When manufacturing parts on CNC machines, it is necessary to quickly obtain a control program that will increase the accuracy of the product by ensuring the optimal trajectory of the mutual movement of the part and the tool. Therefore, the creation of a comprehensive methodology for the design of car parts, their basic inspection and obtaining design documentation is an important technical task. Goal. The main goal of the work is to determine the rational sequence and basic principles of computer-aided design, research and manufacturing of car and engine parts, and the development of mathematical spatial models of surfaces in order to optimize control programs for CNC machines. Methodology. The approaches adopted in the work for solving the set goal are based on general design principles. The methods of conducting static and dynamic calculations, as well as spatial mathematical modeling of the processes of manufacturing parts were also used. Results. A comprehensive methodology for designing parts of engines and cars has been developed. The main principles of creating spatial models have been determined in order to achieve their flexibility and simple editing. Methods of automating the development of drawings and design documentation are described. The software options for the primary verification of static strength and frequency analysis are considered. A calculation program has been developed for constructing transient and amplitude-phase-frequency characteristics. The mathematical spatial modeling of typical surfaces of the camshaft is provided with the aim of further research and optimization of the development of a control program for processing on a CNC machine tool. Originality. The defined basic principles of designing parts of engines and cars provide an opportunity to create a flexible spatial model and more fully automate the process of drawing up technical documentation. The developed mathematical spatial models of the supporting and working surfaces of the camshaft make it possible to write a control program with the determination of the optimal trajectory of the mutual movement of the part and the tool. Practical value. The obtained results can be recommended when designing car parts and assemblies.

Retracted: Research on Automobile Assembly Line Optimization Based on Industrial Engineering Technology and Machine Learning Algorithm

Retracted: Research on Automobile Assembly Line Optimization Based on Industrial Engineering Technology and Machine Learning Algorithm

An efficient automobile assembly state monitoring system based on channel-pruned YOLOv4 algorithm

ABSTRACT An efficient automobile assembly state monitoring system in industrial environment is presented in this paper. The system only needs to input a video that contains the whole detected parts and manually label it in the first frame. By finding the best point for tracking and tracking the point, the dataset can be automatically generated which saves time spent on manufacturing the dataset and makes the assembly state monitoring system easy to deploy into a practical industrial environment. The target detection algorithm uses the channel-pruned YOLOv4 neural network. The experimental result shows the algorithm balances speed and accuracy. Compared to original YOLOv4, the proposed method is two times faster and the performance is nearly equal to it. Comparative experiments show that the proposed algorithm performs better and is faster than other lightweight models which demonstrates that the channel pruning process dynamically improves the speed of the forward propagation without sacrificing accuracy. Additionally, the algorithms are deployed on two common embedded systems. The results show that in the industrial environment, the speed can fully meet real-time requirements.

Dividend Policy and Firm Performance with Moderating Effect of Ownership Structure: Evidence from The Manufacturing Firms in Pakistan

The objective of this paper is to check the impact of dividend policy and firm performance with a moderating effect of ownership structure. Many studies have been made to check the impact of dividend policy on firm performance but there is not a single study that has been made yet to check the impact of dividend policy and firm performance with moderating effect of ownership structure. This research had been carried out on different non-financial sectors of Pakistan covering the firms of automobile assembler and parts manufacturing, cement and chemical manufacturing, fertilizer manufacturing, pharmaceuticals, and textile manufacturing etc. The data used in carrying out this research was taken from the companies that was paying dividends and were listen in KSE-100 index of Pakistan Stock Exchange. Ownership data of the firms were collected by the company’s pattern of shareholding which are published or mentioned in the company’s annual report as per Securities and Exchange Commission of Pakistan (SECP) requirements that can be taken from the company’s website. The study shows significant positive relationship between dividend policy and firm performance. The study also shows that there is positive significant relationship between dividend policy and firm performance with moderating effect of ownership structure. Dividend policy can also have implications for a company's investment and growth prospects. In general, firms that retain more of their earnings for reinvestment are likely to have higher levels of capital expenditures and research and development spending. However, dividend payments can also signal to investors that a company has confidence in its future growth prospects.

Design and research on pneumatic system of tire wheel hub matching machine

With the increasingly fierce competition in the automotive industry, the requirements for the automobile wheel hub assembly line weregetting higher and higher, and the matching machine has become an indispensable part of the assembly line. In this paper, the structure of pneumatic drive and servo drive was adopted by the tire wheel hub matching machine, which ensured the production rhythm and the matching accuracy requirements. According to the overall index requirements of the automatic matching machine,the pneumatic system schematic diagram was determined by the detailed research and comparison of the principle and system composition of the representative pneumatic circuit.

Assessing the transition from mass production to lean manufacturing using a hybrid simulation model of a LEGO® automotive assembly line

PurposeWith hybrid simulation techniques getting popular for systems improvement in multiple fields, this study aims to provide insight on the use of hybrid simulation to assess the effect of lean manufacturing (LM) techniques on manufacturing facilities and the transition of a mass production (MP) facility to incorporating LM techniques.Design/methodology/approachIn this paper, the authors apply a hybrid simulation approach to improve an educational automotive assembly line and provide guidelines for implementing different LM techniques. Specifically, the authors describe the design, development, verification and validation of a hybrid discrete-event and agent-based simulation model of a LEGO® car assembly line to analyze, improve and assess the system’s performance. The simulation approach examines the base model (MP) and an alternative scenario (just-in-time [JIT] with Heijunka).FindingsThe hybrid simulation approach effectively models the facility. The alternative simulation scenario (implementing JIT and Heijunka LM techniques) improved all examined performance metrics. In more detail, the system’s lead time was reduced by 47.37%, the throughput increased by 5.99% and the work-in-progress for workstations decreased by up to 56.73%.Originality/valueThis novel hybrid simulation approach provides insight and can be potentially extrapolated to model other manufacturing facilities and evaluate transition scenarios from MP to LM.

The Manufacturing Holonic Structured Virtual Team Organization: Labour Cost Reduction?

It is an undeniable fact that business endeavours in modern times are centred on offshore operations. This scenario demands the employment of virtual teams to be seen as an essential requirement. The experiences from COVID-19 pandemic corroborate this reality. Corporations ought to revise their strategies to successfully incorporate technology as a part of their habitual actions. Holonic manufacturing emerges as a powerful communication and collaboration tool for virtual teams to optimize resources and human power. This study is based on an experiment consisting in a simulated automobile assembly industry. Holonic organizations are compared to non-holonic structures in terms of labour costs based on effective work times. The study employed a Holonic Manufacturing Structure to analyse its possible impact on labour cost reduction among several other advantages.

Self-organising multiple human–robot collaboration: A temporal subgraph reasoning-based method

Multiple Human–Robot Collaboration (HRC) requires self-organising task allocation to adapt to varying operation goals and workspace changes. However, nowadays an HRC system relies on predefined task arrangements for human and robot agents, which fails to accomplish complicated manufacturing tasks consisting of various operation sequences and different mechanical parts. To overcome the bottleneck, this paper proposes a temporal subgraph reasoning-based method for self-organising HRC task planning between multiple agents. Firstly, a tri-layer Knowledge Graph (KG) is defined to depict task-agent-operation relations in HRC tasks. Then, a subgraph mechanism is introduced to learn node embeddings from subregions of the HRC KG, which distills implicit information from local object sets. Thirdly, a temporal reasoning module is leveraged to integrate features from previous records and update the HRC KG for forecasting humans’ and robots’ subsequent operations. Finally, a car engine assembly task is demonstrated to evaluate the effectiveness of the proposed method, which outperforms other benchmarks in experimental results.

Improved Cross-tension Strength of a Friction-element-welding Joint by Tempering Treatment Using Electric Heating

Resistance spot welding is widely used in automobile assembly, but reduction of the cross-tension strength (CTS) is inevitable when using high-strength steel sheets. Therefore, we focused on joining high-strength steel sheets using friction element welding (FEW), which is mainly used to join steel sheets and aluminum alloy sheets, to improve the CTS. However, even when using FEW, the CTS decreases at the joint, when a high-carbon-content steel sheet is used as the lower sheet. This CTS decrease is presumed to be due to the low local ductility caused by the large hardness difference (ΔHn) between martensite (M) and ferrite (α) in the inter-critically annealed and quenched area during joining. The local ductility of the material having complex structure with hard and soft phases is affected by the structural morphology and ΔHn. Therefore, we attempted to improve the CTS by creating joints whose α of the inter-critically annealed and quenched area was equiaxed or acicular, and performed tempering treatment using electric heating to reduce ΔHn of this part. The results showed that ΔHn decreased at the annealed joint and the CTS improved. Furthermore, the highest CTS was observed at the joint having acicular α when tempering was performed. It was inferred that the reduction of ΔHn using electric heating improved the local ductility and CTS of the area. Furthermore, it was considered that the local ductility and CTS are also influenced by the α morphology.