Journal of Intelligent Manufacturing最新文献

{"title":"Evaluation of data augmentation and loss functions in semantic image segmentation for drilling tool wear detection","authors":"","doi":"10.1007/s10845-023-02313-y","DOIUrl":"https://doi.org/10.1007/s10845-023-02313-y","url":null,"abstract":"<h3>Abstract</h3> <p>Tool wear monitoring is crucial for quality control and cost reduction in manufacturing processes, of which drilling applications are one example. Identification of the wear area in images of cutting inserts is important to building a reliable ground truth for the development of indirect monitoring approaches. Therefore, we present a semantic image segmentation pipeline for wear detection on microscopy images of cutting inserts. A broadly used convolutional neural net, namely a U-Net, is trained with different preprocessing and optimisation task configurations: On the one hand the problem is considered as binary problem, and on the other hand as multiclass problem by differentiating the wear into two different types. By comparing these two problem formulations we investigate whether the separation of the two wear structures improves the performance of the recognition of the wear types. For both problem formulations three loss functions, i. e., Cross Entropy, Focal Cross Entropy, and a loss based on the Intersection over Union (IoU), are investigated.The use of different augmentation intensities during training suggests adequate but not too excessive augmentation, and that with optimal augmentation the choice of loss function gets less important. Furthermore, models are trained on image tiles of different sizes, which has an impact on producing artefacts on the whole image predictions performed by the overlap-tile strategy. In summary, the best performing models are binary models, trained on data with moderate augmentation and an IoU-based loss function.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"13 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-02-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139772891","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"NG-Net: No-Grasp annotation grasp detection network for stacked scenes","authors":"Min Shi, Jingzhao Hou, Zhaoxin Li, Dengming Zhu","doi":"10.1007/s10845-024-02321-6","DOIUrl":"https://doi.org/10.1007/s10845-024-02321-6","url":null,"abstract":"<p>Achieving a high grasping success rate in a stacked environment is the core of the robot’s grasping task. Most methods achieve a high grasping success rate by training the network on a dataset containing a large number of grasping annotations which requires a lot of manpower and material resources. Therefore, achieving a high grasping success rate for stacked scenes without grasping annotations is a challenging task. To address this, we propose a No-Grasp annotation grasp detection network for stacked scenes (NG-Net). Our network consists of two modules: an object selection module and a grasp generation module. Specifically, the object selection module performs instance segmentation on the raw point cloud to select the object with the highest score as the object to be grasped, and the grasp generation module uses mathematical methods to analyze the geometric features of the point cloud surface to achieve grasping pose generation without grasping annotations. Experiments show that on the modified IPA-Binpicking dataset <i>G</i>, NG-Net has an average grasp success rate of 97% in the stacked scene grasp experiment, 14–22% higher than PointNetGPD.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"17 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139755415","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Flexible dual-mode sensor with accurate contact pressure sensing and contactless distance detection functions for robotic perception","authors":"Zhijian Chen, Yancheng Wang, Zhongtan Zhang, Deqing Mei, Weijie Liu","doi":"10.1007/s10845-023-02314-x","DOIUrl":"https://doi.org/10.1007/s10845-023-02314-x","url":null,"abstract":"<p>This paper presents a novel flexible dual-mode sensor with both contact pressure and distance sensing abilities for robotic grasping and manipulation applications. The proposed flexible dual-mode sensor measures contactless distances by flat interdigitated electrodes, based on electrical field detection principle. Meanwhile the sensor detects contact pressures by truncated pyramid-shaped porous composites based on graphene nanoplate and silicone rubber. Both the functions of the sensor are encapsulated by cascading assembly, the different sensing units are nested and arranged to avoid coupling effects between different sensing signals. The structural design, working principle, and fabrication process to make the flexible dual-mode sensor were presented. Characterization tests showed that the developed flexible dual-mode sensor has a high sensitivity of 0.33 V/N and stability for contact pressure sensing, this sensor can also detect the distances between objects and sensor with high accuracy. The dual-mode sensor was then mounted onto a robotic arm to perform object’s grasping and collision experiments, results demonstrated that the sensor can accurately measure the distributed contact force and distance between objects for tactile perception. Thus, our proposed flexible dual-mode sensor would have great prospects in robotic safety detection and manipulation applications.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"13 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139677617","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-scale progressive fusion-based depth image completion and enhancement for industrial collaborative robot applications","authors":"Chuhua Xian, Jun Zhang, Wenhao Yang, Yunbo Zhang","doi":"10.1007/s10845-023-02299-7","DOIUrl":"https://doi.org/10.1007/s10845-023-02299-7","url":null,"abstract":"<p>The depth image obtained by consumer-level depth cameras generally has low resolution and missing regions due to the limitations of the depth camera hardware and the method of depth image generation. Despite the fact that many studies have been done on RGB image completion and super-resolution, a key issue with depth images is that there will be evident jagged boundaries and a significant loss of geometric information. To address these issues, we introduce a multi-scale progressive fusion network for depth image completion and super-resolution in this paper, which has an asymptotic structure for integrating hierarchical features in different domains. We employ two separate branches to learn the features of a multi-scale image given a depth image and its corresponding RGB image. The extracted features are then fused into different level features of these two branches using a step-by-step strategy to recreate the final depth image. To confine distinct borders and geometric features, a multi-dimension loss is also designed. Extensive depth completion and super-resolution studies reveal that our proposed method outperforms state-of-the-art methods both qualitatively and quantitatively. The proposed methods are also applied to two human–robot interaction applications, including a remote-controlled robot based on an unmanned ground vehicle (UGV), AR-based toolpath planning, and automatic toolpath extraction. All these experimental results indicate the effectiveness and potential benefits of the proposed methods.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"15 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139677533","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Digital twin-driven real-time suppression of delamination damage in CFRP drilling","authors":"Jielin Chen, Shuang Li, Hanwei Teng, Xiaolong Leng, Changping Li, Rendi Kurniawan, Tae Jo Ko","doi":"10.1007/s10845-023-02315-w","DOIUrl":"https://doi.org/10.1007/s10845-023-02315-w","url":null,"abstract":"<p>Delamination damage should be avoided because it severely affects the quality of CFRP products. This paper proposes a digital twin (DT) driven method for real-time suppression of delamination damage to ensure the highest quality hole exit. The relationship between the increase in thrust caused by tool wear and CFRP delamination was analyzed through extensive drilling experiments. The evolving twin models were developed to integrate the virtual space of the drilling process. Once the cutting parameters and thrust signals were input into the twin, the Gaussian process regression and mathematical models predicted the current tool wear and thrust curve, respectively. The feedback results from the DT dynamically interact with the real drilling operation after the optimization function solves the current critical feed rate (CFR). A DT scheme was designed, and the performance of the deployed DT was tested through an online service panel. The results show that the DT has excellent real-time prediction capability within 100 hole-making cycles, with maximum errors of 4.1% and 4.2% for tool wear and thrust prediction at the exit, respectively. Compared to conventional drilling (CD), DT technology provides closed-loop feedback on the time-varying CFR for each hole, resulting in no delamination mode I and up to 48.4% suppression of delamination mode III. This research has achieved intelligent virtual-real linkage in the CFRP drilling process, providing important theoretical support for effectively suppressing delamination damage in the automated production process.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"32 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139688683","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi agent reinforcement learning for online layout planning and scheduling in flexible assembly systems","authors":"","doi":"10.1007/s10845-023-02309-8","DOIUrl":"https://doi.org/10.1007/s10845-023-02309-8","url":null,"abstract":"<h3>Abstract</h3> <p>Manufacturing systems are undergoing systematic change facing the trade-off between the customer's needs and the economic and ecological pressure. Especially assembly systems must be more flexible due to many product generations or unpredictable material and demand fluctuations. As a solution line-less mobile assembly systems implement flexible job routes through movable multi-purpose resources and flexible transportation systems. Moreover, a completely reactive rearrangeable layout with mobile resources enables reconfigurations without interrupting production. A scheduling that can handle the complexity of dynamic events is necessary to plan job routes and control transportation in such an assembly system. Conventional approaches for this control task require exponentially rising computational capacities with increasing problem sizes. Therefore, the contribution of this work is an algorithm to dynamically solve the integrated problem of layout optimization and scheduling in line-less mobile assembly systems. The proposed multi agent deep reinforcement learning algorithm uses proximal policy optimization and consists of a decoder and encoder, allowing for various-sized system state descriptions. A simulation study shows that the proposed algorithm performs better in 78% of the scenarios compared to a random agent regarding the makespan optimization objective. This allows for adaptive optimization of line-less mobile assembly systems that can face global challenges.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"295 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-01-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139580905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Multi-objective sustainable production planning for a hybrid multi-stage manufacturing-remanufacturing system with grade-based classification of recovered and remanufactured products","authors":"Houria Lahmar, Mohammed Dahane, Kinza Nadia Mouss, Mohammed Haoues","doi":"10.1007/s10845-023-02308-9","DOIUrl":"https://doi.org/10.1007/s10845-023-02308-9","url":null,"abstract":"<p>In this paper, we address the problem of multi-objective production planning in a hybrid manufacturing and remanufacturing system (HMRS), introducing several significant contributions. First, we propose a new formulation of the problem that extends the existing literature by introducing a multi-objective model. This model aims to minimize both total costs and <span>(CO_2)</span> emissions within a hybrid system composed of various machines in charge of producing new and remanufactured products of different qualities. To efficiently solve this complex problem, we present an innovative approach that integrates several techniques, including NSGA-II, the entropy weight method and the TOPSIS technique. Our research focuses on the economic and environmental aspects of the remanufacturing process, seeking to determine the optimal manufacturing and remanufacturing plan. This plan aims to meet demand for new products and maximize satisfaction for remanufactured products of different qualities, while minimizing the total economic costs and <span>(CO_2)</span> emissions incurred during the various manufacturing and remanufacturing stages, including set-up, production, inventory and disposal. To address the multi-objective nature of this problem, we develop a mathematical model and introduce an approach based on the non-dominated genetic sorting algorithm (NSGA-II). To help decision-making, we use the technique of performance ranking by similarity to the ideal solution (TOPSIS) in combination with the entropy weight method (EWM) to objectively obtain the optimal compromise solution from the Pareto front provided by NSGA-II. Finally, we conduct computational experiments to assess the environmental impact of carbon emissions associated with new, remanufactured and discarded products over a finite production horizon. We illustrate the adaptability of the proposed approach by applying it to two distinct remanufacturing strategies: one where remanufacturing is used to reduce waste, and one where demand for remanufactured products is critical, with a penalty cost associated with any shortfall in demand.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"122 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-01-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139560451","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Causal deep learning for explainable vision-based quality inspection under visual interference","authors":"Tianbiao Liang, Tianyuan Liu, Junliang Wang, Jie Zhang, Pai Zheng","doi":"10.1007/s10845-023-02297-9","DOIUrl":"https://doi.org/10.1007/s10845-023-02297-9","url":null,"abstract":"<p>Vision-based quality inspection is a key step to ensure the quality control of complex industrial products. However, accurate defect recognition for complex products with information-rich, structure-irregular and significantly different patterns is still a tough problem, since it causes the strong visual interference. This paper proposes a causal deep learning method (CDLM) to tackle the explainable vision-based quality inspection under visual interference. First, a structural causal model for defect recognition of complex industrial products is constructed and a causal intervention strategy to overcome the background interference is generated. Second, a defect-guided recognition neural network (DGRNN) is constructed, which can realize accurate defect recognition under the training of CDLM via feature-wise causal intervention using two sub-networks with feature difference mechanism. Finally, the causality between defect features and defective product labels can guide the DGRNN to complete the accurate and explainable learning of defect in a causal direction of optimization. Quantitative experiments show that the proposed method achieves recognition accuracy of 94.09% and 93.95% on two fabric datasets respectively, which outperforms the cutting-edge inspection models. Besides, Grad-CAM visualization experiments show that the proposed method successfully captures the data causality and realizes the explainable defect recognition.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"9 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139560457","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Automated porosity segmentation in laser powder bed fusion part using computed tomography: a validity study","authors":"Catherine Desrosiers, Morgan Letenneur, Fabrice Bernier, Nicolas Piché, Benjamin Provencher, Farida Cheriet, François Guibault, Vladimir Brailovski","doi":"10.1007/s10845-023-02296-w","DOIUrl":"https://doi.org/10.1007/s10845-023-02296-w","url":null,"abstract":"<p>Defect detection in laser powder bed fusion (LPBF) parts is a critical step for in their quality control. Ensuring the integrity of these parts is essential for a broader adoption of this manufacturing process in highly standardized industries such as aerospace. With many challenges to overcome, there is currently no standardized image analysis and segmentation process for the defect analysis of LPBF parts. This process is often manual and operator-dependent, which limits the repeatability and the reproducibility of the analytical methods applied, raising questions about the validity of the analysis. The pore segmentation step is critical for porosity analysis since the pore size and morphology metrics are calculated directly from the results of the segmentation process. In this work, Ti6Al4V specimens with purposely induced and controlled porosity were printed, scanned 5 times on two CT scan systems by two different operators, and then reconstructed as 3D volumes. The porosity in these specimens was analyzed using manual and Otsu thresholding and a convolutional neural network (CNN) deep learning segmentation algorithm. Then, a variance component estimation realized over 75 porosity analyses indicated that, independently of the operator and the CT scan system used, the CNN provided the best repeatability and reproducibility in the LPBF specimens of this study. Finally, a multimodal correlative study using higher resolution laser confocal microscopy observations was used for a multi-scale pore-to-pore comparison and as a reliability assessment of the segmentation algorithms. The validity of the CNN-based pore segmentation was thus assessed through improved repeatability, reproducibility, and reliability.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"66 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139510010","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Stacked encoded cascade error feedback deep extreme learning machine network for manufacturing order completion time","authors":"Waqar Ahmed Khan, Mahmoud Masoud, Abdelrahman E. E. Eltoukhy, Mehran Ullah","doi":"10.1007/s10845-023-02303-0","DOIUrl":"https://doi.org/10.1007/s10845-023-02303-0","url":null,"abstract":"<p>In this paper, a novel stacked encoded cascade error feedback deep extreme learning machine (SEC-E-DELM) network is proposed to predict order completion time (OCT) considering the historical production planning and control data. Usually, the actual OCT significantly deviates from the planned because of recessive disturbances. The disturbances do not shut down production but slow down the production that accumulates over time, causing significant deviation of actual time from planned. The generation of weight parameters in neural networks using a randomization approach has a significant effect on generalization performance. To predict the OCT, firstly, the stacked autoencoder is used to generate input connection weights for the network by learning a deep representation of the real data. Secondly, the learned distribution of the encoder is connected to the network output through output connection weights incrementally learned by the Moore–Penrose inverse. Thirdly, the new hidden unit is added one by one to the network, which receives input connections from the input units and the last layer of the encoder to avoid overfitting and improve model generalization. The input connection weights for the newly added hidden unit are analytically calculated by the error feedback function to enhance the convergence rate by further learning deep features. Lastly, the hidden unit keeps on adding one by one by receiving connections from input units and some of the existing hidden units to make a deep cascade architecture. An extensive comparative study demonstrates that calculating connection weights by the proposed method helps to significantly improve the generalization performance and robustness of the network.</p>","PeriodicalId":16193,"journal":{"name":"Journal of Intelligent Manufacturing","volume":"7 1","pages":""},"PeriodicalIF":8.3,"publicationDate":"2024-01-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139510048","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}