Technical PaperIdentifying manufacturing operational conditions by physics-based feature extraction and ensemble clustering
Introduction
In manufacturing, the condition of production process is in one of its operational conditions. Operational condition (OC) is defined as the health status of system/tool/equipment [1], [2], [3], [4], [5], [6]. A process may have varying OCs due to scheduled/unexpected changes in tool conditions, system configurations, production environments, etc. Correctly and timely identifying the OC of a process enables prevention and quick maintenance reactions to tool wear, machine faults and process degradations.
Traditionally, OC identification can be achieved by model-based approaches [7]. Methods of this kind target reliability and safety issues in industrial processes and build systems for fault diagnosis. Yet, the complicated model design and restrictive mathematical assumptions have compromised their feasibility in real applications. As a contrast, data-driven methods for OC identification are favorable alternatives with the potential of real-time detection of production issues. They handle complicated data types efficiently and generate readily interpretable and comparable results. In today's advanced manufacturing, a huge volume of data can be generated by smart sensors continuously. The resulted data-rich environment lays the foundation for data-driven OC identification.
However, most data-driven OC identification methods are supervised, requiring prior knowledge about the OCs. “Labels” indicating the OCs of the system are needed for model training and validation. Popular data-driven OC identification approaches include control chart, classification analysis, Principal Component Analysis (PCA), Partial Least Squares (PLS), deep learning [8], etc. They all require prior knowledge about OCs for method initialization and implementation.
Unfortunately, in most manufacturing setting, labels about OCs are not readily available at the time of production. For example, a retrospective analysis of maintenance records may reveal process degradation leading up to the failure, but the degradation may not be directly translated to OC labels, leaving the in-line collected data as unlabeled. It's also possible to have partially labeled data. In the hot stamping case, periodic inspection for die wear/equipment damage provides labels for thermal images leading up to the inspection date. Yet, there are numerous thermal images not matched to any inspection dates, resulting in many unlabeled thermal images. Furthermore, production data can be unlabeled due to physical constraints. In laser-based additive manufacturing (LBAM), for example, out-of-control OCs are reflected in part porosity, which is a microstructural issue that cannot be measured precisely with current technology [9]. Therefore, accurate, creditable labels for thermal images from LBAM rarely exist. In manufacturing, in-situ thermal images captured by infrared (IR) camera or pyrometer reflect tool wear, equipment damage, and process degradation. This study aims to develop an unsupervised learning approach based on clustering analysis for data-driven OC identification with unlabeled thermal images.
There are two unique challenges in identifying OCs from thermal images. First, thermal images increase the complexity of feature extraction (FE). According to whether or not labels are required, FE methods can be divided into “supervised” and “unsupervised” FE. Supervised FE is applied to labeled data, with the goal to minimize the distance between samples belonging to the same class [10]. Automated FE with deep learning [6] is a typical practice of this category. However, supervised FE is infeasible for the in-situ thermal images in our study, as our data are unlabeled, and we do not know the ground-truth classes. Unsupervised FE requires no labeled samples. It finds an optimal projection to preserve crucial information [10]. Unsupervised FE has been explored in recent literature about thermal images-based process monitoring. Sun et al. [11] and Li et al. [12] formed features by combining pixels at certain locations in thermal image; Grasso et al. [13] extracted Region of Interest (ROI) by image thresholding; Liu et al. [14] capture image textural information using the gray level co-occurrence matrix. Wang et al. [15] detected the edge lines and performed Hough transform to map edge points to Hough space. Tootoonia et al. [16] converted the image into an unweighted and undirected network graph. Khanzadeh et al. [17] obtained heat-affected zone (HAZ) boundary characteristics with Functional Principal Component (FPCA). It is noted that most of these studies leveraged the physical meanings of HAZs in their FE, instead of applying pure image processing techniques on thermal images as in conventional FE. Such physics-based FE approaches characterize the object from more explainable and intuitive ways. A limitation of the above works, however, is that they emphasized either the object shape or temperature but not both, so the extracted features may not be sufficiently informative for OC identification. It is of strong interest to have an unsupervised FE method that thoroughly characterizes the physical behavior of HAZs in thermal images.
The second challenge is a lack of consistency and robustness when using clustering analysis for OC identification. The major branches of clustering analysis include centroid-based approach (e.g., -means), connectivity-based approach (e.g., hierarchical clustering), density-based approach (e.g., DBSCAN, OPTICS), condensation-based approach (e.g., BIRCH), grid-based approach (e.g., WaveCluster), model-based approach (e.g., K clusters), and randomized search approach (e.g., CLARANS) [18], [19], [20]. Each branch of methods targets one critical aspect. However, none of them are robust across all kinds of data. Depending on a single clustering analysis for OC identification can be biased. To improve consistency and robustness, ensemble clustering (EC) is promoted in this study, which combines multiple clustering models/results to produce more consistent and robust analysis [21]. The major steps in EC include base (individual clustering) selection and ensemble with consensus function [22]. There have been numerous works on EC, e.g., robust clustering ensemble (RCESCC) [23], multiple -means clustering [24], ultra-scalable ensemble clustering (U-SENC) [25], self-paced clustering ensemble (SPCE) [26], cluster-level weighting [22], etc. However, these works mostly relies on certain knowledge of the ground truth, e.g., a handful of labeled instances, for base selection or consensus function, thus are infeasible when no information about the ground truth is available. Besides, they lack a domain-knowledge-guided mapping mechanism from the EC results to OCs. There is a research gap for a practical, unsupervised method for discovering OCs from the HAZ features.
Targeting the above challenges, this study aims to develop an unsupervised learning framework that extracts informative features from thermal images, assigns labels using EC and associate cluster labels to OCs based on empirical evidence from the manufacturing plant. In the proposed framework, two most practically meaningful perspectives, part temperature and shape, are consistently considered to ensure the correspondence between statistically formed clusters and the OCs identified. According to physical properties, FE characterizes the temperature magnitude and shape of a part's HAZ. EC is adopted to integrate individual clustering results for comprehensive, informative analysis. Specifically, -means [20], Ordering Points to Identify Cluster Structure (OPTICS) [27] and HC with complete linkage [20] are respectively applied on temperature profile and shape features using varying model parameters, generating a large group of clustering outcomes. These results are fused with our proposed EC approach to create ensemble clusters. Physics of the part such as the heat allocation and distribution is leveraged to map the ensemble clusters to OCs and then further categorize the OCs into “normal”, “warning”, and “abnormal” working conditions. The three stages, i.e., from features to ensemble clusters, from ensemble clusters to OCs, and from OCs to working conditions, constitute the proposed framework for OC identification.
The remainder of the paper is organized as follows. Section 2 will introduce the data to be used in case study. Section 3 will elaborate the proposed framework with technical details, followed by a case study in Section 4 to demonstrate the generality and effectiveness of the proposed framework. Section 5 will end the paper with conclusion and future directions.
Section snippets
Thermal images from hot stamping
Hot stamping is a forming technique of a thinner flat part that is consisted of approach phase, part forming and part quenching [28]. It is intensively used in manufacturing automotive parts [28], [29]. During hot stamping, a metal blank is heated towards 900 ∼ 950 °C, during 5 ∼ 10 min, in a furnace until complete austenitization. The hot sheet is then transferred quickly from the furnace to the press and immediately stamped, with the stamping tool held closed for about 20 seconds to quench
Method
This section elaborates the technical details of the proposed framework. An offline OC identification framework is formulated by three crucial components: (1) feature extraction for HAZ, (2) EC for image labeling, and (3) building mappings between clustering results and OCs. An online OC identification algorithm is developed on top of the proposed framework to facilitate in-situ process monitoring with unsupervised thermal images.
Results on OC Identification for Type 2 Parts
To verify feasibility and generality, we implement the proposed framework on the 594 Type 2 part thermal images for offline OC identification. Preliminary inspection shows that Type 2 part has a lower temperature magnitude in general – no HAZs exceeding 200°C. Relatively low and should better preserve and reveal the HAZ characteristics than large ones. Correspondingly, we use , in feature extraction for this case study.
Individual clustering with the same
Conclusion and future work
In this study, an OC identification framework for thermal images has been proposed based on unsupervised feature extraction and EC. Three key components constitute the framework: HAZ feature extraction, EC with OPTICS, and domain-knowledge-informed OC identification. They have respectively enabled precise characterization of HAZ temperature profile, comprehensive thermal image clustering, and solid mappings from cluster labels to OCs. In addition, an online extension has been incorporated into
Declaration of Competing Interest
The authors report no declarations of interest.
Acknowledgment
The authors would like to thank Ford Motor Company for providing the data and related domain knowledge.
References (38)
- et al.
Deep learning for smart manufacturing: Methods and applications
J Manufact Syst
(2018) - et al.
In situ monitoring of selective laser melting of zinc powder via infrared imaging of the process plume
Robot. Comput-Integrated Manuf
(2018) - et al.
Image analysis-based closed loop quality control for additive manufacturing with fused filament fabrication
J Manuf Syst
(2019) - et al.
Machine vision intelligence for product defect inspection based on deep learning and hough transform
J Manuf Syst
(2019) - et al.
Online non-contact surface finish measurement in machining using graph theory-based image analysis
J Manuf Syst
(2016) - et al.
Porosity prediction: Supervised-learning of thermal history for direct laser deposition
J Manuf Syst
(2018) - et al.
A review of clustering techniques and developments
Neurocomputing
(2017) - et al.
Experimental and theoretical study of thermal aspects of the hot stamping process
Appl Thermal Eng
(2011) - et al.
Application of hot press forming process to manufacture an automotive part and its finite element analysis considering phase transformation plasticity
Int J Mech
(2009) - et al.
An efficient and scalable family of algorithms for combining clusterings
Eng Appl Artif Intel
(2013)
Silhouettes: a graphical aid to the interpretation and validation of cluster analysis
J Comput Appl Math
Preventive maintenance of a batch production system under time-varying operational condition
Int J Prod Res
The real-time multiple operational condition monitoring of ethylene cracking furnace based on the principal component analysis
Chem Eng Trans
Tool and process condition monitoring
Mechanik
Predictive process control for change in chemical dry etching equipment condition
in: 2018 International Symposium on Semiconductor Manufacturing (ISSM)
Three-tier modular structural health monitoring framework using environmental and operational condition clustering for data normalization: Validation on an operational wind turbine system
Proc IEEE
Deep learning for infrared thermal image based machine health monitoring
IEEE/ASME Trans Mechatr
A review on basic data-driven approaches for industrial process monitoring
IEEE Trans Ind Electr
Hierarchical spatial-temporal modeling and monitoring of melt pool evolution in laser-based additive manufacturing
IISE Trans
Cited by (6)
Approach to derive golden paths under time-varying machine performance in multistage manufacturing process
2021, Journal of Manufacturing SystemsCitation Excerpt :Such developed HIs are utilized to simplify machine degradation and accurately predict future health state of the machine [12,14]. In terms of the MMP, most existing studies related to HI development and prognostics have a common limitation in that they focus on an individual machine [15–17]. Relatively limited attention is paid to the entire manufacturing process-level prognostics [18–20].
Data-driven dynamic bottleneck detection in complex manufacturing systems
2021, Journal of Manufacturing SystemsCitation Excerpt :Its working principles should be intuitive for industrial engineers and line experts to understand, so that with the correct understanding of the methodology, domain experts can make adjustment when necessary based on local line configurations. More research attention in data-driven modeling for manufacturing systems in recent years has been attracted to include such domain knowledge and physical understanding to generate models that are more robust and explainable [13–16]. Therefore, the goal of this paper is to improve the current data-driven dynamic bottleneck detection methods, “the Turning Point Method”, so that it adapts to a wider variety of industrial production systems.
Knowledge distillation-based information sharing for online process monitoring in decentralized manufacturing system
2024, Journal of Intelligent ManufacturingMachine Learning in Manufacturing towards Industry 4.0: From ‘For Now’ to ‘Four-Know’
2023, Applied Sciences (Switzerland)An ensemble clustering method based on consistency cluster consensus approach and MapReduce model
2023, Transactions on Emerging Telecommunications TechnologiesUIR-Net: Object Detection in Infrared Imaging of Thermomechanical Processes in Automotive Manufacturing
2022, IEEE Transactions on Automation Science and Engineering