Multifeature, Sparse-Based Approach for Defects Detection and Classification in Semiconductor Units

Haddad, Boualem; Yang, Sen; Karam, Lina J.; Ye, Jieping; Patel, Neil; Braun, Martin

doi:10.1109/tase.2016.2594288

Cited by 37 publications

(17 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Their proposed method could discover new failure map patterns, identify the cause, and resolve them quickly. Haddad et al [47] studied a defect detection and classification problem with the very small sample size of available data in semiconductor units. They developed a defect characterization framework, including feature extraction, feature subtraction, and sparse code generation, to identify defect patterns.…”

Section: B Applying Big Data and Data Mining To Enhance Product Yieldmentioning

confidence: 99%

A Data-Driven Approach for Identifying Possible Manufacturing Processes and Production Parameters That Cause Product Defects: A Thin-Film Filter Company Case Study

2020

View full text Add to dashboard Cite

A semiconductor or photoelectric manufacturer faces a more competitive market with small quantities of many products. These products require hundreds of processes for production, thereby generating huge manufacturing data. With the help of the Internet of Things (IoT) technology, the manufacturer can collect manufacturing process data in a timely manner. Due to the massive quantities of manufacturing process data, it has become difficult for manufacturers to determine the causes of product defects, by which machine, and by what manufacturing process (or recipe) parameters. This research proposes a six-step datadriven solution to this problem. The chi-square test of independence, the Apriori algorithm, and the decision tree method identify the process that is generating the defective products and extract rules to identify the lot identification of product defects and their associated manufacturing process parameters. An empirical study was conducted at an optical thin-film filter (TFF) company in Taiwan. Based on the data of the optical TFF production lines, the coating process was identified as the source of the defective products, and the extracted rules were validated and implemented. The product defect rate decreased from 20% to 5%. Hence, the proposed data-driven solution was found to be capable of helping manufacturers enhance their product yield. INDEX TERMS Data-driven, data mining, process improvement, association rule mining, decision tree, thin-film filter (TFF).

show abstract

Section: B Applying Big Data and Data Mining To Enhance Product Yieldmentioning

confidence: 99%

A Data-Driven Approach for Identifying Possible Manufacturing Processes and Production Parameters That Cause Product Defects: A Thin-Film Filter Company Case Study

2020

View full text Add to dashboard Cite

show abstract

“…The training set is used to train the predictive models, and the test set, modified to include the new process behaviour, is used to evaluate their performance. In order to assess the capacity of each dynamic sampling algorithm to detect the new behaviour, we employ two different approaches based on control limits defined on the training dataset, which will be referred to as fault detection and anomaly detection [30]:…”

Section: Dealing With New Process Behaviourmentioning

confidence: 99%

Induced Start Dynamic Sampling for Wafer Metrology Optimization

Susto

Maggipinto

Zocco

et al. 2020

IEEE Trans. Automat. Sci. Eng.

View full text Add to dashboard Cite

Metrology, which plays an important role in ensuring production quality in modern manufacturing industries, incurs substantial costs, both in terms of the infrastructure required, and the time needed to perform measurements. In particular, in the semiconductor manufacturing industry, measuring fundamental quantities on different sites of a wafer surface is associated with increased production time. To increase metrology efficiency, a typical strategy is to limit the number of sites measured and to exploit statistical models (soft sensing) to reconstruct the wafer profile. Moreover, for quality reasons, spatial dynamic sampling strategies may be employed to ensure that all regions of a wafer surface are checked periodically during production. In this work, we propose a new sampling strategy, called Induced Start Dynamic Sampling (ISDS), that adapts greedy feature selection algorithms to the spatial dynamic sampling problem such that the number of measured sites at each process run is minimized while achieving good wafer profile reconstruction accuracy and process visibility. The superiority of the proposed strategy with respect to the state-of-the-art is demonstrated using both simulated data and an industrial chemical vapour deposition case study. Note To Practitioners-In this work we tackle a practical metrology problem encountered in semiconductor manufacturing, namely, the design of a dynamic wafer measurement plan to monitor the accuracy of a process across the whole wafer surface. The measurement plan is called 'dynamic', since, the measurement locations, which are drawn from a candidate set that provides coverage of the whole wafer, change at each process iteration. Our methodology addresses the challenge of finding an optimized trade-off between the number of measurements performed on each wafer and the reconstruction accuracy that can be achieved for the unmeasured areas on the wafer, while at the same time, for quality assurance purposes, ensuring that all locations on a wafer are visited in a finite number of process runs. The major benefit of the methodology is that it can significantly reduce the number of sites that need to be measured on each wafer enabling greater throughput on metrology tools without sacrificing process monitoring and anomaly detection capability.

show abstract

“…Unlike bagging and Ada-Boost, stacked generalization (SG) [24] combines the yield which is produced by various base learners in the first level, and then, by utilizing a metalearner, it tries to combine the outcomes from these base learners in an ideal method to augment the generalization ability. It is well known that SG has been applied in the manufacturing areas, e.g., scheduling of flexible manufacturing systems [25], defect detection and classification in semiconductor units [26], and engine RUL prediction [27]. However, there are very few examples of SG applied to recognize tool wear states.…”

Section: Introductionmentioning

confidence: 99%

Milling Tool Wear State Recognition by Vibration Signal Using a Stacked Generalization Ensemble Model

Yang

Mei

Jiang

et al. 2019

Shock and Vibration

View full text Add to dashboard Cite

Milling tool wear state recognition plays an important role in controlling the quality of milled parts and reducing machine tool downtime. However, the characteristics of milling process limit the accuracy and stability of tool condition monitoring (TCM) employing vibration signals. To improve this problem, this paper explores the use of vibration signals as sensing approach for recognizing tool wear states during milling operation by using the stacked generalization (SG) ensemble model. In this study, vibration signals collected during the milling process are analyzed through the time domain, frequency domain, and time-frequency domain to extract signal features. The support vector machine recursive feature elimination (SVM-RFE) algorithm is used to select the main features which are most relevant to tool wear states. The SG ensemble model based on SVM, decision tree (DT), naive Bayes (NB), and SG ensemble strategy is constructed to recognize tool wear states. The proposed method is experimental verified, and the results show that the recognition accuracy of the established SG ensemble model is 98.74% and the overall G-mean and AUC evaluation value of the model is 0.98 and 0.98, respectively. In addition, compared with other ensemble models and single models, the SG ensemble model based on vibration signals has better recognition accuracy and stability than other models.

show abstract

Multifeature, Sparse-Based Approach for Defects Detection and Classification in Semiconductor Units

Cited by 37 publications

References 43 publications

A Data-Driven Approach for Identifying Possible Manufacturing Processes and Production Parameters That Cause Product Defects: A Thin-Film Filter Company Case Study

A Data-Driven Approach for Identifying Possible Manufacturing Processes and Production Parameters That Cause Product Defects: A Thin-Film Filter Company Case Study

Induced Start Dynamic Sampling for Wafer Metrology Optimization

Milling Tool Wear State Recognition by Vibration Signal Using a Stacked Generalization Ensemble Model

Contact Info

Product

Resources

About