Exploring Machine Learning for Semiconductor Process Optimization: A Systematic Review

Advanced Process Control (APC) is an important research area in Semiconductor Manufacturing (SM) to improve process stability crucial for product quality. In low-volume-high-mixture fabrication plants (fabs), Knowledge Discovery in Databases is extremely challenging due to complex technology mixtures and reduced availability of data for comparable process steps. High Density Plasma Chemical Vapor Deposition (HDP CVD) appears to be a process area in SM predestinated for application of Data Mining (DM). Enhancing physical metrology by predictive models leads to smart future fabs. Actual research focuses on Virtual Metrology (VM) using high sophisticated Machine Learning (ML) methods to model unknown functional interrelations and to predict the thickness of dielectric layers deposited onto a metallization layer of the manufactured wafers. Decision Trees (DT), Neural Networks (NN) and Support Vector Regression (SVR) have been investigated to maximize the accuracy of the regression. For data of various logistical granularities promising results have been achieved by implementing these statistical models.

Due to diversity and many possibilities for data collection in semiconductor manufacturing, various complex machine learning approaches exist for different process steps. However, a systematic overview of these approaches is missing. This study, therefore, systematically reviews machine learning applications for process monitoring and control in semiconductor manufacturing based on peer-reviewed literature. To structure the review, we use the wafer fabrication plant-wide framework for process monitoring and control and the framework of continuous process improvement based on machine learning technique. We identify respective application areas and future research needs of machine learning for process monitoring and control in semiconductor manufactnring.

Decision-based virtual metrology for advanced process control to empower smart production and an empirical study for semiconductor manufacturing

This paper proposes a novel highly accurate virtual metrology (VM) method based on adaptive online time-series learning. The method comprises a novel time-series prediction algorithm robust against data drifts and shifts observed in semi-conductor manufacturing. The prediction algorithm incorporates time-aware data normalization and adaptive online learning. The time-aware normalizer transforms the data to suppress the effect of drifts. The adaptive online learner captures the time-varying relationship between response and inputs caused by shifts. On both simulated and real process data our approach outperforms conventional VM approaches. When applied to advanced process control (APC) on high-volume production lines at a major semiconductor manufacturer, our VM method substantially reduced variability of process outcomes.

With the establishment of machine learning (ML) techniques in the scientific community, the construction of ML potential energy surfaces (ML-PES) has become a standard process in physics and chemistry. So far, improvements in the construction of ML-PES models have been conducted independently, creating an initial hurdle for new users to overcome and complicating the reproducibility of results. Aiming to reduce the bar for the extensive use of ML-PES, we introduce Asparagus, a software package encompassing the different parts into one coherent implementation that allows an autonomous, user-guided construction of ML-PES models. Asparagus combines capabilities of initial data sampling with interfaces to ab initio calculation programs, ML model training, as well as model evaluation and its application within other codes such as ASE or CHARMM. The functionalities of the code are illustrated in different examples, including the dynamics of small molecules, the representation of reactive potentials in organometallic compounds, and atom diffusion on periodic surface structures. The modular framework of Asparagus is designed to allow simple implementations of further ML-related methods and models to provide constant user-friendly access to state-of-the-art ML techniques. Program summaryProgram Title:AsparagusCPC Library link to program files:https://doi.org/10.17632/9w9xw7mp2h.1Developer's repository link:https://github.com/MMunibas/AsparagusLicensing provisions: MITProgramming language: PythonSupplementary material: Access to Documentation at https://asparagus-bundle.readthedocs.ioNature of problem: Constructing machine-learning (ML) based potential energy surfaces (PESs) for atomistic simulations is a multi-step process that requires a broad knowledge in quantum chemistry, nuclear dynamics and programming. So far, efforts mainly focused on developing and improving ML model architectures. However, there was less effort spent on providing tools for consistent and reproducible workflows that support the construction of ML-PES for a variety of chemical systems for the broader science community.Solution method:Asparagus is a program package written in Python that provides a streamlined and extensible workflow with a user-friendly command structure to support the construction of ML-PESs. This is achieved by bundling and linking data generation and sampling techniques, data management, model training, testing and evaluation tools into one modular, comprehensive workflow including interfaces to other simulation packages for the application of the ML-PESs. By lowering the entrance barriers especially for new users, Asparagus supports the generation and adjustment of ML-PESs that allow an increased focus on the physico-chemical evaluation of the chemical system or application in molecular dynamics simulations.Additional comments including restrictions and unusual features:Asparagus is a modular package written in Python providing an underlying structure for further extensions and maintenance. Currently, methods based on message-passing neural network (NN) models using the PyTorch Python package are available. Additions of new models and interfaces to already implemented modules are possible. The NN architecture and hyperparameters are stored in a global configuration module and as a json file for documentation. Except for essential input information, default input parameters are used if not specifically defined otherwise, which allows a quick setup for the construction of a ML-PES but also the fine-tuning for specific needs.

Chemical heuristics have been fundamental to the advancement of chemistry and materials science. These heuristics are typically established by scientists using knowledge and creativity to extract patterns from limited datasets. Machine learning offers opportunities to perfect this approach using computers and larger datasets. Here, we discuss the relationships between traditional heuristics and machine learning approaches. We show how traditional rules can be challenged by large-scale statistical assessment and how traditional concepts commonly used as features are feeding the machine learning techniques. We stress the waste involved in relearning chemical rules and the challenges in terms of data size requirements for purely data-driven approaches. Our view is that heuristic and machine learning approaches are at their best when they work together.

A comprehensive review on detection of plant disease using machine learning and deep learning approaches

With the popularization of machine learning (ML) techniques and the increased chipset’s performance, the application of ML to pedestrian localization systems has received significant attention in the last years. Several survey papers have attempted to provide a state-of-the-art overview, but they usually limit their scope to a particular type of positioning system or technology. In addition, they are written from the point of view of ML techniques and their practice, not from the point of view of the localization system and the specific problems that ML techniques can help to solve. This article is intended to offer a comprehensive state-of-the-art survey of the ML techniques that have been adopted over the last ten years to improve the performance of pedestrian localization systems, addressing the applicability of ML techniques in this domain, along with the main localization strategies. It concludes by indicating the underlying open issues and challenges associated with the existing systems, and possible future directions in which ML techniques could improve the performance of pedestrian localization systems. Among other open issues, most previous authors have focused their attention on position estimation accuracy, which wastes the potential of ML techniques to improve other performance parameters (e.g., response time, computational complexity, robustness, scalability or energy efficiency). This study shows that there is a strong trend towards the application of supervised learning. Consequently, there are many potential research opportunities in the use of other learning types, such as unsupervised and reinforcement learning, to improve the performance of pedestrian localization systems.

Landslide disaster risk reduction necessitates the investigation of different geotechnical causal factors for slope failures. Machine learning (ML) techniques have been proposed to study causal factors across many application areas. However, the development of ensemble ML techniques for identifying the geotechnical causal factors for slope failures and their subsequent prediction has lacked in literature. The primary goal of this research is to develop and evaluate novel feature selection methods for identifying causal factors for slope failures and assess the potential of ensemble and individual ML techniques for slope failure prediction. Twenty-one geotechnical causal factors were obtained from 60 sites (both landslide and non-landslide) spread across a landslide-prone area in Mandi, India. Relevant causal factors were evaluated by developing a novel ensemble feature selection method that involved an average of different individual feature selection methods like correlation, information-gain, gain-ratio, OneR, and F-ratio. Furthermore, different ensemble ML techniques (Random Forest (RF), AdaBoost (AB), Bagging, Stacking, and Voting) and individual ML techniques (Bayesian network (BN), decision tree (DT), multilayer perceptron (MLP), and support vector machine (SVM)) were calibrated to 70% of the locations and tested on 30% of the sites. The ensemble feature selection method yielded six major contributing parameters to slope failures: relative compaction, porosity, saturated permeability, slope angle, angle of the internal friction, and in-situ moisture content. Furthermore, the ensemble RF and AB techniques performed the best compared to other ensemble and individual ML techniques on test data. The present study discusses the implications of different causal factors for slope failure prediction.

Machine Learning (ML) techniques are pervasive tools in various emerging commercial applications, but have to be accommodated by powerful computer systems to process very large data. Although general-purpose CPUs and GPUs have provided straightforward solutions, their energy-efficiencies are limited due to their excessive supports for flexibility. Hardware accelerators may achieve better energy-efficiencies, but each accelerator often accommodates only a single ML technique (family). According to the famous No-Free-Lunch theorem in the ML domain, however, an ML technique performs well on a dataset may perform poorly on another dataset, which implies that such accelerator may sometimes lead to poor learning accuracy. Even if regardless of the learning accuracy, such accelerator can still become inapplicable simply because the concrete ML task is altered, or the user chooses another ML technique.In this study, we present an ML accelerator called PuDianNao, which accommodates seven representative ML techniques, including k-means, k-nearest neighbors, naive bayes, support vector machine, linear regression, classification tree, and deep neural network. Benefited from our thorough analysis on computational primitives and locality properties of different ML techniques, PuDianNao can perform up to 1056 GOP/s (e.g., additions and multiplications) in an area of 3.51 mm^2, and consumes 596 mW only. Compared with the NVIDIA K20M GPU (28nm process), PuDianNao (65nm process) is 1.20x faster, and can reduce the energy by 128.41x.

Machine learning (ML) techniques have numerous applications in many fields, including healthcare, medicine, finance, marketing, and cyber security. For example, ML techniques are being applied to determine whether to give a loan to a customer or whether the computing system has been attacked. However, the ML techniques themselves may be subject to attacks and may discriminate when determining who should get the loan. Therefore, the ML techniques have to be secure, ensure privacy of the individuals, incorporate fairness and be accurate. Such collection of ML techniques has come to be known as trustworthy machine learning (trustworthy ML). This article describes an architecture to support scalable trustworthy ML and describes the features that have to be incorporated into the ML techniques to ensure that they are trustworthy.

Background: E-beam metrologies, both critical dimension scanning electron microscope metrology and defect scan metrology, have been playing a very critical role in gating patterning quality. SEM images can provide rich visual information for engineers to do qualitative and quantitative analyses. However, the low e-beam metrology tool throughput makes it impossible to obtain SEM images for larger area. Monte Carlo-based SEM image simulations or other SEM image simulations require postlithography or postetch pattern three-dimensional structures as prerequisite, and the simulation speed is not sufficiently fast for full chip implementation. Aim: We aim to develop machine learning SEM models with sufficient accuracy and speed for full chip application in semiconductor manufacturing environment. Approach: We have proposed a virtual SEM metrology solution based on U-Net neural network with physics-based feature maps as model input. With information in aerial image space encoded properly, SEM images of both postlithography and postetch can be predicted accurately enough for practical applications using our proposed virtual SEM metrology models. Equipped with GPUs, the machine learning-based SEM image models are fast enough to make it possible to realize full chip SEM generation from post-OPC data. Results: Our machine learning SEM image models can predict SEM images with normalized cross correlation around 0.95 in reference to ground truth SEM images, each SEM image (512 × 512 in size) takes about 800 ms using single CPU, and the speed can be accelerated to about 10 ms with single GPU. Conclusions: Using U-Net structure and physics-based feature maps as model inputs, machine learning-based SEM image models can be developed. The models are sufficiently accurate and fast to find their applications in semiconductor manufacturing, and they can be used as independent model for OPC data verification or generate SEM images as reference for SEM defect scan metrology.

In semiconductor manufacturing, the state of the art for wafer quality control is based on product monitoring and feedback control loops; the related metrology operations, that usually involve scanning electron microscopes, are particularly cost-intensive and time-consuming. It is therefore not possible to evaluate every wafer: commonly, a small subset of a productive lot is measured at the metrology station and delegated to represent the whole lot. Virtual Metrology (VM) methodologies aim to obtain reliable estimates of metrology results without actually performing measurement operations; this goal is usually achieved by means of statistical models, linking process data and context information to target measurements. In this paper, we tackle two of the most important issues in VM: (i) regression in high dimensional spaces where few variables are meaningful, and (ii) data heterogeneity caused by inhomogeneous production and equipment logistics. We propose a hierarchical framework based on ℓ <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">1</sub> -penalized machine learning techniques and solved by means of multitask learning strategies. The proposed methodology is validated on actual process and measurement data from the semiconductor manufacturing industry.

Code smells are considered symptoms of poor implementation choices, which may hamper the software maintainability. Hence, code smells should be detected as early as possible to avoid software quality degradation. Unfortunately, detecting code smells is not a trivial task. Some preliminary studies investigated and concluded that machine learning (ML) techniques are a promising way to better support smell detection. However, these techniques are hard to be customized to promote an early and accurate detection of specific smell types. Yet, ML techniques usually require numerous code examples to be trained (composing a relevant dataset) in order to achieve satisfactory accuracy. Unfortunately, such a dependency on a large validated dataset is impractical and leads to late detection of code smells. Thus, a prevailing challenge is the early customized detection of code smells taking into account the typical limited training data. In this direction, this paper reports a study in which we collected code smells, from ten active projects, that were actually refactored by developers, differently from studies that rely on code smells inferred by researchers. These smells were used for evaluating the accuracy regarding early detection of code smells by using seven ML techniques. Once we take into account such smells that were considered as important by developers, the ML techniques are able to customize the detection in order to focus on smells observed as relevant in the investigated systems. The results showed that all the analyzed techniques are sensitive to the type of smell and obtained good results for the majority of them, especially JRip and Random Forest. We also observe that the ML techniques did not need a high number of examples to reach their best accuracy results. This finding implies that ML techniques can be successfully used for early detection of smells without depending on the curation of a large dataset.

A Convolutional Autoencoder Approach for Feature Extraction in Virtual Metrology