GLGuoyi Li

Guoyi Li

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Guoyi Li has a total of 24 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Deep Learning based Optical Inspection with Centralized Analysis for High Volume Smart Manufacturing

    by Guoyi Li, Chao Feng, Addishiwot Woldesenbet, Bruce King, Hamid Hadavi, Vikasini Moku, Kurt Loken, Gary Kunkel

    Increased capabilities in data storage and exploration provide significant insights for quality assurance in a high volume manufacturing environment. However, these opportunities are associated with great challenges in analytical model development, application deployment, system throughput and reliability. While no commercial software system fully meets the needs of recording head factories in Seagate, a novel strategy named optical inspection with centralized analysis has been developed to detect defects of trailing edges of the recording heads of hard disk drives, and fail the parts when necessary. Leveraging the state-of-the-art artificial intelligence technologies, a deep learning based semantic segmentation engine is built using convolutional neural networks for optical inspection. It has shown an improved accuracy to that of visual inspection performed by human. Meanwhile, a high performance computation engine has been built as a Kubernetes cluster with multiple GPU and CPU units. It is able to achieve the target throughput of three million high-resolution images in each day (i.e., 12 TB image data and 35 images per second). With the high fidelity offered by Kubernetes cluster, the developed applications (inference engine, preprocessor, postprocessor, etc.) serve as containerized microservices independently. Such an architecture ensures the vertical and horizontal scalabilities according to the computation of each individual deployment, while all deployments communicate through an Advanced Message Queuing Protocol cluster without human interference. This analytic framework enables Industry 4.0 recording head manufacturing by integrating advanced AI technologies with a robust edge computation architecture.


  • Detection of Sealant Delamination in Integrated Circuit Package Using Nondestructive Evaluation

    by Guoyi Li, Hyunseong Lee, Aditi Chattopadhyay, Rajesh Kumar Neerukatti, Kuang C. Liu

    This paper presents the development of a nondestructive evaluation method for delamination detection in integrated circuit (IC) packages. The developed method allows for quantitative detection of sealant delamination between integrated heat sink (IHS) and substrate, which is considered a potential failure mechanism in IC packages. The proposed method is expected to overcome the destructive nature of most existing techniques, while maintaining low cost of development. The reported development consists of the following processes: First, flash thermographic analysis is investigated and shown to be incapable of detecting a realistic sealant delamination. Then, ultrasonic guided waves (UGWs) are used as the chosen interrogation method due to their sensitivity to small-size damage and structural thickness variations. The complexity of the received ultrasonic signals caused by the geometric heterogeneity is resolved using a time–frequency signal processing technique. The extracted ultrasonic information, including time-of-arrival (TOA) and amplitude of wave modes received from different sensing paths under multiple excitation frequencies, is then used to construct a feature space. A multivariate Gaussian model is implemented as an information fusion and delamination detection tool, which investigates the distribution of feature space including correlations between features. Results from the developed model are compared with two existing evaluation methods, including pullout force measurement and a delamination metric. The developed method is seen to possess a similar level of accuracy with a nondestructive nature.


  • Open Access

    Health Monitoring Framework for Aircraft Engine System using Deep Neural Network

    by Hyunseong Lee, Guoyi Li, Ashwin Rai, Aditi Chattopadhyay

    A real-time monitoring framework is developed to detect operational anomalies in aircraft engine performance. A historical flight dataset recorded from commercial aircraft is utilized to perform the proposed method. Sampling frequency synchronization and denoise are performed on the flight dataset using signal processing techniques. A robust detection algorithm using the deep neural network is developed to capture flight performance anomalies that show significant off-nominal behavior in engine related and flight dynamic features. The accuracy and efficiency of the proposed monitoring method are validated through a demonstration of anomaly detection in the aircraft engine system associated with dynamic flight behavior.


  • An online‐offline prognosis model for fatigue life prediction under biaxial cyclic loading with overloads

    by Guoyi Li, Siddhant Datta, Aditi Chattopadhyay, Nagaraja Iyyer, Nam Phan

    This paper presents a robust online‐offline model for the prediction of crack propagation under complex in‐phase biaxial fatigue loading in the presence of overloads of different magnitudes. The online prognosis model comprises a combination of finite element analysis and data‐driven regression to predict the crack propagation under constant loading, while the offline model is trained using experimental data to inform the post‐overload crack growth retardation behavior to the online model. The developed methodology is validated by conducting biaxial fatigue experiments using aluminum AA7075‐T651 alloy cruciform specimens. A close correlation is observed between the experimental results and model predictions. The results show that the model successfully predicts the crack retardation behavior under the influence of overloads with different magnitudes occurring at different stages of fatigue crack growth. Error analysis is conducted to investigate the sensitivities of the number of training points and crack increments to the prediction accuracy. In addition, the error propagation with respect to the crack length is studied, which provides constructive suggestions for further model improvement.


  • Reference-free damage localization in time-space domain for structural health monitoring of X-COR sandwich composites

    by Guoyi Li, Aditi Chattopadhyay

    This article presents a guided wave based damage localization framework using a time-space analysis for structural health monitoring of X-COR sandwich composites with a reference-free perspective to overcome the difficulty in detecting reflected guided waves in a highly attenuated media. Transducers, including macro-fiber composites and piezoelectric wafers, are used to design the sensing paths. The time-space domain is constructed using de-noised signals that are processed by signal processing techniques including matching pursuit decomposition and Hilbert transform. The localization framework is then validated across a wide range of excitation frequencies in X-COR sandwich composites with seeded facesheet delamination. The results indicate that time-space analysis offers a high accuracy for detection and localization of internal damages and serves as a promising framework for structural health monitoring of complex sandwich composites with reinforcements. This work also provides a comprehensive study of the changes in group velocities, attenuation tendencies, and time-space resolution of actuated and converted modes under different excitation frequencies across a range of ultrasonic transducer sizes, thereby helping to improve reliability and accuracy of damage localization in time-space domain.


  • Open Access

    Operational Anomaly Detection in Flight Data Using a Multivariate Gaussian Mixture Model

    by Guoyi Li, Ashwin Rai, Hyunseong Lee, Aditi Chattopadhyay

    This paper presents a robust real-time aircraft health monitoring framework using a machine learning based approach, specifically the multivariate Gaussian mixture model (mGMM), for the detection of in-air operational anomalies of an aircraft system. Sensor fusion and noise filtering algorithms have also been adopted to reduce dimensionality of the feature space while avoiding the elimination of useful information from the original flight data. Random noise in each feature, induced by the aircraft sensors and data acquisition system, is filtered out using a weighted averaging window while maintaining inherent variances. The filtered dataset is then fused according to the underlying physics of each sensed feature to reduce redundant features and subsequently trained using the mGMM. The methodology allows monitoring the behavior of each feature as well as correlations between features, significantly improving detection sensitivity. The high computational efficiency of this approach permits real-time monitoring of an aircraft system.


  • A Reference-Free Guided Wave Based Damage Localization Approach for Highly Dispersive Structures

    by Guoyi Li, Abhishek Rajadas, Aditi Chattopadhyay, Daniel W. Huff

    A Reference-Free Guided Wave Based Damage Localization Approach for Highly Dispersive Structures

  • Ultrasonic Guided Wave Inspection and Characterization of Integrated Circuit Packages with Delamination

    by Javaid Ikram, Guoyi Li, Rajesh K. Neerukatti, Aditi Chattopadhyay

    Ultrasonic Guided Wave Inspection and Characterization of Integrated Circuit Packages with Delamination

  • Ultrasonic guided wave propagation in composites including damage using high-fidelity local interaction simulation

    by Guoyi Li, Rajesh Kumar Neerukatti, Aditi Chattopadhyay

    Composite materials are used in many advanced engineering applications because of high specific strength and stiffness. Their complex damage mechanisms and failure modes, however, are still not well-understood, thus challenging the application safety. Ultrasonic guided waves are promising structural health monitoring tools used to determine the operational safety of composite materials. In this article, a fully coupled numerical simulation model is used to study wave propagation and dispersion in composites under varying sensor locations, propagating orientations, excitation frequencies, and damage locations. The model is based on the local interaction simulation approaches/sharp interface model wherein output sensor signals are processed using the matching pursuit decomposition algorithm to study the signal features in the time–frequency domain. The changes in signals due to varying damage locations with respect to the through-thickness direction are studied under anti-symmetrical and symmetrical excitation scenarios. The results show that the signal from symmetric excitation is more sensitive to the damage location, while the signal from anti-symmetric excitation is less dispersive. It indicates that comprising effective feature extraction technique with the accurate physics-based numerical simulation model can be implemented to develop robust structural health monitoring framework for composites.