Jionghua (Judy) Jin
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jionghua (Judy) Jin has a total of 60 co-authors. Here are the top co-authors by number of shared publications:
- Kamran Paynabar
Georgia Institute of Technology, Atlanta, GA, 30332
3 shared publications - S. Jack Hu
Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI 48109
3 shared publications - Wenbo Sun
University of Michigan Transportation Research Institute, University of Michigan
3 shared publications - Chenhui Shao
Department of Mechanical Science and Engineering University of Illinois at Urbana-Champaign 61801 Urbana IL USA
2 shared publications - Hadi Ibrahim Masoud
Department of Industrial Engineering, University of Jeddah, Jeddah, Saudi Arabia
2 shared publications - Matthew P. Reed
University of Michigan Transportation Research Institute, Ann Arbor, USA
2 shared publications
Latest Publications
- 31 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 46 (2) , 944-956
WOOD: Wasserstein-Based Out-of-Distribution Detection
The training and testing data for deep-neural-network-based classifiers are usually assumed to be sampled from the same distribution. When part of the testing samples are drawn from a distribution that is sufficiently far away from that of the training samples (a.k.a. out-of-distribution (OOD) samples), the trained neural network has a tendency to make high-confidence predictions for these OOD samples. Detection of the OOD samples is critical when training a neural network used for image classification, object detection, etc. It can enhance the classifier's robustness to irrelevant inputs, and improve the system's resilience and security under different forms of attacks. Detection of OOD samples has three main challenges: (i) the proposed OOD detection method should be compatible with various architectures of classifiers (e.g., DenseNet, ResNet) without significantly increasing the model complexity and requirements on computational resources; (ii) the OOD samples may come from multiple distributions, whose class labels are commonly unavailable; (iii) a score function needs to be defined to effectively separate OOD samples from in-distribution (InD) samples. To overcome these challenges, we propose a Wasserstein-based out-of-distribution detection (WOOD) method. The basic idea is to define a Wasserstein-based score that evaluates the dissimilarity between a test sample and the distribution of InD samples. An optimization problem is then formulated and solved based on the proposed score function. The statistical learning bound of the proposed method is investigated to guarantee that the loss value achieved by the empirical optimizer approximates the global optimum. The comparison study results demonstrate that the proposed WOOD consistently outperforms other existing OOD detection methods.
- 31 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 46 (2) , 944-956
WOOD: Wasserstein-Based Out-of-Distribution Detection
The training and testing data for deep-neural-network-based classifiers are usually assumed to be sampled from the same distribution. When part of the testing samples are drawn from a distribution that is sufficiently far away from that of the training samples (a.k.a. out-of-distribution (OOD) samples), the trained neural network has a tendency to make high-confidence predictions for these OOD samples. Detection of the OOD samples is critical when training a neural network used for image classification, object detection, etc. It can enhance the classifier's robustness to irrelevant inputs, and improve the system's resilience and security under different forms of attacks. Detection of OOD samples has three main challenges: (i) the proposed OOD detection method should be compatible with various architectures of classifiers (e.g., DenseNet, ResNet) without significantly increasing the model complexity and requirements on computational resources; (ii) the OOD samples may come from multiple distributions, whose class labels are commonly unavailable; (iii) a score function needs to be defined to effectively separate OOD samples from in-distribution (InD) samples. To overcome these challenges, we propose a Wasserstein-based out-of-distribution detection (WOOD) method. The basic idea is to define a Wasserstein-based score that evaluates the dissimilarity between a test sample and the distribution of InD samples. An optimization problem is then formulated and solved based on the proposed score function. The statistical learning bound of the proposed method is investigated to guarantee that the loss value achieved by the empirical optimizer approximates the global optimum. The comparison study results demonstrate that the proposed WOOD consistently outperforms other existing OOD detection methods.
- 2 December 2022
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 145 (3) , 1-44
MVGCN: Multi-View Graph Convolutional Neural Network for Surface Defect Identification Using Three-Dimensional Point Cloud
Surface defect identification is a crucial task in many manufacturing systems, including automotive, aircraft, steel rolling, and precast concrete. Although image-based surface defect identification methods have been proposed, these methods usually have two limitations: images may lose partial information, such as depths of surface defects, and their precision is vulnerable to many factors, such as the inspection angle, light, color, noise, etc. Given that a three-dimensional (3D) point cloud can precisely represent the multidimensional structure of surface defects, we aim to detect and classify surface defects using a 3D point cloud. This has two major challenges: (i) the defects are often sparsely distributed over the surface, which makes their features prone to be hidden by the normal surface and (ii) different permutations and transformations of 3D point cloud may represent the same surface, so the proposed model needs to be permutation and transformation invariant. In this paper, a two-step surface defect identification approach is developed to investigate the defects’ patterns in 3D point cloud data. The proposed approach consists of an unsupervised method for defect detection and a multi-view deep learning model for defect classification, which can keep track of the features from both defective and non-defective regions. We prove that the proposed approach is invariant to different permutations and transformations. Two case studies are conducted for defect identification on the surfaces of synthetic aircraft fuselage and the real precast concrete specimen, respectively. The results show that our approach receives the best defect detection and classification accuracy compared with other benchmark methods.
- 2 December 2022
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 145 (3) , 1-44
MVGCN: Multi-View Graph Convolutional Neural Network for Surface Defect Identification Using Three-Dimensional Point Cloud
Surface defect identification is a crucial task in many manufacturing systems, including automotive, aircraft, steel rolling, and precast concrete. Although image-based surface defect identification methods have been proposed, these methods usually have two limitations: images may lose partial information, such as depths of surface defects, and their precision is vulnerable to many factors, such as the inspection angle, light, color, noise, etc. Given that a three-dimensional (3D) point cloud can precisely represent the multidimensional structure of surface defects, we aim to detect and classify surface defects using a 3D point cloud. This has two major challenges: (i) the defects are often sparsely distributed over the surface, which makes their features prone to be hidden by the normal surface and (ii) different permutations and transformations of 3D point cloud may represent the same surface, so the proposed model needs to be permutation and transformation invariant. In this paper, a two-step surface defect identification approach is developed to investigate the defects’ patterns in 3D point cloud data. The proposed approach consists of an unsupervised method for defect detection and a multi-view deep learning model for defect classification, which can keep track of the features from both defective and non-defective regions. We prove that the proposed approach is invariant to different permutations and transformations. Two case studies are conducted for defect identification on the surfaces of synthetic aircraft fuselage and the real precast concrete specimen, respectively. The results show that our approach receives the best defect detection and classification accuracy compared with other benchmark methods.
- Open Access
- 10 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 156071-156113
The Internet of Federated Things (IoFT)
The Internet of Things (IoT) is on the verge of a major paradigm shift. In the IoT system of the future, IoFT, the “cloud” will be substituted by the “crowd” where model training is brought to the edge, allowing IoT devices to collaboratively extract knowledge and build smart analytics/models while keeping their personal data stored locally. This paradigm shift was set into motion by the tremendous increase in computational power on IoT devices and the recent advances in decentralized and privacy-preserving model training, coined as federated learning (FL). This article provides a vision for IoFT and a systematic overview of current efforts towards realizing this vision. Specifically, we first introduce the defining characteristics of IoFT and discuss FL data-driven approaches, opportunities, and challenges that allow decentralized inference within three dimensions: (i) a global model that maximizes utility across all IoT devices, (ii) a personalized model that borrows strengths across all devices yet retains its own model, (iii) a meta-learning model that quickly adapts to new devices or learning tasks. We end by describing the vision and challenges of IoFT in reshaping different industries through the lens of domain experts. Those industries include manufacturing, transportation, energy, healthcare, quality & reliability, business, and computing.
- 10 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 156071-156113
Open AccessThe Internet of Federated Things (IoFT)
The Internet of Things (IoT) is on the verge of a major paradigm shift. In the IoT system of the future, IoFT, the “cloud” will be substituted by the “crowd” where model training is brought to the edge, allowing IoT devices to collaboratively extract knowledge and build smart analytics/models while keeping their personal data stored locally. This paradigm shift was set into motion by the tremendous increase in computational power on IoT devices and the recent advances in decentralized and privacy-preserving model training, coined as federated learning (FL). This article provides a vision for IoFT and a systematic overview of current efforts towards realizing this vision. Specifically, we first introduce the defining characteristics of IoFT and discuss FL data-driven approaches, opportunities, and challenges that allow decentralized inference within three dimensions: (i) a global model that maximizes utility across all IoT devices, (ii) a personalized model that borrows strengths across all devices yet retains its own model, (iii) a meta-learning model that quickly adapts to new devices or learning tasks. We end by describing the vision and challenges of IoFT in reshaping different industries through the lens of domain experts. Those industries include manufacturing, transportation, energy, healthcare, quality & reliability, business, and computing.
- 1 July 2021
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 144 (1)
In Situ Monitoring of Optical Emission Spectra for Microscopic Pores in Metal Additive Manufacturing
Quality assurance techniques are increasingly demanded in additive manufacturing. Going beyond most of the existing research that focuses on the melt pool temperature monitoring, we develop a new method that monitors the in situ optical emission spectra signals. Optical emission spectra signals have been showing a potential capability of detecting microscopic pores. The concept is to extract features from the optical emission spectra via deep auto-encoders and then cluster the features into two quality groups to consider both unlabeled and labeled samples in a semi-supervised manner. The method is integrated with multitask learning to make it adaptable for the samples collected from multiple processes. Both a simulation example and a case study are performed to demonstrate the effectiveness of the proposed method.
- 1 July 2021
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 144 (1)
In Situ Monitoring of Optical Emission Spectra for Microscopic Pores in Metal Additive Manufacturing
Quality assurance techniques are increasingly demanded in additive manufacturing. Going beyond most of the existing research that focuses on the melt pool temperature monitoring, we develop a new method that monitors the in situ optical emission spectra signals. Optical emission spectra signals have been showing a potential capability of detecting microscopic pores. The concept is to extract features from the optical emission spectra via deep auto-encoders and then cluster the features into two quality groups to consider both unlabeled and labeled samples in a semi-supervised manner. The method is integrated with multitask learning to make it adaptable for the samples collected from multiple processes. Both a simulation example and a case study are performed to demonstrate the effectiveness of the proposed method.
- 10 June 2019
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 141 (8) , 1-34
Profile Monitoring and Fault Diagnosis Via Sensor Fusion for Ultrasonic Welding
Sensor signals acquired during the manufacturing process contain rich information that can be used to facilitate effective monitoring of operational quality, early detection of system anomalies, and quick diagnosis of fault root causes. This paper develops a method for effective monitoring and diagnosis of multisensor heterogeneous profile data based on multilinear discriminant analysis. The proposed method operates directly on the multistream profiles and then extracts uncorrelated discriminative features through tensor-to-vector projection, and thus, preserving the interrelationship of different sensors. The extracted features are then fed into classifiers to detect faulty operations and recognize fault types. The developed method is demonstrated with both simulated and real data from ultrasonic metal welding.
- 10 June 2019
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 141 (8) , 1-34
Profile Monitoring and Fault Diagnosis Via Sensor Fusion for Ultrasonic Welding
Sensor signals acquired during the manufacturing process contain rich information that can be used to facilitate effective monitoring of operational quality, early detection of system anomalies, and quick diagnosis of fault root causes. This paper develops a method for effective monitoring and diagnosis of multisensor heterogeneous profile data based on multilinear discriminant analysis. The proposed method operates directly on the multistream profiles and then extracts uncorrelated discriminative features through tensor-to-vector projection, and thus, preserving the interrelationship of different sensors. The extracted features are then fed into classifiers to detect faulty operations and recognize fault types. The developed method is demonstrated with both simulated and real data from ultrasonic metal welding.
- 4 October 2018
- Published by Elsevier in International Journal of Industrial Ergonomics
- Vol. 69, 23-28
Sample size calculations for a functional human motion analysis: Application to vehicle ingress discomfort prediction
- 4 October 2018
- Published by Elsevier in International Journal of Industrial Ergonomics
- Vol. 69, 23-28
Sample size calculations for a functional human motion analysis: Application to vehicle ingress discomfort prediction
- 24 August 2017
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 139 (10) , 101002
Dynamic Sampling Design for Characterizing Spatiotemporal Processes in Manufacturing
Fine-scale characterization and monitoring of spatiotemporal processes are crucial for high-performance quality control of manufacturing processes, such as ultrasonic metal welding and high-precision machining. However, it is generally expensive to acquire high-resolution spatiotemporal data in manufacturing due to the high cost of the three-dimensional (3D) measurement system or the time-consuming measurement process. In this paper, we develop a novel dynamic sampling design algorithm to cost-effectively characterize spatiotemporal processes in manufacturing. A spatiotemporal state-space model and Kalman filter are used to predictively determine the measurement locations using a criterion considering both the prediction performance and the measurement cost. The determination of measurement locations is formulated as a binary integer programming problem, and genetic algorithm (GA) is applied for searching the optimal design. In addition, a new test statistic is proposed to monitor and update the surface progression rate. Both simulated and real-world spatiotemporal data are used to demonstrate the effectiveness of the proposed method.
- 24 August 2017
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 139 (10) , 101002
Dynamic Sampling Design for Characterizing Spatiotemporal Processes in Manufacturing
Fine-scale characterization and monitoring of spatiotemporal processes are crucial for high-performance quality control of manufacturing processes, such as ultrasonic metal welding and high-precision machining. However, it is generally expensive to acquire high-resolution spatiotemporal data in manufacturing due to the high cost of the three-dimensional (3D) measurement system or the time-consuming measurement process. In this paper, we develop a novel dynamic sampling design algorithm to cost-effectively characterize spatiotemporal processes in manufacturing. A spatiotemporal state-space model and Kalman filter are used to predictively determine the measurement locations using a criterion considering both the prediction performance and the measurement cost. The determination of measurement locations is formulated as a binary integer programming problem, and genetic algorithm (GA) is applied for searching the optimal design. In addition, a new test statistic is proposed to monitor and update the surface progression rate. Both simulated and real-world spatiotemporal data are used to demonstrate the effectiveness of the proposed method.
- Open Access
Surrogate Measures for the Robust Scheduling of Stochastic Job Shop Scheduling Problems
This study focuses on surrogate measures (SMs) of robustness for the stochastic job shop scheduling problems (SJSSP) with uncertain processing times. The objective is to provide the robust predictive schedule to the decision makers. The mathematical model of SJSSP is formulated by considering the railway execution strategy, which defined that the starting time of each operation cannot be earlier than its predictive starting time. Robustness is defined as the expected relative deviation between the realized makespan and the predictive makespan. In view of the time-consuming characteristic of simulation-based robustness measure (RMsim), this paper puts forward new SMs and investigates their performance through simulations. By utilizing the structure of schedule and the available information of stochastic processing times, two SMs on the basis of minimizing the robustness degradation on the critical path and the non-critical path are suggested. For this purpose, a hybrid estimation of distribution algorithm (HEDA) is adopted to conduct the simulations. To analyze the performance of the presented SMs, two computational experiments are carried out. Specifically, the correlation analysis is firstly conducted by comparing the coefficient of determination between the presented SMs and the corresponding simulation-based robustness values with those of the existing SMs. Secondly, the effectiveness and the performance of the presented SMs are further validated by comparing with the simulation-based robustness measure under different uncertainty levels. The experimental results demonstrate that the presented SMs are not only effective for assessing the robustness of SJSSP no matter the uncertainty levels, but also require a tremendously lower computational burden than the simulation-based robustness measure.
Open AccessSurrogate Measures for the Robust Scheduling of Stochastic Job Shop Scheduling Problems
This study focuses on surrogate measures (SMs) of robustness for the stochastic job shop scheduling problems (SJSSP) with uncertain processing times. The objective is to provide the robust predictive schedule to the decision makers. The mathematical model of SJSSP is formulated by considering the railway execution strategy, which defined that the starting time of each operation cannot be earlier than its predictive starting time. Robustness is defined as the expected relative deviation between the realized makespan and the predictive makespan. In view of the time-consuming characteristic of simulation-based robustness measure (RMsim), this paper puts forward new SMs and investigates their performance through simulations. By utilizing the structure of schedule and the available information of stochastic processing times, two SMs on the basis of minimizing the robustness degradation on the critical path and the non-critical path are suggested. For this purpose, a hybrid estimation of distribution algorithm (HEDA) is adopted to conduct the simulations. To analyze the performance of the presented SMs, two computational experiments are carried out. Specifically, the correlation analysis is firstly conducted by comparing the coefficient of determination between the presented SMs and the corresponding simulation-based robustness values with those of the existing SMs. Secondly, the effectiveness and the performance of the presented SMs are further validated by comparing with the simulation-based robustness measure under different uncertainty levels. The experimental results demonstrate that the presented SMs are not only effective for assessing the robustness of SJSSP no matter the uncertainty levels, but also require a tremendously lower computational burden than the simulation-based robustness measure.
- 1 March 2017
- Published by Elsevier in Applied Ergonomics
- Vol. 59, 401-409
Analysis of human motion variation patterns using UMPCA
- 1 March 2017
- Published by Elsevier in Applied Ergonomics
- Vol. 59, 401-409
Analysis of human motion variation patterns using UMPCA
- Open Access
- 29 September 2016
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 139 (1)
Improving Machined Surface Shape Prediction by Integrating Multi-Task Learning With Cutting Force Variation Modeling
The shapes of machined surfaces play a critical role affecting powertrain performance, and therefore, it is necessary to characterize the shapes with high resolution. State-of-the-art approaches for surface shape characterization are mostly data-driven by interpolating and extrapolating the spatial data but its precision is limited by the density of measurements. This paper explores the new opportunity of improving surface shape prediction through considering the similarity of multiple similar manufacturing processes. It is a common scenario when the process of interest lacks sufficient data whereas rich data could be available from other similar-but-not-identical processes. It is reasonable to transfer the insights gained from other relevant processes into the surface shape prediction. This paper develops an engineering-guided multitask learning (EG-MTL) surface model by fusing surface cutting physics in engineering processes and the spatial data from a number of similar-but-not-identical processes. An iterative multitask Gaussian process learning algorithm is developed to learn the model parameters. Compared with the conventional multitask learning, the proposed method has the advantages in incorporating the insights on cutting force variation during machining and is potentially able to improve the prediction performance given limited measurement data. The methodology is demonstrated based on the data from real-world machining processes in an engine plant.
- 29 September 2016
- Published by ASME International in Journal of Manufacturing Science and Engineering
- Vol. 139 (1)
Open AccessImproving Machined Surface Shape Prediction by Integrating Multi-Task Learning With Cutting Force Variation Modeling
The shapes of machined surfaces play a critical role affecting powertrain performance, and therefore, it is necessary to characterize the shapes with high resolution. State-of-the-art approaches for surface shape characterization are mostly data-driven by interpolating and extrapolating the spatial data but its precision is limited by the density of measurements. This paper explores the new opportunity of improving surface shape prediction through considering the similarity of multiple similar manufacturing processes. It is a common scenario when the process of interest lacks sufficient data whereas rich data could be available from other similar-but-not-identical processes. It is reasonable to transfer the insights gained from other relevant processes into the surface shape prediction. This paper develops an engineering-guided multitask learning (EG-MTL) surface model by fusing surface cutting physics in engineering processes and the spatial data from a number of similar-but-not-identical processes. An iterative multitask Gaussian process learning algorithm is developed to learn the model parameters. Compared with the conventional multitask learning, the proposed method has the advantages in incorporating the insights on cutting force variation during machining and is potentially able to improve the prediction performance given limited measurement data. The methodology is demonstrated based on the data from real-world machining processes in an engine plant.