Selected work

Representative Papers

Accelerating nanodrug development in continuous flow systems using informed prediction models based on low-cost surrogate nanoparticles

Aug 06, 2026

This study addresses the challenges in nanomedicine development posed by the high sensitivity of nanoparticle size and polydispersity index to process parameters, which renders traditional trial-and-error approaches costly and time-consuming. To overcome this, the work integrates microfluidic experimentation with expert knowledge and, for the first time, incorporates shape constraints into a machine learning model. By leveraging a small amount of low-cost surrogate data, the approach accurately predicts critical physicochemical properties of lipid-based nanoparticles. The method achieves high-fidelity modeling of both liposome and lipid nanoparticle size and dispersity with minimal experimental samples, substantially reducing the need for extensive screening experiments and enabling rational, efficient design of nanomedicine manufacturing processes in continuous-flow systems.

0 citationsRead paper

Schmidt Decomposition-Based Methods for Efficient Quantum Image Encoding

Jun 09, 2026

This work addresses the high circuit depth, excessive gate count, and substantial qubit requirements of mainstream quantum image encoding schemes—such as FRQI, QPIE, and NEQR—which hinder their deployment on noisy intermediate-scale quantum (NISQ) devices. To overcome these limitations, the study introduces Schmidt decomposition into quantum image encoding for the first time, leveraging low-rank quantum state approximation to preserve essential image information while drastically reducing circuit complexity. Experimental results demonstrate that the proposed approach achieves near-perfect image reconstruction in FRQI (MSE ≈ 0.27) with a 97% reduction in circuit depth, significantly enhancing feasibility on NISQ hardware and effectively balancing reconstruction accuracy with resource efficiency.

0 citationsRead paper

Automated Batch Distillation Process Simulation for a Large Hybrid Dataset for Deep Anomaly Detection

Apr 10, 2026

This work addresses the scarcity of large-scale, well-annotated time-series datasets for deep anomaly detection in chemical processes. The authors propose an automated workflow that pioneers the translation of experimental logs into simulation scenarios, leveraging a tailored model-order reduction strategy to efficiently solve differential-algebraic equations and enable fully automatic, consistent simulations across diverse operating conditions. By aligning experimental and simulated data and establishing structured anomaly mappings, the framework integrates real and synthetic data to construct the first hybrid anomaly detection dataset specifically for distillation processes. This dataset encompasses a variety of actuator and control anomalies and supports simulation-to-experiment style transfer and pseudo-experimental data generation, thereby providing a high-quality benchmark for advancing anomaly detection in chemical engineering.

0 citationsRead paper

Hybrid Quantum-Classical AI for Industrial Defect Classification in Welding Images

Mar 30, 2026

This work addresses the task of defect classification in industrial welding images by proposing two hybrid quantum-classical approaches: a quantum kernel-based classifier and a variational quantum circuit model. The methodology first employs a convolutional neural network to extract image features, which are subsequently processed using parameterized quantum feature maps, angle encoding, and a variational quantum linear solver for classification. Notably, this study introduces, for the first time in industrial quality inspection, an analysis of quantum kernel condition numbers. The proposed methods are evaluated on a real-world welding dataset, demonstrating competitive performance against classical CNNs in both binary and multiclass classification tasks. Experimental results indicate that the hybrid models achieve comparable accuracy to their classical counterparts, highlighting their potential for near-term practical deployment.

0 citationsRead paper

Synthetic Defect Geometries of Cast Metal Objects Modeled via 2d Voronoi Tessellations

Feb 05, 2026

This work addresses the scarcity of high-quality, pixel-level annotated defect data in industrial quality inspection, which hinders the training of automated detection models. To overcome this limitation, the authors propose a transferable, parameterized defect modeling approach that constructs three-dimensional geometric defect models based on two-dimensional Voronoi tessellation, embeds them into digital twins of castings, and leverages physics-driven Monte Carlo simulation to generate large-scale, realistic, and diverse synthetic defect datasets. The method enables controllable simulation of rare defects while simultaneously providing precise pixel-level ground truth annotations. It is adaptable across multiple non-destructive testing scenarios and significantly enhances the training and validation performance of vision-based surface inspection algorithms.

0 citationsRead paper

Recent publications

Latest Papers

Accelerating nanodrug development in continuous flow systems using informed prediction models based on low-cost surrogate nanoparticles

Aug 06, 2026

This study addresses the challenges in nanomedicine development posed by the high sensitivity of nanoparticle size and polydispersity index to process parameters, which renders traditional trial-and-error approaches costly and time-consuming. To overcome this, the work integrates microfluidic experimentation with expert knowledge and, for the first time, incorporates shape constraints into a machine learning model. By leveraging a small amount of low-cost surrogate data, the approach accurately predicts critical physicochemical properties of lipid-based nanoparticles. The method achieves high-fidelity modeling of both liposome and lipid nanoparticle size and dispersity with minimal experimental samples, substantially reducing the need for extensive screening experiments and enabling rational, efficient design of nanomedicine manufacturing processes in continuous-flow systems.

0 citationsRead paper

Schmidt Decomposition-Based Methods for Efficient Quantum Image Encoding

Jun 09, 2026

This work addresses the high circuit depth, excessive gate count, and substantial qubit requirements of mainstream quantum image encoding schemes—such as FRQI, QPIE, and NEQR—which hinder their deployment on noisy intermediate-scale quantum (NISQ) devices. To overcome these limitations, the study introduces Schmidt decomposition into quantum image encoding for the first time, leveraging low-rank quantum state approximation to preserve essential image information while drastically reducing circuit complexity. Experimental results demonstrate that the proposed approach achieves near-perfect image reconstruction in FRQI (MSE ≈ 0.27) with a 97% reduction in circuit depth, significantly enhancing feasibility on NISQ hardware and effectively balancing reconstruction accuracy with resource efficiency.

0 citationsRead paper

Automated Batch Distillation Process Simulation for a Large Hybrid Dataset for Deep Anomaly Detection

Apr 10, 2026

This work addresses the scarcity of large-scale, well-annotated time-series datasets for deep anomaly detection in chemical processes. The authors propose an automated workflow that pioneers the translation of experimental logs into simulation scenarios, leveraging a tailored model-order reduction strategy to efficiently solve differential-algebraic equations and enable fully automatic, consistent simulations across diverse operating conditions. By aligning experimental and simulated data and establishing structured anomaly mappings, the framework integrates real and synthetic data to construct the first hybrid anomaly detection dataset specifically for distillation processes. This dataset encompasses a variety of actuator and control anomalies and supports simulation-to-experiment style transfer and pseudo-experimental data generation, thereby providing a high-quality benchmark for advancing anomaly detection in chemical engineering.

0 citationsRead paper

Hybrid Quantum-Classical AI for Industrial Defect Classification in Welding Images

Mar 30, 2026

This work addresses the task of defect classification in industrial welding images by proposing two hybrid quantum-classical approaches: a quantum kernel-based classifier and a variational quantum circuit model. The methodology first employs a convolutional neural network to extract image features, which are subsequently processed using parameterized quantum feature maps, angle encoding, and a variational quantum linear solver for classification. Notably, this study introduces, for the first time in industrial quality inspection, an analysis of quantum kernel condition numbers. The proposed methods are evaluated on a real-world welding dataset, demonstrating competitive performance against classical CNNs in both binary and multiclass classification tasks. Experimental results indicate that the hybrid models achieve comparable accuracy to their classical counterparts, highlighting their potential for near-term practical deployment.

0 citationsRead paper

Synthetic Defect Geometries of Cast Metal Objects Modeled via 2d Voronoi Tessellations

Feb 05, 2026

This work addresses the scarcity of high-quality, pixel-level annotated defect data in industrial quality inspection, which hinders the training of automated detection models. To overcome this limitation, the authors propose a transferable, parameterized defect modeling approach that constructs three-dimensional geometric defect models based on two-dimensional Voronoi tessellation, embeds them into digital twins of castings, and leverages physics-driven Monte Carlo simulation to generate large-scale, realistic, and diverse synthetic defect datasets. The method enables controllable simulation of rare defects while simultaneously providing precise pixel-level ground truth annotations. It is adaptable across multiple non-destructive testing scenarios and significantly enhances the training and validation performance of vision-based surface inspection algorithms.

0 citationsRead paper