Invited Speeches

Weifei Hu
Zhejiang University, China

Title: AI-Driven Reliability-Based Design Optimization (AI-RBDO)
Abstract: Complex engineering systems often involves multidisciplinary and multiphysics analyses, which exhibit challenges such as non-steady-state loading, nonlinear performance response, and non-explicit failure mechanism, posing significant difficulties to reliability-based design optimization (RBDO) in both theoretical research and practical applications. In order to address these challenges, this talk will present some recent research in AI-driven reliability-based design optimization (AI-RBDO), mainly including: 1. A reduced-order modeling method based on parameterized artificial neural networks, which improves the efficiency and accuracy of global design performance evaluation; 2. A spatiotemporal reliability analysis (RA) method based on physics-informed neural networks achieving efficient and accurate global time-varying reliability; 3.RBDO for complex engineering systems driven by deep reinforcement learning and large language model (LLM). The proposed AI-RBDO theories and methods have been successfully applied in aeroelastic coupling analysis of large wind turbine (WT) blades, rigid-flexible coupling analysis of WT drivetrain, WT safety factor calibration, novel airfoil design, and RA and RBDO of other engineering systems..

Bio: Weifei Hu, tenured associate professor and Ph.D. supervisor, was selected into the national young talent program in 2020. He received the B.S. degree in 2008 from Zhejiang University, Hangzhou, China, the M.S. degree in 2010 from Hanyang University, Seoul, South Korea, and the Ph.D. degree in 2015 from the University of Iowa, Iowa city, Iowa, USA, all in mechanical engineering. From February 2016 to September 2018, Dr. Hu was a postdoctoral fellow at Cornell University, Ithaca, New York, USA. His research interests include digital twin, design optimization under uncertainty, artificial intelligence, and wind energy. He has published more than 80 peer reviewed SCI papers and authored 2 English books published by Springer. He has been listed among the top 2% of the world's top scientists by Elsevier. He has been authorized 25 Chinese invention patents and 1 US patent. He has been participating in the formulation of 1 national standard for the design of wind power equipment. He won 1 Second Prize of Zhejiang Provincial Science and Technology Progress Award. He has served as the organizers and session chairs of several renowned domestic and international academic conferences, including the World Congress of Structural and Multidisciplinary Optimization (WCSMO), the Asian Congress of Structural and Multidisciplinary Optimization (ACSMO), and the International Conference on Machine Design of the Chinese Mechanical Engineering Society. He serves as an editorial board member of the SCI journal Wind Energy, an Associate Editor of Wind Energy Science, a Review Editor of Structural and Multidisciplinary Optimization, a member of the Committee on Renewable and Advanced Energy Systems of the American Society of Mechanical Engineers, the vice president of the Zhejiang Engineering Graphics Society, and a member of the Product Information Modeling Professional Committee of the Chinese Graphics Society.

Tangfan Xiahou
University of Electronic Science and Technology of China, China

Title: AI-Enabled Fault Diagnosis and Reliability Assessment of Control Rod Drive Mechanisms in Nuclear Reactors
Abstract: Control rod drive mechanism (CRDM) is the only moving mechanism in pressurized water reactors. Its primary functions include lifting, inserting, holding, and power-off release, thereby enabling the startup and shut down, and rapid shut down in emergencies for nuclear reactors. Its reliability is paramount for the economic and safe operation of the reactors. During the operational stage of CRDMs, the vibration and current data can be collected from the instrumentation and control system. However, due to complex working environment and the weak fault features in the condition monitoring data, fault diagnosis and reliability assessment of CRDM is challenging. This talk introduces some novel deep learning approaches to CRDM fault diagnosis and reliability assessment. For the CRDM fault diagnosis, we developed some sampling generation methods, such as generative adversarial network and diffusion model to address the imbalanced data challenge. For CRDM reliability assessment, we proposed a physics-informed probabilistic neural network (PIPNN) with multi-task learning method to assess the probability distribution of the operational time under multi-physics coupling environment. The proposed PIPNN not only provides accurate predictions of CRDM motion indicators but also reduces the need for extensive data collection. The proposed models’ accuracies are demonstrated through case studies based on simulation data and real condition monitoring data, showing that they outperform the state-of-the-art models.

Bio: He is currently an Associate Professor with the School of Mechanical and Electrical Engineering, University of Electronic Science and Technology of China. He has received the Ph.D. degree in mechanical engineering from UESTC in 2022. He has been doing research on reliability engineering and prognostics and health management (PHM) for nuclear power plants more than ten years. He has published more than 40 peer-reviewed papers in international journals such as IISE Trans. and IEEE Trans. on Reliab. He has received several international awards, including IEEE Reliability Society Student Achievement Award, IISE annual conference 2022 QCRE Best Track Paper Award Finalist, and Best Paper Award in APARM2018&QR2MSE2018. He currently serving as the guest editor of the Reliability Engineering and System Safety and the Editorial Board of International Journal of Reliability and Safety and Journal of Reliability Science and Engineering. He is also a member of IEEE, IEEE reliability society, and IISE.

Xinan Chen
University of Nottingham-Ningbo, China

Title: Large Language Model based Automatic Heuristic Design in Fault Detection
Abstract: For decades, the design of effective fault diagnosis algorithms has been an artisanal process, demanding deep domain knowledge and extensive manual tuning. This talk will explore how Automatic Heuristic Design (AHD) is transforming this craft into a scalable science. We will journey from the foundational principles of AHD to the current revolution powered by Large Language Models (LLMs), which have elevated the field from merely optimizing code to optimizing the very strategies behind diagnostic reasoning.

This new wave of AHD can autonomously discover sophisticated heuristics for interpreting complex sensor signals and identifying subtle failure signatures that elude traditional methods. However, the most profound shift lies ahead. I will present a forward-looking vision for the next generation of LLM powered AHD, where diagnostic algorithms will break free from the constraints of purely numerical data. By fusing quantitative sensor readings with the rich, contextual information buried in non-numerical sources—such as technician maintenance logs, operator comments, and system manuals—AHD is poised to create truly holistic and context-aware diagnostic systems. This leap from signals to semantics will redefine the boundaries of system reliability and safety.

Bio: Dr. Xinan Chen is an Assistant Professor in the School of Computer Science at the University of Nottingham Ningbo China. He previously served as an Assistant Professor at the International Business School of Xi'an Jiaotong-Liverpool University. Dr. Chen holds a Ph.D. from the University of Nottingham, a Master's degree from the University of Liverpool, and a Bachelor's degree from the University of Electronic Science and Technology of China.

His research focuses on applying advanced AI technologies to solve complex, real-world optimization challenges, with a particular emphasis on system reliability, robustness, and decision-making under uncertainty. In practice, Dr. Chen contributed to the development of the reliability and maintenance system for the Shenguang-III (SG-III) laser facility, a major national scientific installation. He has also collaborated with Honda and Ningbo Port on research to enhance the robustness of production line supply chains under uncertainty.

To tackle complex fault diagnosis, he has conducted in-depth explorations into leveraging multi-modal data for failure detection and has separately developed MELA, an automated algorithm design tool. Beyond manufacturing and port logistics, his research extends to interdisciplinary areas including transportation systems, optical device design, and protein optimization. Through these innovative AI tools, Dr. Chen aims to enhance the operational performance and resource allocation of complex industrial and scientific systems while ensuring their reliable operation and intelligent automation.

Tian-li Huang
Central South University, China

Title: Time-dependent Reliability Assessment and Time-varying Force Identification for Stay Cables in Cable-stayed Bridges
Abstract: Cables perform a crucial role in cable-stayed bridges due to their excellent load-bearing capacity. However, their continuous exposure to dynamic loads and harsh environmental factors leads to various forms of damage, including prestress loss, corrosion, fatigue, and wire breaking. These damages compromise the structural integrity of cables and pose significant safety risks to the entire bridge. Therefore, the condition assessment and the time-varying force identification for stay cables are two key issues for the reliability and safety of cable-stayed bridges.

Existing reliability methods for assessing the corrosion fatigue of stay cables in cable-stayed bridges are predominantly based on fatigue damage accumulation theory. However, these methods fail to account for the randomness of load variations and do not adequately incorporate the uncertainties associated with the fatigue performance of corroded steel wires. To address these shortcomings, a novel framework for time-dependent reliability assessment of corrosion fatigue in stay cables is proposed, which takes into consideration multiple sources of uncertainty. Firstly, the corrosion levels of steel wires based on the rate of corrosion mass loss are clarified and the future corrosion states are predicted using a Markov chain model. Then, using the WIM monitoring data and finite element models, a random traffic flow model is established and the stress time histories are extracted. Utilizing the Miner’s linear cumulative damage theory, a time-dependent reliability model for the corrosion fatigue of a single steel wire is derived and the correlations in the failure processes of steel wires within the cable are considered. Furthermore, a time-dependent reliability model for corrosion fatigue of stay cables based on a dynamic Bayesian network (DBN) is established. Finally, the corrosion fatigue reliability analysis on a typical real cable-stayed bridge is demonstrated the proposed framework.

Most current methods for identifying the cable force can only provide an average cable force over a period. The vibration method correlates the cable force with cable frequency via a formula, transforming the time-varying cable force identification into the instantaneous frequency (IF) identification of the cable. Synchroextracting transform (SET) is a powerful time-frequency method with the ability to identify the cable’s IF from the measured response. However, the performance of SET in identifying the IF is heavily influenced by the choice of window length, limiting its practical application. To overcome these limitations, an improved SET (ISET) is proposed to identify the time-varying force in bridge cables. In ISET, the variational mode decomposition is first employed to decompose the measured response into several mono-component modes. A hyperparameter optimization algorithm, adopting R′enyi entropy as the evaluation index, is then introduced to determine the optimal window length for each mono-component mode.

Furthermore, the proposed ISET with the optimal window length is used to identify the cable’s IF, thereby calculating the time-varying cable force using the vibration method formula. Cases involving finite element model of the cable, along with laboratory experiments on a scaled cable, show that the proposed method successfully addresses the limitations of SET and accurately identifies time-varying cable force. Finally, to validate the practicality of the proposed method, the field test data from an unsymmetric cable-stayed bridge is adopted. The results demonstrate that the proposed method is straightforward and efficient, capable of accurately monitoring the time-varying cable force in both new and existing cables of cable-stayed bridges.

Bio: Dr. Tianli Huang joined the Central South University (CSU), China in June 2007 after obtaining PhD from Tongji University, China. He is currently a Professor in Bridge Engineering in School of Civil Engineering at CSU. His main areas of research cover structural health monitoring, time-varying and nonlinear system identification, stochastic deterioration modelling, performance assessment for deteriorating bridge and optimum maintenance strategy. Dr. Huang is the PI of a few research projects related to bridge engineering funded by National Natural Science Foundation of China (NSFC). He is also the co-PI of a Newton Research Collaboration Programme project funded by the British Royal Academy of Engineering. He was awarded the Science and Technology Advancement Award (First Class/8) from the Chinese Ministry of Education in 2009. He has published over 90 technical papers in international and Chinese journals, including Probabilistic Engineering Mechanics, Mechanical Systems and Signal Processing and Engineering Structures etc. He is currently the paper reviewer of over 30 international and Chinese journals. Dr. Huang has more than 20 years' experience in China civil engineering industry as a licensed National Structural Engineer (First Class).

Piao Chen
Zhejiang University, China

Title: Enhancing Battery RUL Prediction: Domain Robustness & Regeneration Modeling
Abstract: Reliable remaining useful life (RUL) prediction of lithium‑ion batteries underpins proactive maintenance and lifecycle optimization. Two pervasive issues compromise prediction fidelity: (1) domain heterogeneity caused by batch‑to‑batch variations, and (2) non‑monotonic capacity regeneration events. This talk focuses on our dual contributions: a robust transfer‑learning ensemble that leverages early‑cycle kernel regression with domain‑distance–based weighting and transfer component analysis for cross‑batch alignment; and a monotone decomposition technique that segregates the capacity signal into a monotonically decreasing component and a regeneration term, each forecasted via Gaussian processes and deep autoregression for uncertainty‑aware RUL estimates. Results on multiple datasets demonstrate the efficacy of our approaches in real‑world scenarios.

Bio: Dr. Piao Chen is currently an Associate Professor at the ZJUI Institute, Zhejiang University. He previously served as an Assistant Professor in statistics at TU Delft, the Netherlands. His research interests include quality and reliability, statistical learning, and decision optimization. He has published over 30 papers in leading journals across management, engineering, and statistics, such as Management Science, Production and Operations Management, and IEEE Transactions on Information Theory. His work has received several Best Paper Awards at international conferences, including INFORMS QSR and SRSE.

Weiwen Peng
Sun Yat-sen University, China

Title: Towards Second-Life Battery Health Prognostics: A Physics-Feature Enhanced Deep Learning Framework
Abstract: With the booming growth of the new energy vehicle industry, recycling and disposal of retired power batteries have become an increasingly challenging issue, as improper treatment can cause severe environmental pollution and resource waste. Repurposing retired batteries for second-life applications offers a promising, economical, and environmentally friendly solution. However, due to nonlinear degradation patterns under variable working conditions, ensuring their reliable operation in second-life usage has become a critical challenge. Accurate health prognostics, specifically state of health (SOH) estimation and remaining useful life (RUL) prediction, plays a vital role in battery reliability and guiding their safe repurposing. In this talk, we are going to introduce a physics-enhanced deep learning framework for health prognostics, integrating a reduced-order electrochemical model with real-world monitoring data. The proposed approach reconstructs full charge-discharge cycles from incomplete cycling data and incorporates a feature alignment loss function to embed physics-based degradation behaviors directly into the deep learning model. This hybrid framework effectively improves the reliability of health management under diverse working conditions. The results demonstrate substantial improvements over existing methods, with a 54.12% improvement in SOH estimation and a 23% improvement in RUL prediction. This framework provides a scalable method for the health management and safe deployment of second-life LIBs in real-world applications.

Bio: Weiwen Peng is an Associate Professor in the School of Intelligent Systems Engineering, Sun Yat-sen University, China. Dr. Peng received the B.E., M.E. and Ph.D. degrees in Mechanical Engineering from the University of Electronic Science and Technology of China, Chengdu, China, in 2009, 2012 and 2015, respectively. From 2014 to 2015, he worked as a Visiting Graduate Researcher in University of California, Los Angeles. From 2016 to 2019, he worked as a Research Fellow in National University of Singapore, Singapore. Dr. Peng’s research interests include system reliability, Bayesian deep learning, prognostics and health management.

Ke Feng
Xi’an Jiaotong University, China

Title: Digital Twin-driven Health Management and Remaining Useful Life Prediction of the Gearbox Transmission System
Abstract: The gearbox transmission system plays a vital role in advanced manufacturing, aerospace, renewable energy, vehicle, and mining system. Its degradation and failure would cause unexpected economic loss and even serious accidents. For example, the degradation and failure of the gearbox will impair the performance of the machine tool, affecting the production quality and quantity significantly and resulting in enormous economic loss. Therefore, monitoring the health condition of the gearbox transmission system is of great significance. However, the gearbox transmission system usually operates in harsh working environments, and it is difficult to conduct regular manual inspections and maintenance. Thus, the use of advanced online algorithms to monitor the degradation status of the gearbox transmission system and predict its remaining useful life (RUL) can bring significant benefits to industry practices.

Digital twin (DT) is a virtual representation (mirror) of a physical structure or a system in real space along its lifecycles. Through real-time interaction between the virtual model and physical structure, the degradation status of the system and its RUL can be reflected and evaluated effectively. Thanks to its unique specialty, DT has recently received considerable attention from the research community. However, due to the complex structures and harsh operation conditions, research on DT-based gearbox transmission system RUL prediction is limited. Moreover, existing conceptual approaches have limitations in indicating the specific contact status and providing insights into the degradation stages of gearbox transmission systems, which greatly benefit RUL prediction.

To this end, this work presents a systematic and practical digital-twin technology for gearbox transmission systems RUL prediction, including the development of the realistic virtual model, real-time interaction between the virtual model and physical structures, and ‘transfer learning’ for a wider mechanical transmission system RUL prediction. This work can significantly benefit the health management of the gearbox transmission system and bring significant benefits to various industrial applications, including advanced manufacturing equipment/machinery, industrial machinery, aerospace applications, and wind turbines.

Bio: Ke Feng, Professor at Xi'an Jiaotong University, is a Marie Curie Fellow and ranked among the “Stanford/Elsevier Top 2% Scientists”. He earned his bachelor's and master's degrees from the University of Electronic Science and Technology of China and his Ph.D. from the University of New South Wales. He has held positions at renowned institutions such as the University of British Columbia, the National University of Singapore, and Imperial College London. His research areas include digital twins, signal processing, fault diagnosis, fatigue, and wear analysis, among others. In 2023, he was awarded the title of “Emerging Leader” by the Royal Physical Society Journal.

He currently serves as an Associate Editor and Editorial Board Member for several international journals, including IEEE Transactions on Industrial Informatics, Information Fusion, IEEE Internet of Things Journal, and Structural Health Monitoring. His research findings have been published in prestigious academic journals such as IEEE Transactions on Fuzzy Systems & Mechanical Systems and Signal Processing. He has led numerous international collaborative projects, including the Horizon Europe, UKRI projects, the National Natural Science Foundation Excellent Young Scientists Fund, and the key projects under the National Key Research and Development Program of China. He has also received the “Second Prize of the China Aviation Science and Technology Award” and the “Second Prize of the Vibration Engineering Society Science and Technology Award”.

Jiawen Hu
University of Electronic Science and Technology of China, China

Title: Optimal Abort Policy for Mission-Critical Systems under Imperfect Condition Monitoring
Abstract: While most on-demand mission-critical systems are engineered to be reliable to support critical tasks, occasional failures may still occur during missions. To increase system survivability, a common practice is to abort the mission before an imminent failure. We consider optimal mission abort for a system whose deterioration follows a general three-state (normal, defective, failed) semi-Markov chain. The failure is assumed self-revealed, while the healthy and defective states have to be inferred from imperfect condition monitoring data. Due to the non-Markovian process dynamics, optimal mission abort for this partially observable system is an intractable stopping problem. For a tractable solution, we introduce a novel tool of Erlang mixtures to approximate non-exponential sojourn times in the semi-Markov chain. This allows us to approximate the original process by a surrogate continuous-time Markov chain whose optimal control policy can be solved through a partially observable Markov decision process (POMDP). We show that the POMDP optimal policies converge almost surely to the optimal abort decision rules when the Erlang rate parameter diverges. This implies that the expected cost by adopting the POMDP solution converges to the optimal expected cost. Next, we provide comprehensive structural results on the optimal policy of the surrogate POMDP. Based on the results, we develop a modified point-based value iteration algorithm to numerically solve the surrogate POMDP. We further consider mission abort in a multi-task setting where a system executes several tasks consecutively before a thorough inspection. Through a case study on an unmanned aerial vehicle, we demonstrate the capability of real-time implementation of our model, even when the conditionmonitoring signals are generated with high frequency.

Bio: Jiawen Hu is an associate professor in School of Astronautics and Aeronautic, University of Electronic Science and Technology of China, Chengdu, China. He received the B.S. degree in mechanical engineering from Shanghai Jiao Tong University, Shanghai, China, in 2009, the M.S. degree in mechanical engineering from Chinese Academy of Sciences, Beijing, China, in 2012, and the Ph.D. degree in industrial engineering from Shanghai Jiao Tong University, Shanghai, China, in 2017. He was a research fellow with the Department of Industrial Systems Engineering and Management, National University of Singapore from 2017 to 2020. His research interests include maintenance optimization, degradation modeling. His work has appeared in journals including Operations Research, Naval Research Logistics, IISE Transactions, IEEE Transactions on Industrial Informatics, IEEE Transactions on Reliability, Reliability Engineering & System Safety, Journal of Manufacturing Systems, International Journal of Production Research.

Dapeng Tian
Changchun Institute of Optics, Fine Mechanics and Physics, China

Title: Research on Dynamic Compensation Control in Optical Imaging and Related Issues
Abstract: Optical imaging requires a stable imaging environment, a stable optical system, and a stable object-image relationship. However, for modern optical systems to perform better and meet higher observation requirements, they often need to be mounted on moving carriers and image moving targets under complex movements and changes in multiple physical factors. Dynamic physical factors affect the optical imaging process, directly reducing the observation performance. It is necessary to compensate for the complex dynamic factors in the dynamic process of optical imaging. The presenter reported the latest research results in this direction in recent years. Taking the problem of space-based optical imaging as a typical research object, under the constraints of the optical imaging mechanism, multi-disciplinary cross-research is conducted from the perspectives of optics, mechanics, control science, and signal sensing to support the realization of high-performance dynamic optical imaging equipment manufacturing. Facing this challenging research, it is hoped to solve the contradiction between dynamic disturbances and the relative static conditions required by optical imaging, improve the imaging quality, and meet the increasingly higher application requirements.

Bio: Dapeng Tian: His is the director of the Department of Dynamic Imaging Foundation, Vice Director of the State Key Laboratory of Dynamic Optics Imaging and Measurement, Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences. He has graduated with a bachelor's degree from Beijing Institute of Technology, and a doctoral degree from Beihang University. He studied abroad in Japan as a co-research fellow from 2009 to 2011. Then, he has been working at the Changchun Institute of Optics, Fine Mechanics and Physics of the Chinese Academy of Sciences since 2012. He is the fellow of IET, senior member of IEEE, senior member of the Chinese Society of Optics Engineering. He is also a Technical Editor of IEEE/ASME Trans. on Mechtronics, member of the editorial boards of Optics and Precision Engineering, Light: advanced manufacturing, Infrared and Laser Engineering. He has received awards such as the National Science and Technology Progress Award First Prize (ranked 13th), the Young Scientist Award of the Chinese Academy of Sciences, the Outstanding Scientific and Technological Achievement Award of the Chinese Academy of Sciences, the Second Prize of the Technical Invention Award of Jilin Province (ranked 1st), and the First Prize of the Academic Achievement Award of Natural Sciences of Jilin Province (ranked 1st). He is also the chief scientist of the Leading Program of the Chinese Academy of Sciences. Has completed several tasks such as the National Excellent Young Fund and equipment pre-research.

Di Cao
University of Electronic Science and Technology of China, China

Title: Robust Decentralized Energy Management Strategy for Modern Distribution System Considering Anomalous Measurements
Abstract: Frequent and sizable voltage fluctuation, a common issue faced by the modern distribution system (DS), could lead to potential equipment failures and power interruption. This brings huge negative impact on the power supply reliability of the DS. Existing voltage regulation methods typically rely on the precise physical parameters, complete measurements, and perfect communication, all of these premises are difficult to meet in practice. To this end, a decentralized control method that is robust to measurement acquisition errors is developed for DS. Specifically, a graph learning-based surrogate network is first built to simulate the power flow computing procedure and capture the structural characteristics of the DS. The centralized surrogate model is, then, divided into several decentralized representation networks according to the network partition results to obtain the robust embedding of the regional information of each sub-network. Subsequently, the representation networks are embedded in the front of the actor networks of the multi-agent soft actor-critic algorithm, the agents of which are learned in a centralized fashion according to the reward value estimated by the centralized surrogate model. The systematic integration of the three components allows us to achieve cooperation between different sub-regions and robustness against anomalous measurements without the reliance on precise circuit parameters. Comparative studies on IEEE test system illustrate the robustness of the proposed approach.

Bio: Cao Di is currently an Associate Professor at the University of Electronic Science and Technology of China, specializing in research on urban power grid resilience. He has been selected for several prestigious programs and honors, including the Sichuan Provincial "Tianfu Qingcheng Plan" for Young Sci-Tech Talent, the Sichuan Provincial "Postdoctoral Innovative Talent Support Program" (one of only 25 recipients in the province), Stanford's Global Top 2% Most-Cited Scientists (2023–2025), and the "Young Talent Support Project" by the Chinese Society for Electrical Engineering. He has presided over multiple research projects, including the National Natural Science Foundation Youth Program (Category C), the Special Funding of China Postdoctoral Science Foundation, the General Funding of China Postdoctoral Science Foundation, and Open Funds from State Key Laboratories.

In the past five years, he has authored or co-authored 24 IEEE Transactions papers as the first or corresponding author, with one first-authored paper receiving the IEEE Transactions on Smart Grid Best Paper Award. Four of his papers are recognized as ESI Highly Cited Papers, and his work has accumulated over 4,800 Google Scholar citations. His accolades include the Tibet Autonomous Region First Prize for Scientific and Technological Progress (4/30), the China Instrument and Control Society Second Prize for Scientific and Technological Progress (2/15), and the First Prize for Scientific and Technological Progress in Electric Power Construction. He also serves as a Guest Editor for special issues of multiple SCI-indexed journals and as a peer reviewer.