• ---

   

• Special Session #10

• Artificial Intelligence in Environmental Science

• Session Organizers:

Prof. Dr. Jin Zhang, Hohai University, China

Email: jin.zhang@hhu.edu.cn

Associate Prof. Dr. Yun Bai, Chongqing Technology and Business University, China

Email: yunbai@ctbu.edu.cn

Associate Prof. Dr. Pei Hua, South China Normal University, China

Email: pei.hua@m.scnu.edu.cn

Dr. Peifeng Li, Technische Universität Dresden, Germany

Email: peifeng.li@mailbox.tu-dresden.de

• Download: Special Session #10.pdf

Environmental science is an interdisciplinary field that focuses on studying the natural environment, including its physical, chemical, and biological characteristics, as well as the interactions between humans and the natural environment. Due to the nonlinear natural systems, implicit changes in the environment, and multifaceted environmental issues, there remain significant challenges in properly deciphering the complexity and dynamics of environmental science.

Artificial intelligence (AI) is on a steep upward trajectory, and its benefits are unearthed and implemented in business, traffic, and especially scientific research. The powerful nonlinearity and flexible model architecture of AI allow it to clarify complex relationships between variables and uncover underlying patterns behind large datasets with few computing resources. As a result, it has tremendous potential to address the challenges in environmental science and facilitate a better understanding of it. Based on recent research statistics, AI techniques have infiltrated many fields of environmental science, including the collection and analysis of informing environmental data, fast prediction of environment-related parameter values, chemical screening analysis, risk assessment and management, and environmental decision-making.

Accordingly, the primary purpose of this Special Session is to provide recent studies on novel machine learning approaches for tackling problems in environmental science. Theoretical and practical advancements in physics-informed and/or theory-guided machine learning approaches are also welcomed.

The topics of interest include, but are not limited to:

• Deep learning tools

• Novel machine learning algorithms

• Intelligent forecasting

• Uncertainty quantification

• Neural networks

• Water supply/distribution systems

• Data-driven techniques

• Water quality model

• Predicting classical and emerging contaminants

• Low carbon–water quality-based forecasting and decision making

• Special Session #11

• Human Activity and Scene Understanding

• Session Organizers:

• Prof. Liang Zhang, Xidian University

  Email: liangzhang@xidian.edu.cn

  Assoc. Prof. Guangming Zhu, Xidian University

  Email: gmzhu@xidian.edu.cn

  Prof. Mohammed Bennamoun, The University of Western Australia

  Email: mohammed.bennamoun@uwa.edu.au

•  

  Download: Special Session #11.pdf

•  

  Human activity and scene understanding is becoming a trendy area of research due to its wide range of applications such as security, surveillance, human–computer interaction, patients monitoring analysis system, sports and robotics. With the development of deep learning video analysis techniques, scene understanding, natural language processing, multimodal features (including appearance features, spatial features and semantic features) based on video frames, skeleton data and semantic labels, have been used to improve the performance of human activity and scene understanding. Vision Transformers and graph models have achieved exemplary performance for a broad range of computer vision tasks e.g., image recognition, object detection, segmentation, and image captioning. This special session seeks original contributions towards advancing the theory and algorithmic design for vision transformers and graph models, and focuses on presenting the state-of-the-art vision transformers and graph models based human activity understanding techniques that are developed for solving important problems in action/activity recognition, understanding, prediction, and so on.

•  

  Topics of interest include but are not limited to:

  • Vision Transformers

  • Graph Models

  • Action Recognition

  • Graph neural network

  • Action Predictions

  • Scene Understanding

  • Human Object Interaction

  • Object Recognition

•  

  ---

   

• Special Session #12

• Intelligent Energy Sensing and Conversion technologies and Devices

• Session Organizers:

Assoc. Prof. Yun Zhang, Xidian University, China

Email: yunzhang@xidian.edu.cn

Research Fellow Hongzhi Yao, Shaanxi Applied Physics-Chemistry Research Institute, China

Email: yhz305@163.com

Assoc. Prof. Donglin Zou, Shanghai Jiao Tong University, China

Email: zoudonglin.520@sjtu.edu.cn

• Download: Special Session #12.pdf

Almost all behaviors in nature are based on energy sensing and conversion, so as to realize the perception of external characteristics and convert one form of energy into another desired form of energy to meet the needs of the actual scene. Therefore, how to effectively detect and convert energy to become an effective means of information feedback or execution is crucial. In recent years, with the advancement of basic technology, energy sensing and conversion technology has developed rapidly. This session aims to show and share new ideas and achievements of the new development and application of intelligent energy sensing and conversion technologies and devices with relevant experts, scholars and engineers around the world, discuss sensing technology of different energy, conversion technology between different energy, new sensing, conversion and actuation devices, etc.

Topics of interest include but are not limited to:

• Micro Electro-thermal Conversion Elements

• • Microwave Energy Conversion Technologies
• • Vibration Energy Conversion and Devices
• • New Energy and Energy Storage Systems
• • Intelligent Information Sensing Methods
• • Microsensors and Microactuators
• • Micro Energy Harvesting Technologies and Devices
• • Fabrication of Energy Sensing and Conversion Devices
• • Electromagnetic Energy Coupling Testing
• • Vibration and Noise Energy Sensing
•  

• ---

•  

• Special Session #13

• Signal Decomposition-Integrated Mechanical Fault Detection and Remaining Useful Life Prediction

• Session Organizers:

Prof. Zong Meng, Yanshan University

Email: mzysu@ysu.edu.cn 

Prof. Yuejian Chen, Tongji University

Email: yuejianchen@tongji.edu.cn 

Dr. Rui Yuan, Wuhan University of Science and Technology

Email: yuanrui@wust.edu.cn

Download: Special Session #13.pdf

Practical mechanical systems usually operate under variable speed, load, and strong noise conditions, which complicate the vibration signal. The system is also subjected to multiple complex excitations, and the vibration signal contains many complex sub-signal components, where fault characteristic components are submerged. Particularly, the early weak fault features are easily overwhelmed by noise and interference components. Advanced signal decomposition algorithms are needed to decompose vibration signal into a series of sub-signal component with clear physical interpretation. The accurate fault characteristic extraction of univariate or multivariate signal is fundamental to mechanical fault detection (FD) and remaining useful life (RUL) prediction. Furthermore, when the deep learning model uses the original vibration signal as input, the strong noise and interference affect its efficiency greatly in extracting fault features. The time-varying characteristics of the vibration signal demands high requirements on the generalization ability of the model, making it difficult to guarantee the mechanical diagnostic accuracy under unknown working conditions. It is of great significance to integrate signal decomposition algorithms into deep learning to achieve end-to-end mechanical FD and RUL prediction.

The topics of interest include, but are not limited to:

• • Novel signal decomposition algorithms
• • Physical mechanism of mechanical fault diagnosis
• • Physical interpretation of signal decomposition algorithms and deep learning methods
• • Signal decomposition-integrated deep learning models
• • Signal decomposition-integrated deep learning for mechanical FD and RUL prediction
• • Multivariate signal decomposition-integrated mechanical FD and RUL prediction
• • Multi-sensor data fusion models
• • Applications of signal decomposition algorithms in mechanical fault diagnosis
•  

• ---

•  

• Special Session #14

• Digital Twin Empowered Intelligent Operation and Maintenance of Renewable Energy Equipment

• Session Organizers:

• Assoc. Prof. Haidong Shao, Hunan University

  Email: hdshao@hnu.edu.cn

  Assoc. Prof. Te Han, Beijing Institute of Technology

  Email: hante@bit.edu.cn

  Assoc. Prof. Yun Kong, Beijing Institute of Technology

  Email: kongyun@bit.edu.cn

•  
• Download: Special Session #14.pdf
•  

• As the world embraces sustainable energy sources, renewable energy equipment's efficient and reliable operation becomes paramount. The concept of digital twins, with its ability to create virtual replicas of physical assets, has shown remarkable promise in revolutionizing the way we manage and optimize renewable energy systems. This special session aims to explore the latest advancements and cutting-edge research in leveraging digital twin technology for intelligent operation and maintenance of renewable energy equipment. The proposed special session seeks to facilitate a dynamic exchange of ideas, innovations, and best practices within the realm of digital twin applications in renewable energy. Potential topics include but are not limited to the following:

   

• • Novel approaches to predictive maintenance and early anomaly detection using data-driven techniques, machine learning algorithms, and sensor fusion in digital twin-enabled renewable energy systems.
• • Exploration of optimization strategies and control methodologies that utilize digital twin models to enhance the performance and lifespan of renewable energy assets.
• • Integrating Internet of Things (IoT) devices and advanced sensors for real-time data acquisition and control in digital twin-based renewable energy operation and maintenance.
• • Leveraging AI and ML algorithms for intelligent decision-making in renewable energy systems, including fault diagnosis, load forecasting, and energy management.
• • Addressing the challenges of ensuring robust cybersecurity and safeguarding sensitive data in digital twin ecosystems for renewable energy equipment.
• • Exploring the integration of digital twin technology in hybrid renewable energy systems for improved stability and efficient operation.
• • Examining how digital twins contribute to the development of Industry 4.0 concepts and their impact on smart grid deployment in renewable energy infrastructure.
• • Case studies and practical applications of developing digital twin models for various renewable energy systems, including solar photovoltaic arrays, wind turbines, hydroelectric generators, and more.
•  

• ---

•  

• Special Session #15

• In-situ Inspection and Maintenance Robotics: Intelligent Sensing,Control, Locomotion, and Detection

• Session Organizers:

• Dr. Laihao Yang, School of Mechanical Engineering, Xi’an Jiaotong University, China  

  Email: yanglaihao@xjtu.edu.cn

  Dr. Jun Dai, School of Mechatronic Engineering, Beijing Institute of Technology, China

  Email: daijun@bit.edu.cn

  Dr. Xin Zhang, School of Mechanical Engineering, Southwest Jiaotong University, China

  Email: xylon.zhang@swjtu.edu.cn

  Dr. Yu Sun, School of Mechanical Engineering, Xi’an Jiaotong University, China  

  Email: yu.sun@xjtu.edu.cn

•  

• Download: Special Session #15.pdf
•  

• Intelligent inspection and maintenance are of great significance for the safety insurance of mechanical equipments in the field of aero-engine, aircraft, and nuclear power. However, the harsh inspection environment, long-narrow space, and scarce label data make it very challenging for the intelligent, efficient, and accurate inspection and maintenance of these equipments. Under this context, the concept of “robotics + autonomous intelligence” emerges in the field of inspection and maintenance for mechanical equipments, and attracts extensive attention from OEMs, end-users, and academics. The new paradigm of inspection and maintenance will promote robotics, sensors, and detection technology to an upper level due to the high demand on reachability in crammed space, multi-mode environment sensing, pattern identification, and classification accuracy. This topic mainly addresses the emerging theory and technology in the field of intelligent robotized in-situ inspection and maintenance for high-end mechanical equipments, and focuses on intelligent robotics, multi-mode sensing, and explainable machine learning, which is intended for intelligent, efficient, and accurate inspection and maintenance.

   

  The topics of interest include, but are not limited to:

• • Latest progress of state-of-the-art technologies in the field of robotics, sensing, control, and pattern identification for intelligent inspection and maintenance
• • Emerging robotics technologies such as soft robotics, micro-crawling robotics, and bio-inspired robotics for intelligent inspection and maintenance
• • Embedded intelligence for robotics
• • Intelligent fault diagnosis and health monitoring of the key components (gear, bearing, etc.) in industrial robotics
• • Intelligent control and path planning for in-situ inspection and maintenance
• • Intelligent sensing technologies such as flexible sensing, multi-mode sensing integration, and damage imaging method
• • Deep learning methods in vision-based damage detection
• • Explainable machine learning in the field of intelligent structure design, control, and detection 
•  

• ---

• Special Session #16

• Multi-Physical Sensors-Based Detection & Diagnosis Solution Empowered by AI Technology

• Session Organizers:

• Prof. Yanpeng Cao, Zhejiang University

  Email: caoyp@zju.edu.cn

  Prof. Xin Xiong, Shanghai University

  Email: xxiong@shu.edu.cn

  Dr. Jun He, Zhejiang University

  Email: hjshenhua@zju.edu.cn

•  

• Download: Special Session #16.pdf

To precisely evaluate the health status of mechanical devices, a significant amount of attention has been paid to multi-physical sensors-based detection & diagnosis technology, including vibration, laser ultrasonic, infrared thermography, acoustic emission, and their combinations. Artificial intelligence, with its powerful data processing capabilities, provides more possibilities for the accurate detection & diagnosis of mechanical faults. This special session aims to share and envision the achievements and innovative ideas of integrating artificial intelligence to enrich our tools for fault detection & diagnosis based on multi-physical sensors.

Topics of interest include, but are not limited to:

• Artificial intelligence in multi-physical sensors fusion modeling

• Artificial intelligence in laser ultrasonic testing

• Artificial intelligence in infrared thermography detection

• Artificial intelligence in vibration testing

• Artificial intelligence in acoustic emission testing

• Artificial intelligence in industrial data processing

• Multi-vibration signals fusion modeling

• New technological routines for multi-physical sensors-based detection & diagnosis

• Artificial intelligence in other NDT methods

• Multi-physical sensors-based detection & diagnosis of critical equipment

•  

• Special Session #17

• Principles, methods and applications of photoelectron spectroscopy analytical instruments

• Session Organizers:

Chair：

Prof. Yu Chen, Xi’an Jiaotong University,

E-mail: chenyu@xjtu.edu.cn

Co-Chair：

Prof. Shulin Liu, Institute of High Energy Physics, Chinese Academy of Science,

E-mail: liusl@ihep.ac.cn

Prof. Yu Huang, Changchun Institute of Optics, Fine Mechanics and Physics, Chinese Academy of Sciences,

E-mail: ssshycn@163.com

Prof. Zicai Shen, Beijing Institute of Spacecraft Environment Engineering, China Academy of Space Technology,

E-mail: zicaishen@163.com

Dr. Zhengxu Huang, Guangzhou Hexin Instrument Co., Ltd.,

E-mail: zx.huang@hxmass.com

• Download: Special Session #17.pdf

As the demand for high-performance materials in critical fields such as aerospace, semiconductor chips, and photovoltaics continues to grow, the importance of material surface processing is also increasing. Photoelectron Spectroscopy instrument and analysis techniques, such as PYS, UPS, XPS, IPES, etc., play a significant role in understanding the surface chemical properties of materials, investigating surface modifications, and developing new devices. This special session aims to bring together experts and scholars in the research and application of photoelectron spectroscopy instruments to discuss the latest progress of different photoelectron spectroscopy instruments and their applications in various fields.

Research interests include but are not limited to:

• Photoelectron spectroscopy instrument technology (PYS, XPS, UPS, IPES, etc.)

• Key components of photoelectron spectroscopy instrument

• Photoelectron spectroscopy analysis technology

• Surface characteristic analysis method

• Theoretical modeling and calculation in photoelectron spectroscopy field

•  Applications of photoelectron spectroscopy instrument technology

• Special Session #18

• Advanced Sensing, Monitoring and Diagnosis in Smart Grid

• Session Organizers:

• Chair：

• Associate Prof. Yu Gao, Tianjin University

• E-mail: hmgaoyu@tju.edu.cn

  Co-Chair：

• Associate Prof. Peng Wang, Sichuan University

•  E-mail: pwang@scu.edu.cn

  Associate Prof. Jiale Mao, Xi’an Jiaotong University

• E-mail: jmaoab2019@xjtu.edu.cn

  Dr. Shuang Wang, Xi’an Jiaotong University 

• E-mail: shuang@xjtu.edu.cn

•  

• Download: Special Session #18.pdf
•  

• With the deepening of clean and low-carbon transformation of energy, large-scale development and utilization of renewable energy, distributed energy, and energy storage are developing rapidly. Digital technology is used to empower the traditional power grid, continuously improve the IntelliSense ability, interaction level, and operation efficiency of the power grid, effectively support various energy access and comprehensive utilization, and continuously improve energy efficiency. Based on the research of new sensing, monitoring, and diagnosis technology, the modern information technology and advanced communication technology, such as mobile Internet and artificial intelligence, are fully applied to realize the comprehensive condition IntelliSense, efficient information processing, convenient and flexible application for key equipment in smart grid, such as transformers, rotating machines, switches, cables, power transmission lines, wind turbines, solar photovoltaic systems. This session is intended to focus on the intellisense, monitoring, and diagnosis in the smart grid.

   

  Research interests include but are not limited to:

  • Intellisense technology and its application for power equipment and renewable energy conversion system.

  • Conditioning monitoring and diagnosis for transformer, rotating machines, GIS/GIL, cable, power transmission line, wind turbine, photovoltaic system, etc.

  • Qualification, monitoring and diagnosis for power equipment connected with power electronics.

  • Application of image processing in the conditioning monitoring and diagnosis field.

  • Emerging new applications with 5G mobile internet and IoT in electricity.

  • Related research on edge computing, artificial intelligence, and big data in smart grid.

•  

• ---

   

•  

• ​Special Session #19

• Artificial Intelligence for Sensor Data analytics

• Session Organizers:

• Zhenghua Chen, Institute for Infocomm Research (I2R) A*STAR, Singapore

  E-mail: Chen_zhenghua@i2r.a-star.edu.sg 

  Fangfang Yang, Sun Yat-sen University, China

  E-mail: Yangff7@mail.sysu.edu.cn

  Jun Zhu, Northwestern Polytechnical University, China

  E-mail: j.zhu@nwpu.edu.cn

  Xiaoli Zhao, Nanjing University of Science and Technology, China

  E-mail: xlzhao@njust.edu.cn

•  

   

• Download: Special Session #19.pdf
•  

• Sensor data analytics has gained significant attention due to the advancements in sensing and communication technologies. It holds great potential for various real-world applications, such as human activity recognition, machine health monitoring, prognostics and health management of energy systems, localization and tracking, etc. Conventional machine learning approaches have demonstrated success in sensor data analytics, typically involving two main steps: feature engineering and inference. However, the feature engineering process can be laborious, relying on domain knowledge and expert input, which may not always be readily available. Moreover, this manual approach might overlook implicit features, leading to performance limitations in conventional machine learning methods. To overcome these challenges, deep learning offers a promising solution by automatically learning latent features from raw sensor readings. Bypassing the need for explicit feature engineering, deep learning can uncover intricate patterns and relationships hidden in the data.

  This special session aims to showcase recent advances in artificial Intelligence for sensor data analytics. All submitted papers will undergo peer-review, and the selection will be based on both quality and relevance. Original innovative research papers addressing the following technology areas and their potential applications are invited:

• • Machine learning for sensor data analytics
• • Deep learning for sensor data analytics
• • Transfer learning for sensor data analytics
• • Self-supervised learning for sensor data analytics
• • Reinforcement learning for sensor data analytics
• • Knowledge distillation for sensor data analytics
• • Foundation models for sensor data analytics
• • Deep learning for system prognostics and health management
• • Application of equipment monitoring
• • Application of fault diagnosis
• • Application of battery prognostics
• • Application of activity recognition
• • Application of localization and tracking

•  

•  

•