Assoc. Prof. Song-Kyoo Kim, Macao Polytechnic University, China

Dr. Song-Kyoo (Amang) Kim received an M.S. degree in computer engineering and a Ph.D. degree in operations research from the Florida Institute of Technology in 1999 and 2002, respectively. He is currently an Associate Professor of the computing program at the Macao Polytechnic University, Macau, and a Research Scholar at Khalifa University, Abu Dhabi. He used to be an Associate Professor at several United Arab Emirates universities. Before moving to the Gulf Region, he was a Core Faculty Member of the Asian Institute of Management, providing courses in technology, innovation, and operations. Before his academic career, he was a Technical Manager with the Mobile Communications Division, Samsung Electronics, for more than ten years and mainly dealt with technology management in the information technology industry. He is the author of more than 70 research articles and ten patents relating to the mobile technology industries. He has been an Invited Speaker at many international conferences concerning technology management, innovation processes, operations research, and data sciences. He is also an external reviewer of various prestige journals including IEEE Access; ACM Transactions on Multimedia Computing, Communications, and Applications; and the Journal of Information Security and Applications.

Invited Speech: Novel Public Transport Prediction Systems with Dynamic Statistical Attention Technique

Abstract: A novel machine learning (ML) framework for accurate real-time bus arrival time prediction integrates the ML model with a lightweight Dynamic Statistical Attention (DSA) technique. Real-time GPS data from Macao bus routes, synchronized with weather records, constitute the primary dataset. The DSA technique blends ML predictions with historical average arrival times to mitigate random fluctuations and improve result stability. Experimental results indicate that the proposed models, particularly KNN combined with DSA, deliver superior performance through lower mean absolute error, root mean square error, and mean absolute percentage error relative to previous hybrid neural network approaches. This framework provides a computationally efficient solution suitable for deployment in resource-constrained smart city public transport systems while sustaining high prediction accuracy under varying traffic and weather conditions.

Assoc. Prof. Tianyu Li, Sanjiang University, China

Li Tianyu, Associate Professor of Sanjiang University, holds a Ph.D. in Civil Engineering and a postdoctoral degree in Hydraulic Engineering from Hohai University. He has published over 60 academic papers in core domestic and international journals, seven educational reform papers, and holds more than 30 authorized invention patents (including two international patents) and over ten computer software copyrights. He has authored two monographs, commercialized six research outcomes, led the drafting of one group standard, and participated in the formulation of one national standard and one group standard. He has delivered over ten academic presentations both domestically and internationally. He is the selected candidate for the 2023 Jiangsu Outstanding Postdoctoral Program and the 2025 Jiangsu Young Scientific and Technological Talent Support Project. Additionally, he serves as a review expert for the Academic Degrees & Graduate Education Development Center of the Ministry of Education, the Science and Technology Vice President of Jiangsu Wanstar New Materials Technology Co., Ltd., a peer reviewer for Engineering Structures, an expert committee member of the Ready-Mixed Concrete Branch of the China Concrete and Cement Products Association, a council member of the UHPC Branch of the China Concrete and Cement Products Association, a youth committee member of the Solid Waste and Ecological Materials Branch of the Chinese Ceramic Society, a member of the China Association for Standardization, a guest mentor at the Macau Science Publishing House - Strong Nation Youth Center, and a committee member of Youth Innovation and Exploration.

Invited Speech: Research Progress on Resilient Maintenance of Road Infrastructure in the Southern Coastal Region and AI-Based Decision-Making Technologies

Abstract: In the southern coastal region, highway infrastructure has long been subjected to the combined effects of high temperature and humidity, salt-mist corrosion, typhoon-induced storm surges, and soft-soil settlement, leading to accelerated structural degradation and heightened risks of functional disruptions. As a result, the resilience of lifeline systems faces severe challenges. The resilience-oriented maintenance paradigm focuses on the system’s ability to absorb, adapt to, and rapidly recover its functions under disturbances. By quantifying resilience through metrics such as robustness, redundancy, resource availability, and responsiveness, the maintenance objectives have shifted from solely controlling failure probabilities to minimizing functional losses during disasters and shortening recovery times. The rapid advancement of artificial intelligence (AI) technologies has provided sophisticated sensing, prediction, and decision-making tools for resilience-driven highway maintenance, thereby driving a transformative shift in maintenance practices.
In damage identification and condition assessment, deep convolutional neural networks, generative adversarial networks, and Transformer models have significantly improved the accuracy of surface damage detection. Long Short-Term Memory networks and autoencoders can capture anomalous features in time-series monitoring data, while Bayesian fusion methods enable component-level condition classification. For performance degradation prediction, physics-informed neural networks embed corrosion diffusion mechanisms into the learning framework, enhancing their extrapolation capabilities with limited data; deep reinforcement learning and multi-objective evolutionary optimization offer solutions for maximizing network-level resilience under high-dimensional states and complex constraints, while also addressing temporal scheduling and resource allocation limitations.
This report systematically reviews typical damage patterns and resilience maintenance needs for highways in the southern coastal region, summarizes the latest advances in AI-based decision-making technologies, and highlights adaptive improvements tailored specifically to the unique coastal environment: these include domain-adaptive transfer learning to mitigate perceptual data drift caused by salt-mist corrosion; Bayesian networks and graph neural networks for constructing cascading failure inference models that account for multiple disaster interactions and identify vulnerable nodes in the road network; physics-informed approaches (such as PINNs) for intelligent prediction of soft-soil foundation settlement and bridge pier scouring; and meta-learning and federated learning techniques to tackle the challenge of scarce degradation data. Finally, this report identifies key development directions—such as physics-data fusion-driven approaches, dynamic decision-making based on digital twins, interpretable human-machine collaboration, and integrated climate-change scenario modeling—as critical pathways for achieving comprehensive lifecycle resilience management. These developments aim to provide a systematic theoretical reference for resilient operation and intelligent decision-making in the infrastructure of southern coastal highways.