This presentation introduces the Korean UAM (K-UAM) national R&D program, a comprehensive initiative developing the core technologies required for safe and scalable UAM operations. The program spans airspace design, surveillance, vertiport infrastructure and operations, traffic flow management, weather prediction, airworthiness certification, and flight demonstration.

Within this framework, we present the safety risk management component, which addresses a critical gap in current UAM development. As most existing efforts focus on optimizing nominal operations, little attention has been paid to abnormal and unforeseen conditions. We introduce a proactive approach that generates synthetic data to simulate abnormal scenarios and validate the robustness of to-be-developed UAM algorithms.

Findings from preliminary research collectively demonstrate that robust UAM safety requires proactively engineering for the unknown, not solely optimizing nominal operations.

Dr. Jungwoo Cho is an Assistant Professor in the Department of Aerospace Engineering at Inha University. His research centers on modeling, simulation, and data analytics to advance the safety of aviation and Urban Air Mobility (UAM). His current work focuses extensively on data-driven Safety Risk Management (SRM) for UAM, including ground risk assessment frameworks and the integration of advanced AI technologies into aviation safety systems.

He received his B.S., M.S., and Ph.D. in Civil and Environmental Engineering from KAIST in 2014, 2015, and 2020, respectively. Prior to his current academic appointment, Dr. Cho served as an Associate Research Fellow in the Air Transport Department at the Korea Transport Institute (KOTI). His research has been widely published in leading academic journals, including Transportation Research Part C, Transportation Research Part D, IEEE Transactions on Intelligent Transportation Systems (TITS), and Transport Policy.

The Traveling Salesman Problem with Drone (TSP-D) models cooperative delivery systems in which a truck and a drone jointly serve customers. This talk presents recent advances in algorithms, neural acceleration, and probabilistic analysis for TSP-D. We introduce the Iterative Chainlet Partitioning (ICP) algorithm, which iteratively improves chainlet segments using a dynamic programming subroutine, achieving substantial performance gains on large benchmark sets. To further accelerate the method, we develop Neuro-ICP, which integrates a graph neural network to predict promising improvements and reduce expensive subroutine calls. We also extend the Beardwood–Halton–Hammersley theorem to TSP-D, establishing the asymptotic scaling of the optimal makespan and deriving bounds for Euclidean and mixed metric models. Together, these results provide both efficient solution methods and theoretical insights for large-scale truck–drone routing systems.

Changhyun Kwon (권창현) is a Professor in Industrial and Systems Engineering at KAIST. His research aims to advance computational optimization methods for efficient transportation and logistics systems. His current focus is to improve the efficiency of heuristic and exact algorithms using machine-learning approaches to solve large-scale vehicle routing problems and mobility service operations problems. He received a Ph.D. in Industrial Engineering in 2008 from Penn State and a B.S. in Mechanical Engineering from KAIST in 2000. His research has been published in Operations Research, Transportation Science, Transportation Research Part B, INFORMS Journal on Computing, etc. Before joining KAIST, he was a faculty member at the University at Buffalo and the University of South Florida. He was the Chair of the Urban Transportation SIG of the INFORMS TSL Society and the International Liaison for Asia/Oceania. He wrote the book Julia Programming for Operations Research, and he is a member of the JuMP steering committee, a NumFOCUS-sponsored open-source community for developing mathematical optimization tools in Julia. He is a recipient of the NSF CAREER Award, and his research has been funded by the National Science Foundation, the U.S. Department of Transportation, the National Research Foundation of Korea, and several industrial partners.

This study aims to identify and classify expected off-nominal situations that urban air mobility (UAM) aircraft may encounter when operating in high-density urban airspace within the K-UAM urban entry environment. Drawing on expert surveys, the study focuses on potential hazards such as communication delays, adverse weather conditions, and bird strikes, analyzing the progression of off-nominal situations in a linked manner and presenting possible outcomes. This research is intended to serve as a foundational reference for future UAM research and development, and can also inform the establishment of risk assessment frameworks.

Dr. Seyun Kim is an Associate Research Fellow in the Department of Air and Space Transportation at the Korea Transport Institute (KOTI). He received his Ph.D. from the Department of Civil and Environmental Engineering at the Korea Advanced Institute of Science and Technology (KAIST) in 2021, and subsequently served as a Research Assistant Professor at KAIST until 2025 before joining KOTI. His research interests span aviation data analytics, artificial intelligence applications in aviation, and urban air mobility (UAM). Through his work, Dr. Kim aims to advance data-driven approaches and intelligent systems that address emerging challenges in the rapidly evolving aviation sector.

Conflict management in air traffic management (ATM) traditionally consists of strategic planning and tactical avoidance. However, such a two-level structure is insufficient for the UAV System Traffic Management (UTM) context because UTM operations are characterized by mixed scheduled and instant demands. The instant demand cannot be captured by strategic Demand Capacity Balancing (DCB), and will give too much computational burden on the tactical level for self-separation. To address this limitation, a pre-tactical conflict management is proposed between the strategic and tactical levels. Operating within a short planning horizon prior to departure, this layer performs centralized conflict management to proactively mitigate large-scale conflicts and reduce real-time tactical workload. To enable efficient and scalable decision making, three key components are developed. An adaptive airspace abstraction transforms submitted free-flight trajectories into a structured conflict graph, balancing operational flexibility and computational tractability. The pre-tactical conflict management problem is formulated as a multi-agent contextual bandit. A Centralized Training Decentralized Execution (CTDE) learning framework outputs coordinated actions for conflict resolution, including ground delay, speed adjustment, and altitude rerouting. Numerical experiments demonstrate effective conflict resolution performance and strong scalability across varying traffic densities. The proposed architecture provides a structural extension to existing UTM systems for managing dynamic operations.

Dr. Ang Li is currently an Assistant Professor in the Department of Aeronautical and Aviation Engineering, The Hong Kong Polytechnic University. She is the director of Aviation Next Generation and Low-altitude Ecosystem (ANGLE) lab. She is affiliated with the Research Center for Low-Altitude Economy (RCLAE). She also serves as deputy program leader of M.Sc. in Aviation Engineering and Operations Management. Prior to this, Ang received her Ph.D. and M.Sc. Degree in Civil Engineering at the University of California, Berkeley, and her BEng in Transportation Engineering at Tongji University. Her research areas focus on Multimodal transport network modeling, Unmanned Aircraft Systems Traffic Management (UTM), Air Traffic Management (ATM), mechanism design and resource allocation, using operations research and machine learning. Ang teaches air traffic management, airport operations, data science and optimization applications. She has collaborated with NASA and co-developed the course ‘Aviation Data Science’ taught at UC Berkeley and NASA AMES. She received several Best Paper Awards from ICNS, INFORMS, and ATRSCC. She serves as associate editor of AIAA Journal of Air Transportation. Ang also received the Robert Wadell Endowed Fellowship for Engineering Innovation in 2022. She is selected as Rising Star in CEE by Carnegie Mellon University in 2022.

Runway capacity is a crucial determinant of the efficiency of near-ground flight operations influenced by various operational constraints. In particular, aircraft wake separation is a key constraint towards enhanced runway throughput. Compared to the traditional aircraft wake separation standard set forth by the International Civil Aviation Organisation (ICAO), both the European Union Aviation Safety Agency (EASA) and the Federal Aviation Administration (FAA) have suggested reclassifying aircraft types to reduce wake separation. Hence, dynamic wake separation based on weather and aircraft pairs remains an active area of research. Furthermore, the effects of dynamic wake separation on terminal traffic control deserves to be investigated. This study introduces two deep-learning models designed to predict dynamic aircraft wake separation using Light Detection and Ranging (LiDAR) data and aviation weather reports (METAR) at Hong Kong International Airport (HKIA). We also present two tiers of wind-related wake separation matrices and compare them with the RECAT-EU standards. The impact of dynamic wake separation on terminal arrival flight management is assessed under both high- and low-traffic scenarios. To tackle the challenges, we propose an integrated solution to predict wake decay performance and project the weather hazards from weather radar and other meteorological data source, then we further optimise the approach flight path. Our results indicate that a reduced wake separation, especially dynamic pairwise separation during the final approach, may lead to congestion at the initial approach fix and increase scheduling pressure at this juncture. However, it can be alleviated through terminal flight path planning, resulting in enhancements in hourly runway arrival throughput by approximately 10% compared to traditional RECAT-EU standards with subtle increase in average flight delay. This advancement provides a promising strategy for the joint optimisation of terminal arrival control and runway scheduling, thereby mitigating supply-demand imbalance and enhancing operational efficiency for airports with constrained capacity and configuration.

Prof. Kam K.H. Ng is currently Associate Head and Associate Professor in the Department of Aeronautical and Aviation Engineering at The Hong Kong Polytechnic University, Hong Kong SAR. He is a Fellow of the Royal Aeronautical Society (RAeS), the Royal Meteorological Society (RMetS), and the Hong Kong Meteorological Society (HKMetS), as well as a member of the American Institute of Aeronautics and Astronautics (AIAA) and the Institute of Electrical and Electronics Engineers (IEEE).

His research interests encompass air traffic management, unmanned aircraft system traffic management, aviation meteorology, operations research, and deep learning. Key contributions include prescriptive optimisation under uncertainty and deep learning models for wake vortex decay and weather nowcasting. His work has demonstrated significant real-world impact, particularly in dynamic wake vortex prediction and separation standards. Utilising LiDAR data, flight tracks, and meteorological inputs—often drawn from observations at Hong Kong International Airport (HKIA)—this research directly supports enhanced runway throughput and safety. By leveraging local LiDAR observations in collaboration with stakeholders such as the Hong Kong Observatory (HKO), it advances operational efficiency, especially in the context of HKIA's three-runway system expansion and efforts to characterise wake vortex behaviour under local meteorological conditions.

Prof. Ng is ranked among the World’s Top 2% Scientists. He has secured major grants from the Research Grants Council, Hong Kong SAR, and his research team has obtained over HK$38 million (~ USD $4.8 million) in external funding over the past five years. His accolades include the 2025 INFORMS Air Transportation Section Best Dissertation Runner-up award (for a supervisee) and the 2025 AIAA Best Paper Award. He has authored and co-authored numerous papers in leading journals, such as Reliability Engineering & System Safety, Transportation Research Part A: Policy and Practice, Transportation Research Part C: Emerging Technologies, Transportation Research Part E: Logistics and Transportation Review, Journal of Air Transport Management, and Transport Policy.

This study presents a comprehensive overview of research on risk-based traffic management in aerospace systems, focusing on unmanned aircraft operations, urban airspace assessment, and emerging Advanced Air Mobility (AAM) environments. The research emphasizes integrating safety, efficiency, and operational uncertainty in complex urban airspace, where increasing traffic demand and environmental constraints pose significant challenges. A data-driven and risk-informed perspective is highlighted to support more adaptive and scalable traffic management, enabling more flexible decision-making under uncertainty. These efforts contribute to advancing next-generation air traffic management frameworks for low-altitude operations and provide practical insights for managing increasingly complex aerospace systems.

Soohwan Oh received the B.S. degree in Air Transportation and Logistics from Korea Aerospace University, Goyang, Republic of Korea, in 2016, and the Ph.D. degree in Civil and Environmental Engineering from the Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Republic of Korea, in 2023. He is currently an Assistant Professor with the Department of Aerospace Industrial and Systems Engineering at Hanseo University, Seosan, Republic of Korea. His research focuses on advanced air mobility, air traffic management, and space traffic management, particularly risk-aware and data-driven modeling using simulation and large-scale operational data.

The safe and scalable integration of Advanced Air Mobility (AAM) relies on effectively translating complex operational data into robust, risk-aware decisions. This talk presents a systematic approach to risk modeling and mission planning, bridging theoretical foundations with practical implementations for AAM decision making. We first introduce a multi-dimensional risk model that integrates first-party operational hazards with third-party exposures via probabilistic collision modelling. To operationalise this data, we discuss techniques for generating spatio-temporal risk maps, alongside model calibration and validation using both empirical and simulated datasets.

Transitioning from modelling to decision-making, the talk explores risk-aware mission planning algorithms that optimise flight trajectories. These formulations are designed to minimise cumulative risk while strictly satisfying operational constraints and balancing safety with efficiency in real-time AAM operations. Finally, we address the critical challenges of verification and validation (V&V) for risk-aware AAM systems, concluding with open research avenues in risk quantification, multi-agent coordination under uncertainty, and collaborative frameworks for cross-disciplinary integration.

Professor Hyo-Sang Shin is a Professor at the Korea Advanced Institute of Science and Technology (KAIST) and an Adjunct Professor in Guidance, Control, and Navigation Systems at Cranfield University. He leads the Autonomous and Intelligent Systems Group at KAIST, where his research focuses on advanced air mobility (AAM), urban air mobility (UAM), unmanned traffic management (UTM), sensor/data/information fusion, and data-driven guidance and control. He received his MSc in Aerospace Engineering from KAIST in 2006, with a specialization in formation flight of unmanned aerial vehicles, and earned his PhD from Cranfield University in 2011, focusing on cooperative missile guidance. Professor Shin has authored 3 books, 11 book chapters, over 110 journal papers, and more than 150 conference papers. He has been frequently invited to deliver lectures, keynote speeches, and technical talks at academic and industry venues around the world. He has participated in numerous research programs and served as coordinator of several international projects, contributing to various research projects. He is also actively engaged in the global research community, serving on various technical, program, and editorial committees.

Urban Air Mobility (UAM), a subdomain of Advanced Air Mobility (AAM), aims to add an aerial layer to urban transportation by operating electric vertical take-off and landing (eVTOL) aircraft between vertiports. Similar to ground transportation systems, UAM operations are constrained by infrastructure capacity, including take-off and landing pads, charging facilities, and parking spaces at vertiports. In addition, spatial and temporal demand imbalances can lead to severe operational bottlenecks, such as pad congestion or shortages of available aircraft. As demonstrated in our previous simulation studies, these imbalances can easily cause system gridlock.

This study develops a data-driven simulation and control framework for joint passenger matching, eVTOL scheduling and routing, and fleet repositioning in a multi-vertiport UAM network for improving the system performance. The simulation models detailed eVTOL states, per-leg energy consumption, vertiport capacity constraints, and passenger queues. Several heuristic policies are implemented as baselines, including FCFS dispatch, an adaptive dynamic-departure policy, and a pressure-based empty-vehicle repositioning policy. Building on these heuristics, we design deep reinforcement learning (DRL) schedulers that adjust two interpretable operational parameters: a departure-intensity multiplier and a repositioning-bias term. A factored dueling Deep Q-Network (DQN) observes aggregated network states and selects parameter settings to optimize system performance. Hyperparameters and reward weights are tuned through grid search and Bayesian optimization. Experimental results demonstrate that DRL controllers, combined with domain-informed heuristics and operational safeguards, can effectively support tactical UAM operations under realistic infrastructure and energy constraints.

Dr. Yu (April) Zhang is a Professor in the Department of Civil and Environmental Engineering at the University of South Florida, where she leads the Smart Urban Mobility Laboratory (SUM-Lab) and directs the Advanced Air Mobility Research Program at the Center for Urban Transportation Research (CUTR). Her research focuses on the development of mathematical programming models, solution algorithms, simulation tools, and machine learning methods to advance efficient, resilient, and sustainable multimodal transportation systems. More recently, her work has centered on the safe, efficient, scalable, and resilient integration of Advanced Air Mobility into both multimodal transportation systems and the national airspace system.

Dr. Zhang is a Senior Member of the National Academy of Inventors and a Senior Member of the American Institute of Aeronautics and Astronautics. She is also the recipient of the 2020 Amazon Research Award, a highly competitive and prestigious program supporting innovative research at academic institutions and nonprofit organizations worldwide.

Dr. Zhang received her Ph.D. and M.S. degrees in Civil and Environmental Engineering from the University of California, Berkeley, and her B.S. degree in Transportation Engineering from Southeast University in China.