18 Mar 2026 (Wed) 10:00-12:00 | Seminar by Prof. Stefan Minner

Decompose-route-improve framework for solving large-scale vehicle routing problems with time windows

Speaker:

Prof. Stefan Minner

School of Management, Technical University of Munich (TUM)

Date:    Mar 18, 2026 (Wednesday)

Time:   10:00 am – 12:00 nn

Venue: The Tam Wing Fan Innovation Wing Two, The University of Hong Kong

Abstract

Several metaheuristics use decomposition and pruning strategies to solve large-scale instances of the vehicle routing problem (VRP). Those complexity reduction techniques often rely on simple, problem-specific rules. However, the growth in available data and advances in computer hardware enable data-based approaches that use machine learning to improve scalability of solution algorithms. We propose a decompose-route-improve (DRI) framework, which first partitions the customers of the VRP with time windows (VRPTW) using clustering. Its dissimilarity metric incorporates customers’ spatial, temporal, and demand data and is formulated to reflect the problem’s objective function and constraints. Second, the resulting sub-routing problems are solved independently using any suitable algorithm. Lastly, we apply pruned local search (LS) between solved subproblems to improve the overall solution. Pruning is based on customers’ similarity information obtained in the decomposition phase. In a computational study, we parameterize and compare existing clustering algorithms and benchmark the DRI against a state-of-the-art solver on large VRPTW instances and very large-scale instances with up to 30,000 customers, which are introduced in this study. Results show that our data-based approach outperforms classic and recent cluster-first, route-second approaches as well as decomposition strategies that are solely based on customers’ spatial information. The newly introduced dissimilarity metric forms separate sub-VRPTWs and improves the selection of LS moves in the improvement phase. Thus, the DRI scales existing metaheuristics to achieve high-quality solutions faster for very large-scale VRPTWs by efficiently reducing complexity. Further, the DRI can be adapted to various solution methods and problem characteristics, such as the distribution of customer locations and demands, the depot location, and different time window scenarios, making it a generalizable approach to solving large- and very large-scale practical routing problems.

Following the research presentation, Professor Minner will also share his insights and experiences regarding research publications, drawing from his role as Editor-in-Chief of the International Journal of Production Economics.

About the speaker

Stefan Minner is a Full Professor for Logistics and Supply Chain Management at the School of Management, Technical University of Munich (TUM) and a core-member of the Munich Data Science Institute (MDSI). Currently, he is Vice Dean of Research and Innovation at TUM School of Management. His research interests using methods of operations research, artificial intelligence and machine learning are in global supply chain design, transportation optimization and inventory management. His work was published in many peer reviewed journals, including Management Science, Manufacturing & Service Operations Management, Operations Research, Production and Operations Management, Transportation Science, Transportation Research Parts B, C and E, European Journal of Operational Research, and the International Journal of Production Research. For his research output, he is currently listed among the top 1% business professors in German speaking countries by Wirtschaftswoche and among the top 2% researchers worldwide in a citation-based ranking by Stanford University. He serves on several editorial boards of logistics and operations research journals. Currently, Stefan Minner is the Editor-in-Chief of the International Journal of Production Economics. Stefan Minner is a fellow of the International Society for Inventory Research (ISIR) and the International Foundation of Production Research (IFPR). He received the PhD supervisory award at TUM School of Management in 2024 and the Science Award for his lifetime achievements by the German Operations Research Society in 2025.

Student Seminar by Mr. Zian Wang Peter on Mar 11, 2026, 1PM

Title: Toward Unified Risk Field Modeling for Interactive Autonomous Driving: A Comparative Study and Future Directions

Speaker: Mr. Zian Wang Peter (Department of Data and Systems Engineering)

Date: Mar 11, 2026 (Wednesday)

Time: 1:00 pm - 2:00 pm

Venue: Room 8-28, Haking Wong Building, The University of Hong Kong

Registration Link: https://forms.office.com/r/WssbpUgZaj

ITS Student Committee will provide a beverage for registered participants.

Abstract: Quantifying interaction risk in mixed traffic remains a fundamental challenge for autonomous driving systems. Traditional approaches rely on low-dimensional metrics such as time-to-collision (TTC), which struggle to capture the spatially distributed and temporally evolving nature of hazards in complex driving scenarios. Recent advances have proposed modeling risk as a continuous field over space and time, offering richer representations that naturally encode spatial uncertainty, multi-agent interactions, and environmental context. However, systematic comparisons of such risk field formulations remain limited, and no unified theoretical foundation has emerged to guide their design.

This talk consists of two parts. In the first half, we present a comparative study of representative risk field models from the literature. Using real-world bird's-eye-view (BEV) datasets spanning diverse interaction scenarios including highway merging, occlusion, and stop-and-go traffic. We then evaluate these models through a unified set of field-level metrics assessing spatial coherence, temporal consistency, and behavioral alignment with human driving decisions. Our analysis reveals that different modeling assumptions yield distinct spatial risk structures, with significant implications for downstream planning and prediction tasks.

In the second half, we discuss emerging directions toward unified field modeling that integrates physics-informed principles with data-driven methods like PINNs. We briefly introduce how partial differential equation (PDE)-based formulations, widely used in macroscopic traffic flow theory, can be adapted for microscopic risk propagation in autonomous driving contexts. Preliminary results on specifically designed robustness benchmarks will be shared, along with open discussions in bridging traffic flow modeling with real-time motion planning for intelligent transportation systems.

Bios: Mr. Zian Wang Peter is a first-year MPhil student in the Department of Data and Systems Engineering at the University of Hong Kong, supervised by Prof. Chen Sun. He received the BEng degree in Electronic and Information Engineering from The Hong Kong Polytechnic University in 2025, where he was advised by Prof. Ivan Ho Wang-hei with experiences on embedded systems, vehicular technology, and robotics. His research now focuses on integration of data-driven methods with risk-aware frameworks for improved prediction and safe planning in autonomous driving, and interdisciplinary topics within intelligent transportation systems.

Student Seminar by Mr. Yuxuan Zhou on Mar 4, 2026, 2PM

Title: Mobility, Segregation, and Inequalities: How experienced income segregation relates to travel behaviour and health inequalities

Speaker: Mr. Yuxuan Zhou (Department of Architecture & Civil Engineering, City University of Hong Kong)

Date: Mar 4, 2026 (Wednesday)

Time: 2:00 pm - 3:00 pm

Venue: Room 1010, CLL, Department of Geography, 10/F, The Jockey Club Tower, The University of Hong Kong

Registration Link: https://forms.office.com/r/4WLBeJ73Y5

ITS Student Committee will provide light refreshments and drinks for registered participants.

Abstract: Income segregation is a barrier to social inclusivity and equality. It is affected by individuals’ travel behaviour and socioeconomic contexts and may be intensified by localized living models emphasizing activities within immediate neighbourhoods. However, how the relationship between mobility-based experienced income segregation and travel behaviour varies across neighbourhood social and urban contexts remains unclear. Moreover, income segregation has been widely linked to social inequalities, including health inequalities. Yet most existing studies rely on residential segregation as a static measure of exposure, and rare studies have examined segregation–health relationships using a purely mobility-based approach that captures dynamic, real-world social exposure that may yield more accurate estimates. This seminar introduces two nationwide studies based on large-scale mobility data from the United States. We examine experienced income segregation and its associations with travel behaviour across different neighbourhood social and urban contexts, as well as how experienced segregation relates to income-related health inequalities. We find that longer travel distances and more diverse activity destinations are associated with lower levels of experienced segregation in less affluent neighbourhoods, particularly in less urbanized areas. In addition, higher levels of experienced segregation are associated with more pronounced income-related health disparities. These findings highlight potential trade-offs between localized living models and adverse social consequences and provide implications for how upward mobility, activity-based social mixing may be considered in future efforts to understand and address social inequalities associated with segregation. 

Bios: Yuxuan’s research focuses on exploring the complex relationships between the built/social environment, human behaviour, and health outcomes using multi-source spatial big data and advanced spatiotemporal statistical methods. His work lies at the intersection of health geography, GIScience, and urban studies. He has published over ten peer-reviewed articles in leading journals, including Nature Communications, Environmental Impact Assessment Review, Social Science & Medicine, and Applied Geography.