Conference Agenda

During the COVID-19 pandemic, courier services for medical specimens have been crucial to the implementation of large-scale testing programs. Such courier services typically offer different order priorities, i.e., deadlines for the pickup and delivery of specimens to a laboratory. However, as the COVID-19 pandemic has shown, transportation demand can fluctuate dramatically, resulting in promised delivery deadlines may not always be feasible for all orders. To minimize the expected cumulative delay, we propose an anticipatory re-optimization approach by integrating an adaptive Cost Function Approximation (CFA) into a Large Neighborhood Search (LNS). The LNS iteratively optimizes the incumbent route plans upon placement of a new order. The CFA, in turn, guides the search of the LNS towards a balance between minimizing delays for pending express and standard orders and minimizing vehicle routing effort in favor of future ones. To this end, we consider the CFA not only in the evaluation of newly generated route plans, but also directly in the (re)insertion heuristic of the LNS. Moreover, we adapt the CFA depending on the current workload by applying Bayesian Optimization to learn the parameterization for the expected demand volume on the one hand and to make dynamic adjustments depending on its realization on the other. We will demonstrate the advantages of our approach in comprehensive computational experiments that will also provide insights into the characteristics of the problem, particularly with respect to the implications of different order priorities.

The 2022 FIFA World Cup in Qatar presented a unique logistical challenge, with up to four matches scheduled daily during the group stage. The close proximity of all locations put considerable strain on Doha's transportation infrastructure, affecting not only match attendees but also participants in parallel public events and festivals. Developing effective transportation plans that prioritize accessibility, sustainability, and safety required a nuanced understanding of the full spectrum of transportation demand.

This paper presents a newly developed, scalable, and comprehensive transportation demand simulation, that was coupled with a passenger counting and inflow control mechanism specifically designed for metro systems to approximate and manage transportation demand during large-scale events. The data-driven simulation is highly adaptable, encompassing several participant categories. The simulation framework draws on multiple large datasets to construct an accurate representation of the local population, event venues, transport hubs, accommodation options, and tourist attractions. Moreover, it considers variables such as the current load on transportation networks and event venues, as well as mode selection based on choice models.

In operation, our simulation successfully identified bottlenecks leading to delays and potential overloads across metro system segments, tourist hotspots, and event locations in preparation for the FIFA World Cup 2022 in Qatar. By using these insights, we developed targeted mitigation strategies and integrated them back into the simulation model to assess their impact. The results provided valuable guidance for policymakers and transportation planners, contributing to the overall safety and efficiency of the FIFA World Cup 2022.