Data-driven optimization of hub locations for smart mobility

智能交通是指使用更清洁、更安全、省油和省时的交通方式，可以改善乘客习惯、交通网络效率、燃油经济性，并减少拥堵和排放。智能交通在加拿大大幅减少温室气体排放和实现安全、可靠、绿色、创新和综合交通系统的长期目标中发挥着重要作用。当前和新兴的智能移动实践包括拼车、电动滑板车（e-scooter）共享和卡车队列。对于所有这些应用程序，集线器位置以及整体网络设计对于操作和基础设施效率以及可访问性至关重要。该研究计划的总体目标是通过优化集线器位置，为智能移动网络提供高效和可持续的设计，并开发集线器网络设计中数据驱动决策支持的分析工具和方法。短期目标侧重于优化三种不同的智能移动应用的中心位置：乘车共享、电动滑板车共享和卡车队列。对于每个应用程序，将通过分析实际数据确定相应的优化问题，并根据这些数据得出的信息开发数学模型。分解方法将被实施，以开发这些具有挑战性的组合优化问题的有效解决算法。除了确定性变量，本研究项目中探索的枢纽位置问题也将在不确定性下建模和解决，使用随机和鲁棒优化技术，以开发弹性枢纽网络。开发的模型和算法将通过现实数据集的关键绩效指标进行测试和验证，包括纽约市出租车，多伦多和蒙特利尔的自行车共享数据。这项研究项目的结果将为智能移动网络的有效设计提供见解，并将对未来的交通产生影响。开发的模型和算法将通过会议和在知名期刊上的出版物进行分享。在这个项目中接受培训的学生毕业时将拥有数据科学和优化方面的技能，可以在学术界、供应链和物流行业工作。

Smart mobility refers to using cleaner, safer, fuel- and time-efficient modes of transportation that can lead to improvements in ridership habits, transit network efficiency, fuel economy, and reduce congestion and emissions. Smart mobility plays a big role in Canada's long-term ambition to drastically reduce greenhouse gas emissions and achieve a safe, secure, green, innovative, and integrated transportation system. Current and emerging smart mobility practices include ridesharing, electric scooter (e-scooter) sharing, and truck platooning. For all these applications, hub locations, and therefore the overall network design, is crucial for operational and infrastructure efficiency and accessibility. The overarching goal of the proposed research program is to provide efficient and sustainable designs for smart mobility networks by optimizing hub locations and to develop the analytical tools and methods for data-driven decision support in hub network design. Short-term objectives focus on the optimization of hub locations for three different smart mobility applications: ridesharing, e-scooter sharing, and truck platooning. For each application, the corresponding optimization problems will be identified through analyzing real-life data and mathematical models will be developed based on information derived from those data. Decomposition methodologies will be implemented to develop efficient solution algorithms for these challenging combinatorial optimization problems. In addition to the deterministic variants, the hub location problems explored within this research program will also be modelled and solved under uncertainty, using stochastic and robust optimization techniques, to develop resilient hub networks. The developed models and algorithms will be tested and verified through key performance indicators on real-life datasets, including New York City taxi, Toronto, and Montreal bike-sharing data. The results from this research program will provide insights into the efficient design of smart mobility networks and will have an impact on the future of transportation. The developed models and algorithms will be shared through conferences and publications in respectable journals. Students trained in this program will graduate with sought-after skills in data science and optimization to work in both academia and the supply chain and logistics industry.