LEAP-HI: On-Demand Multimodal Transit Systems

LEAP-HI：按需多式联运系统

• 批准号: 1854684
• 负责人: Pascal Van Hentenryck
• 金额: $ 176.71万
• 依托单位: Georgia Tech Research Corporation
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-08-01 至 2024-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1854684&HistoricalAwards=false
• 关键词: LEAP; HI; Demand; Multimodal; Transit

In the United States, car ownership is still the best predictor of upward social mobility. Those without a car are disadvantaged in accessing jobs, health care, education, and basic services, including buying groceries. Public transportation has the potential to mitigate congestion and provide environmentally friendly and cost-effective mobility. Existing systems, however, are often plagued by the "first/last mile" problem, i.e., the inability to take travelers all the way from their origin to their destination. As a result, travelers who can afford them prefer private vehicles, creating congestion and harmful emissions. New mobility services such as Uber and Lyft have improved transportation for some population segments by exploiting ubiquitous connectivity to match riders and potential drivers. Unfortunately, they increase congestion and emissions, prompting some cities to limit their numbers. In addition, those services are of limited use to low income citizens and may even lure affluent residents away from transit systems, reducing revenue. In contrast, this Leading Engineering for America's Prosperity, Health, and Infrastructure (LEAP-HI) Program award explores the concept of On-Demand Multimodal Transit Systems (ODMTS). Being multimodal, ODMTS combine on-demand mobility services that serve low-density regions with high-occupancy vehicles (buses or trains) traveling along high-density corridors. They differ from micro-transit solutions by planning, operating, and optimizing transit systems holistically, using state-of-the-art optimization technology and machine learning. As a result, they have the potential to transform accessibility across population segments, decreasing inequalities in transportation and providing a sustainable transportation model for American cities and beyond.To realize this vision, this award researches the data and decision science needed to engineer ODMTS for large, congested cities. This requires a step change in our ability to plan, operate, and optimize ODMTS, which are complex, socio-technical systems deployed over a sophisticated infrastructure. To achieve this objective, the award explores five research threads that address the following, high-level issues: (1) the delivery of scalable optimization and machine learning algorithms for designing the multimodal networks at the core of ODMTS, (2) accurate forecasting of ridership demand in ODMTS and integration of the resulting forecasting models into optimization models for network design, (3) generalization of network design optimization to account for and mitigate congestion, (4) co-design of ODMTS and infrastructure improvements to maximize the performance of transit systems, using concepts such as complete streets and context-sensitive solutions, and (5) the incremental integration of autonomous vehicles into ODMTS as they become available.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

在美国，拥有汽车仍然是社会向上流动的最佳预测因素。那些没有汽车的人在获得工作、医疗保健、教育和基本服务（包括购买杂货）方面处于不利地位。公共交通有可能缓解拥堵，提供环境友好和具有成本效益的流动性。然而，现有系统经常受到“第一/最后一英里”问题的困扰，即，无法将旅行者从出发地一直带到目的地。 因此，负担得起的旅行者更喜欢私家车，造成拥堵和有害排放。 Uber和Lyft等新的移动服务通过利用无处不在的连接来匹配乘客和潜在司机，改善了某些人群的交通状况。不幸的是，它们增加了拥堵和排放，促使一些城市限制它们的数量。 此外，这些服务对低收入公民的用处有限，甚至可能吸引富裕居民远离交通系统，从而减少收入。 相比之下，这个领先的工程为美国的繁荣，健康和基础设施（LEAP-HI）计划奖探讨了按需多式联运系统（ODMTS）的概念。作为多式联运，ODMTS将联合收割机按需移动服务与高乘坐率车辆（公共汽车或火车）沿沿着高密度走廊行驶的低密度区域相结合。它们与微交通解决方案的不同之处在于使用最先进的优化技术和机器学习来整体规划、运营和优化交通系统。因此，他们有可能改变跨人口段的可达性，减少交通不平等，并为美国城市和其他城市提供可持续的交通模式。为了实现这一愿景，该奖项研究了为拥挤的大城市设计ODMTS所需的数据和决策科学。 这就要求我们在规划、运营和优化ODMTS的能力上有一个飞跃性的变化，ODMTS是部署在复杂基础设施上的复杂的社会技术系统。为了实现这一目标，该奖项探索了五个研究主题，解决了以下高层次的问题：（1）提供可扩展的优化和机器学习算法，用于设计ODMTS核心的多模式网络，（2）准确预测ODMTS中的乘客量需求，并将所得预测模型集成到网络设计的优化模型中，（3）网络设计优化的通用化，以解决和缓解拥堵，（4）ODMTS和基础设施改进的协同设计，以最大限度地提高交通系统的性能，使用完整街道和上下文敏感解决方案等概念，以及（5）自动驾驶汽车逐渐融入ODMTS。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

项目成果

Exploring the Impact of Bike Lanes on Transportation Mode Choice: A simulation-based, route-level impact analysis

探索自行车道对交通方式选择的影响：基于模拟的路线级影响分析

• DOI: 10.1016/j.scs.2022.104318
• 发表时间: 2023
• 期刊: Sustainable Cities and Society
• 影响因子: 11.7
• 作者: Hwang, Uijeong;Guhathakurta, Subhrajit
• 通讯作者: Guhathakurta, Subhrajit

Reinforcement Learning from Optimization Proxy for Ride-Hailing Vehicle Relocation

网约车搬迁优化代理的强化学习

• DOI: 10.1613/jair.1.13794
• 发表时间: 2022
• 期刊: Journal of Artificial Intelligence Research
• 影响因子: 5
• 作者: Yuan, Enpeng;Chen, Wenbo;Van Hentenryck, Pascal
• 通讯作者: Van Hentenryck, Pascal

A generalized fluid model of ride-hailing systems

• DOI: 10.1016/j.trb.2021.05.014
• 发表时间: 2021-08
• 期刊: Transportation Research Part B: Methodological
• 作者: Zhengtian Xu;Yafeng Yin;X. Chao;Hongtu Zhu;Jieping Ye

Ridesharing and fleet sizing for On-Demand Multimodal Transit Systems

• DOI: 10.1016/j.trc.2022.103594
• 发表时间: 2021-01
• 期刊: Transportation Research Part C: Emerging Technologies
• 作者: Ramon Auad-Perez;P. Van Hentenryck

The Dual Effects of Team Contest Design on On-Demand Service Work Schedules

• DOI: 10.1287/serv.2023.0320
• 发表时间: 2023-03
• 期刊: Service Science
• 影响因子: 2.3
• 作者: Tingting Dong;Xiaotong Sun;Qi Luo;Jian Wang;Yafeng Yin