COmpetition models and cross-Subsidies for equitable and green MObility - COSMO

移动系统正处于革命的边缘，因为它们遭受着基础设施超载的痛苦，导致用户不满、污染、不平等加剧和健康危险。仅在伦敦，接触NO2每年就造成5900人死亡，医疗成本为1.4BGBP/年。出于这些原因，英国政府将未来移动性确定为四大挑战之一。同时，新形式的移动性和大数据的出现提供了非凡的机遇。在这种背景下，多式联运--不同的运输方式提供互补的服务--是一个很有前途的范例，因为它结合了高效的长途运输和最后一英里的服务。然而，它们的运营也带来了同样多的挑战。最值得注意的是，交通当局很难理解新的移动解决方案应该如何整合到现有的基础设施中，如何协调和监管它们，以及如何确定那些最终将提高公平性和减少系统范围拥堵的解决方案。这些挑战的核心是，私营移动提供商的目标往往与运输当局的目标不一致(例如，利润最大化与最大限度地减少拥堵/不公平)，导致与现有交通方式竞争，而不是补充它们。为了应对这些挑战，Cosmo旨在开发数学模型来描述移动提供商之间的竞争，分析这些模型，并利用它们来设计基于优化和合作诱导的补贴，以协调提供商的利益与改善公平性、最小化拥堵或两者的组合。更详细地说，第一个组成部分将提供三方面的内聚贡献：i)为不同移动提供商之间的竞争开发一个简洁的博弈论模型，ii)研究由此产生的均衡，以及iii)设计高效的均衡计算算法。在第一个组件的基础上，第二个组件将利用最近在优化和博弈论方面的突破来设计基于优化的交叉补贴，在最大化均衡性和最大限度地减少拥堵/排放之间进行权衡。这两个组件将最终发布一个开放获取算法套件，其有效性将在与项目合作伙伴共享和开发的关于美国城市和格林威治区的综合和真实世界案例研究中进行测试。本着这一呼吁的精神，研究将与智能移动(帕沃内博士，斯坦福大学和NVIDIA研究)和交通(奥索里奥博士，HEC蒙特利尔和谷歌研究)的领导者密切合作，与他们共同设计了一系列网络活动，包括延长对斯坦福大学和HEC蒙特利尔的访问，每天访问NVIDIA研究和谷歌研究的特邀讲座。

Mobility systems are on the brink of revolution as they suffer from an overloaded infrastructure causing users' dissatisfaction, pollution, increased inequality, health dangers. In London alone, exposure to NO2 accounts for 5900 fatalities/year, with healthcare costs of 1.4BGBP/year. For these reasons, the UK Government identified Future Mobility as one of the four Grand Challenges.At the same time, the advent of new forms of mobility and big data provides remarkable opportunities. In this context, Intermodal Mobility -- where different modes of transport provide complementary services -- is a promising paradigm, as it combines efficient long-distance transport with last-mile services.However, their operation has resulted in equally many challenges. Most notably, transportation authorities struggle to understand how new mobility solutions should be integrated within the existing infrastructure, how to orchestrate and regulate them in a cohesive way, and how to identify those that will ultimately improve equitability and reduce system-wide congestion. At its core, these challenges stems from the fact that privately-owned mobility providers often have objectives that are misaligned with those of the transportation authority (e.g., maximise profit vs minimise congestion/inequity), and result in competing with existing modes of transport as opposed to complementing them.To address these challenges, COSMO aims to develop mathematical models to describe the competition between mobility providers, to analyse these models, and to exploit them to design optimisation-based and cooperation-inducing subsidies to reconcile the providers interest with that of improving equitability, minimising congestion, or a combination thereof.More in details, the first component will deliver a threefold set of cohesive contributions: i) the development of a concise game-theoretic model for competition between different mobility providers, ii) the study of the resulting equilibria, and iii) the design of efficient equilibrium computing algorithms. Building atop the first, the second component will leverage recent breakthroughs in optimization and game theory to design optimization-based cross-subsidies that trade-off between maximising equitability and minimising congestion/emissions.These two components will culminate in the release of an open-access algorithmic suite, whose effectiveness will be tested on synthetic and real-world case studies on US cities and the Borough of Greenwich, shared and developed jointly with project partners. In the spirit of this call, the research will be carried out in close collaboration with leaders in smart mobility (Dr. Pavone, Stanford University & NVIDIA Research) and transportation (Dr. Osorio, HEC Montreal & Google Research), with whom a number of networking activities have been co-designed including an extended visit at Stanford University and HEC Montreal, daily visits and invited talks at NVIDIA Research and Google Research.