US Ignite: Collaborative Research: Focus Area 1: Social Computing Platform for Multi-Modal Transit

该项目通过直接吸引个人通勤者来解决城市交通和拥堵问题。由于智能设备的广泛使用，用户被建模为共享经济中的活跃代理，其算法旨在激励他们采取对整体交通系统有效的行动。许多商业上可获得的多式联运物联网解决方案侧重于从每个人的当地角度看什么是最适合他或她的。随着这些局部解决方案的数量增加，个人目标和整个系统之间的不一致也会增加。由于市政当局和用户收集了大量数据，但两者都没有资源来开发实时分析和控制，因此也形成了信息瓶颈。目前，在提供平衡包括商业公司、市政服务提供商和个人在内的多方需求的总体解决方案方面所做的工作很少。该项目将配置一个计算和信息共享平台，以克服个人和市政当局之间的激励差距。该平台向出行者提供混合模式的路线建议和一般系统信息，进而为服务提供商提供用户如何使用交通资源的高保真信息。该平台还通过充当虚拟公地，帮助改善社区对政策和监管决策的参与，个人公民可以在其中连接到市政服务提供商。该平台将适用于任何智能城市，并将在华盛顿州西雅图和田纳西州纳什维尔进行测试。研究议程分为三个关键推进：1)适合在分布式平台上实施的分层优化架构；2)从战略用户那里招募资源和激励混合模式路线的机制设计框架；3)软件定义的网络支持的社交计算平台，利用边缘设备进行计算。该研究通过引入事件的概率表示，使用子模块创建了一个接近最优的分布式算法，并将激励机制合并到优化问题中，扩展了现有的求解多模式交通问题的优化技术。此外，通过在社会背景下发展新的适应性契约和激励机制，研究结果将显著促进机制设计理论的发展。将在市政和行业合作伙伴的支持下进行真实世界的实验试验，以验证该平台和支持的算法。

This project addresses the problem of urban transportation and congestion by directly engaging individual commuters. Because of the widespread use of smart devices, users are modeled as active agents in a shared economy, with algorithms designed to incentivize them to take actions that are efficient for the overall transportation system. Many commercially available Internet of Things solutions for multimodal transit focus on what is best for each individual from his or her local perspective. As the number of these local solutions grows, the misalignment between objectives of individual and the overall system also grows. An information bottleneck also forms, since massive data is collected by municipalities and users, but neither has the resources to develop real-time analytics and controls. Currently, very little has been done to provide an overarching solution that balances the needs of multiple parties, including commercial companies, municipal service providers, and individuals. The project will configure a computing and information sharing platform that overcomes the incentive gap between individuals and municipalities. This platform offers mixed-mode routing suggestions and general system information to travelers and in turn provides service providers with high-fidelity information about how users are consuming transportation resources. The platform also help to improve community engagement in policy and regulatory decisions by serving as a virtual commons where individual citizens can connect to municipal service providers. The platform will be suitable for application to any smart city and will be tested in Seattle, WA and Nashville, TN.The research agenda divides into three key thrusts: 1) a hierarchical optimization architecture amenable to implementation on a distributed platform; 2) a mechanism design framework for recruiting resources from strategic users and incentivizing mixed-mode routes; 3) a software defined networking supported social computing platform that utilizes edge devices for computation. The proposed research extends existing optimization techniques for solving the multimodal transit problem by incorporating probabilistic representations of events, creating a near-optimal distributed algorithm by employing submodularity, and folding in incentive mechanisms into the optimization problem. In addition, the results will significantly advance the theory of mechanism design by developing novel adaptive contracting and incentive mechanisms in a societal setting. Real world experimental trials will be conducted with the support of municipal and industry partners to validate the platform and supporting algorithms.