LIDS/IDSS Workshop on Smart URban Infrastructures. To Be Held in Boston, MA

智能传感器网络的进步、高带宽无线通信和智能手机的广泛使用使得在用户、资源和基础设施服务提供商之间整合网络信息层成为可能，从而使智能服务能够通过异构代理的分散决策实现更有效和动态自适应的资源分配。这些服务包括为用户提供实时信息以做出更好决策的新应用程序（例如，交通引导系统），在分散的参与者之间共享资源和管理交易的新平台（例如，优步和Lyft等乘车系统，纽约市的Citi Bike等自行车共享系统），以及新的分散控制方案，通过精心设计激励措施，旨在影响用户行为以更有效地利用资源(例如，电力需求响应计划、节能建筑、智能停车和通勤计划)。这场革命的另一个重要推动者是我们收集和处理大量数据的能力，这使我们能够建立用户特征和消费模式、同伴和社会网络效应的预测模型，并最终设计、测试和调整激励机制，以塑造用户行为。虽然这些服务被誉为真正的突破和提高社会生活质量的先兆，但它们的运作依赖于技术和社会因素的结合，可能导致意想不到和不可预测的后果：向大部分用户提供的实时交通信息导致某些路线极度拥挤，网络层的漏洞导致基础设施遭受严重破坏，大量数据收集和分析导致严重的隐私问题，战略行为和意外的社会同伴效应（通过物理和在线社交网络）导致羊群效应，时尚和低效的结果。这些系统的设计和运行需要一个整体和系统的框架，该框架结合了战略性的人类行为、技术限制、相互联系和网络的本质、隐私和安全影响以及由此产生的系统的弹性。构建这样一个框架需要系统地结合多个学科的工具和方法，包括优化和控制理论、经济学、博弈论和社会科学、统计学和数据科学，这是以前从未做过的。为了提供一个为这样一个框架奠定基础的论坛，组织者计划在麻省理工学院媒体实验室举办为期两天的“智能城市基础设施”研讨会，组织者将来自学术界，工业界和政府的顶级研究人员聚集在一起，分享和讨论这个令人兴奋的新兴领域的研究挑战，问题和最近正在进行的工作。信息与决策系统实验室（LIDS）是一个独特的地方主持和领导这样的讨论，因为它在核心信息和决策科学方面有着悠久的历史和严谨的研究声誉，最近它的研究领域扩展到包括博弈论、信息和网络经济学、自治、优化、推理和统计。LIDS最近还加入了新成立的麻省理工学院数据、系统和社会研究所（IDSS），并在其中发挥领导作用。研讨会的目标是通过在统计学、数据科学、信息和决策系统以及社会科学的交叉领域推进教育和研究来解决复杂的社会挑战，因此提供了解决与智能基础设施服务相关的挑战所需的多方面平台。研讨会围绕五个小组会议组织，还包括两位主题演讲者。特别注意在发言者和小组成员的性别和多样性方面取得平衡。所有小组成员和演讲者都站在这一新兴学科及其相关旗舰领域的最前沿。将有一个学生海报展示学生的作品。研讨会将被记录下来。通过展示和概述智慧城市基础设施研究机会的愿景，研讨会将把以前没有系统连接的想法和社区聚集在一起。通过明确考虑到演讲者和想法的多样性，研讨会将促进和促进进一步的跨学科联系，并将对社区产生广泛影响。

Advances in networks of smart sensors, high-bandwidth wireless communication and widespread use of smart phones allowed incorporating a cyber information layer among users, resources and infrastructure service providers, enabling smart services that achieve more efficient and dynamically adaptive allocation of resources through decentralized decisions of heterogeneous agents. Such services range from new apps that provide real-time information to users for better decisions (e.g., traffic guidance systems), new platforms for sharing resources and governing transactions among decentralized participants (e.g., ride sourcing systems such as Uber and Lyft, bike-sharing systems such as Citi Bike in New York City), and new decentralized control schemes which through careful design of incentives aim to affect user behavior for more efficient use of resources (e.g., electricity demand response programs, energy efficient buildings, and smart parking and commuting programs). Another important enabler of this revolution is our ability to collect and process vast amounts of data, which allows us to build predictive models of user characteristics and consumption patterns, peer and social network effects, and ultimately design, test, and adapt incentive mechanisms for shaping user behavior.While such services are hailed as veritable breakthroughs and harbingers of enhanced quality of life for society, their operation relies on a combination of technological and social factors and may lead to unintended and unpredictable consequences: real-time traffic information provided to a large fraction of users leading to extreme crowding of certain routes, vulnerabilities in cyber layer leading to extreme infrastructure damage, vast amounts of data collection and analytics causing serious privacy concerns, and strategic behavior and unexpected social peer effects (through physical and online social networks) leading to herds, fads and inefficient outcomes. Design and operation of these systems necessitates a holistic and systematic framework that incorporates strategic human behavior, technological constraints, the interconnected and networked nature of interactions, privacy and security implications and resilience of the resulting system.Building such a framework requires systematically combining tools and methodologies from several disciplines including optimization and control theory, economics, game theory and social sciences, and statistics and data science in a way that was not done before. To provide a forum for laying the foundations of such a framework, the organizers plan to host a 2 day workshop on "Smart Urban Infrastuctures" at MIT Media Lab, where the organizers bring together top researchers from academia, industry, and government to share and discuss research challenges, questions and recent ongoing work in this exciting emerging area. The Laboratory for Information and Decision Systems (LIDS) is a unique place to host and lead such a discussion given its long history and reputation of rigorous research in core information and decision sciences and its recent expansion of research areas to include game theory information and network economics, autonomy, optimization, inference and statistics. LIDS has also recently joined and is playing a leadership role in the newly launched MIT Institute for Data, Systems, and Society (IDSS).The goal of the workshop is to address complex societal challenges by advancing education and research at the intersection of statistics, data science, information and decision systems, and social sciences, and therefore provides the multi-faceted platform needed to address the challenges associated with smart infrastructure services. The workshop is organized around five panel sessions and also includes two keynote speakers. Special care has been taken in providing a balance in terms of gender and diversity of speakers and panelists. All panelists and speakers are at the forefront of this emerging discipline and its related flagship domains. There will be a student poster session showcasing the works of students. The workshop will be recorded. By presenting and outlining a vision for research opportunities in smart urban infrastructures, the workshop will bring together ideas and communities that were not systematically connected before. By explicitly taking diversity of speakers and ideas into account, the workshop will promote and foster further interdisciplinary connections and will also broadly impact the community.