Population Analytics through a WiFi-based Edge Computing Platform

通过基于 WiFi 的边缘计算平台进行人口分析

• 批准号: 1525586
• 负责人: Suman Banerjee
• 金额: $ 20万
• 依托单位: University of Wisconsin-Madison
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2015
• 资助国家: 美国
• 起止时间: 2015-06-15 至 2018-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1525586&HistoricalAwards=false
• 关键词: Population; Analytics; through; WiFi; based

The focus of this project is on creating new techniques for understanding population analytics over a space of interest, e.g., a shopping mall, a busy street, or an entire city. Knowledge of population behavior important for many applications. For instance, knowledge of which are the busy corners of city sidewalk can provide city planners with input on where to invest city resources. Knowledge of where people congregate in a shopping mall allows officials to plan where to provide useful services, e.g., information kiosks, floor plans, and more. The process of gathering population analytics today is tedious -- some stores and shops use manual people counters to track how many persons are entering wireless technologies.The technical contributions of this project are two-fold. First, it is attempting to reduce the complexity of determining location of people by reducing the number of infrastructure points needed. Second, automated approaches to population analytics are fraught with privacy concerns, and this project is examining techniques that mitigate such concerns.Personnel involved in this project will be trained in significant technical skills across a broad set of domains including wireless technologies, privacy techniques, and machine learning.To demonstrate the feasibility of this project, the PI team is deploying a version of the system in an urban downtown area of Madison, WI. The team is collaborating with a number of local partners -- the city of Madison, the University of Wisconsin Bookstore, 5NINES (a local Internet Service Provider), and a few local participants. Together they are entering this technology demonstration as part of the Global City Teams Challenge being hosted by NSF and NIST.

该项目的重点是创建新的技术，以了解感兴趣的空间，例如，购物中心，繁忙的街道，或整个城市的人口分析。有关种群行为的知识对许多应用程序都很重要。例如，了解城市人行道上哪些是繁忙的角落，可以为城市规划者提供关于在哪里投资城市资源的信息。了解人们在购物中心聚集的位置，使官员能够计划在哪里提供有用的服务，例如信息亭、平面图等。今天收集人口分析的过程很繁琐--一些商店和商店使用人工人员柜台来跟踪进入无线技术的人数。这个项目的技术贡献有两方面。首先，它试图通过减少所需的基础设施点数来降低确定人员位置的复杂性。其次，人口分析的自动化方法充满了隐私问题，该项目正在研究缓解此类担忧的技术。参与该项目的人员将接受广泛领域的重要技术技能培训，包括无线技术、隐私技术和机器学习。为了证明该项目的可行性，PI团队正在威斯康星州麦迪逊市中心部署该系统的一个版本。该团队正在与许多当地合作伙伴合作--麦迪逊市、威斯康星大学书店、5Nines(当地一家互联网服务提供商)，以及一些当地参与者。作为NSF和NIST主办的全球城市团队挑战赛的一部分，他们将共同进入这项技术演示。