This collaborative project, deploying a large-scale open sensor network testbed for urban monitoring (called CitySense), aims at building a large-scale, urban sensor testbed of nodes powered by city electricity and then operating it as an open facility for academic and industrial groups to experiment with novel distributed sensing architectures. The modular network nodes will be attached to streetlights for power. These nodes contain sufficient processing power and memory to handle a wide range of experimental software, plus radios to interconnect all the nodes by multi-hop networking across Cambridge to its wireline access points. Each node comes with software in flash to provide a robust sensor network architecture including reliable ad hoc networking. CitySense will be tested with a specific application: monitoring air pollution transport in a dense urban environment. Aiming at pushing the envelope for sensor network testbeds, CitySense contains a basic set of meteorological and air quality sensors supporting a broad range of research projects in large-scale sensor networks. Enabling many research experiments with a real large-scale in-situ WSN, the overall research focuses on:
-High-level programming models for WSN, -Resource sharing among multiple users of the testbed, -Data collection from geographically diverse sensor nets, and -Urban-scale air pollution monitoring.
Broader Impact: By providing a permanent, extensible, public testbed, CitySense should have significant impact on the development of large-scale wireless sensor network systems. The open city-wide testbed allows external groups to remotely download their software images into all the nodes in the network, or into a sensor swath when multiple groups share the testbed. Monitoring air pollution transport in a dense urban environment, the application constitutes one of the first high-resolution studies of pollution and its effect on the urban population. Moreover, CitySense will be used as a tool to teach distributed systems concepts. Outreach involves curriculum development for K-12 in sensor networks.
The CitySense project goals were to build a wireless sensor network, deploy it in an urban environment and open it up to the research community. This collaborative industry/university project led by a team from BBN Technologies and Harvard University, began in 2006 and by the end of the project in September of 2010, over 30 nodes have been deployed on street-lights and rooftops around the Cambridge, MA area. See the project web-site at http://citysense.net. The CitySense network consists of four clusters of nodes, where a single node in each cluster acts as the gateway to the Internet. Each node acts as a router forwarding data to and from the other nodes using their wireless radios. Nodes can contain up to three sensors. The sensors deployed on nodes range from meteorological including: rainfall, wind speed and direction, air pressure, temperature and humidity, to carbon dioxide and particulate counters. These data are logged to our sensor database (http://citysense.net/DataQuery/). We have collected over 70 million measurements since the first outdoor nodes were deployed in 2007. Some of the key outcomes of this project were: Design of a low-cost sensor node that withstands harsh outdoor environment Performance analysis of wireless networking in urban deployment Centralized web-site for managing configuration of sensor network Programmable outdoor wireless mesh testbed Watchdog process to reboot nodes into a stable OS image when network/node failure is detected. By the end of the project in 2010, CitySense supported a number of users: Public health and civil engineering researchers from Tufts University used the nodes to track particulate matter spread from roadways; K-12 educators at Cambridge Friends School used data from our weather sensors in the design a middle school science curriculum; RF antenna performance studies for mobile, multi-path rich environments Harvard graduate student studies on WiFi usage patterns in a city: "Mapping the Urban Wireless Landscape with Argos," Ian Rose and Matt Welsh. In Proceedings of the 8th ACM Conference on Embedded Networked Sensor Systems (SenSys'10), November 2010 •Mapping the Urban Wireless Landscape with Argos, Ian Rose and Matt Welsh. In Proceedings of the 8th ACM Conference on Embedded Networked Sensor Systems (SenSys'10), November 2010.
