SenseWeb: Browsing the Physical World in Real Time

Geo-centric web interface such as MSN Virtual Earth (http://local.live.com) and Google Maps (http://maps.google.com) are useful to visualize spatially and geographically related data such as addresses, neighborhoods, weather, traffic, and so on. Desires to augment additional useful information to these interfaces have led people to create custom applications that overlay their own data on top of browsable maps. Examples of such applications overlay housing information (http://www.housingmaps.com) and crime-rate data (http://www.chicagocrime.org/map/) on Google maps, locations of vehicles (http://jprestonsystems.2mydns.com/ vemap.aspx?name=demoacct2) and podcasters (http://www.podlook.com/map.aspx) on MSN Virtual Earth, weather data on custom maps (http://www.wunderground.com), etc. Such applications have been possible after Google Maps and MSN Virtual Earth have published useful APIs to overlay location data on their maps. We envision publishing and querying real-time data (e.g., from sensors) over such geo-centric web interfaces. Existing solutions, although useful to write simple applications as above, have several drawbacks in achieving this vision. First, publishing even a single stream of data as a useful service is a nontrivial task. Many useful data is not being published yet because the data owners do not have enough programming expertise, or publishing it requires too much effort. Second, all the existing applications are mutually incompatible. One can’t bring up a single map that shows both the housing information and crimerates in an area. Third, existing solutions do not provide useful primitives such as querying live sensors based on keywords or location and aggregating the results in useful ways. The SenseWeb project at Microsoft Research Figure 1: SenseWeb Architecture.