Distributed sensing is the least famous member of the architectural family composed of distributed computing (seti@home), distributed storage (OceanStorage), and distributed bandwidth (BitTorrent).

As you might guess from the name, the concept involves networking large numbers of fixed and/or mobile sensing points, collecting and analyzing data generated by those points, and then (usually) redistributing the results with others.

A very cool if highly specialized example is networking hundreds of hobbyist seismographs, thus making it possible to watch an earthquake develop over large areas in real time (at a resolution of one pixel per instrument).  Another might be networking home weather stations in a neighborhood to get a picture of the behavior of the atmosphere over small scales -- so-called microclimates.  Or using the actual routes driven by real people to improve the driving instructions given by online maps. Right now these routes are calculated by algorithms that assume all miles are equal. As we all know, this is far from the case -- sometimes a longer route is far more efficient. If cars could send GPS data collected during their travels direct to the mapping sites, those sites could evaluate choices made by real drivers by measuring how quickly their cars actually traversed given segments. Such information would make driving directions much more useful.

A major application of distributed sensing is its potential to put urban planning, traffic engineering, and crowd control, both in and outside buildings, on a far more empirical basis. Imagine for instance that you wanted to know the full profile of all the acoustic experiences people are exposed to during their entire day. Recruited participants, paid or unpaid, could download a bit of software to their cell phones, which would then send snapshots of the ambient sound levels experienced by the volunteers once a second, every second, during the entire day.

As both amateur and professional historians know all too well, typically the largest fraction of the life of a city goes totally undocumented.  Buildings rise and fall, and neighborhoods change radically, with no sign of these changes appearing on a retrievable or accessible photograph. The right kind of networking among GPS-enabled camera phones, would allow people interested in this project to document a city automatically, just by using their cameras in public.  Anytime anyone took a picture in front of 50 Main St., that picture would go into some archive somewhere (bound by whatever privacy restrictions were imposed by the photographer).

You do not have flash or javascript support.

Average (0 votes)

Reply