A common problem when deploying wide-scale networked solutions is how to power the elements of the net. As reported in a MIT Technology Review article, Perpetua Power Source Technologies based in Oregon developed a solution for effectively powering small devices off excess heat.
According to their product literature, the Perpetua Power Puck generates it’s power using thermoelectric generator (TEG) technologies. It can convert low temperature differences into regulated voltages of 1.8 V, 2.5 V, 3.0 V, and 3.3 V. According to the MIT TR article, each puck is small enough to be placed in tight locations and can generate electricity with a temperature difference as little as 10 degrees C.
I was recently talking to some people at ISOC working with the OECD on research into developing reliable sensor nets. This type of solution plays right into this project as it can vastly simplify many of the power requirements. Assuming a use case in which a wireless sensor can be co-located with a hot water pipe, it is now possible to generate the required power to run the transmitter from excess heat. There is also often more than enough of a temperature differential just a few inches below ground in desert climates, making this type of solution a viable source for long-haul monitoring.
May 4th, 2009 | 
Emergent Innovations: Geo-Tagged Photos Create City Maps
To create this image of San Francisco (he’s currently posted 50 maps), he took the geo-tagged data from photos uploaded to Flickr and Picasa, then banged the locations against OpenStreetMap using Perl and Ghostscript to overlay travel vectors of the photographers. Specifically, he compared photos taken by the same photographer within 10 minutes and bounded by 3 miles to compute and plot their travel vector. The resulting map is color-coded to indicate black=walking (7mph), red=bicycling (19mph), blue=street vehicles (43mph), green=freeway vehicles or rapid transit (>43mph).
I’m not going to argue for/against the privacy issues embedded within geo-tagged photos. That’s a separate issue, but this does clearly illustrate that when people have free and open access to data, they’ll combine them in clever and unique ways to generate something entirely new (and potentially useful).
Provenance: I heard about this via a tweet from @PeteWright, read a blog post (including the comments by Eric explaining his process), and ended up at Eric’s Flickr page.