Slashdot Mirror
Sensors Gone Wild
tulanian writes "forbes.com has an interesting article on networked, intelligent sensors. It mentions an experiment done by DARPA where several dozen magnetic sensors were scattered along a road and passing vehicles could be identified by their magnetic signatures."
These tests were down with MICA sensor motes which can be purchased from Crossbow Technology.
These motes run TinyOS, which was developed at UC Berkeley.
More information about TinyOS:
Yes, my job does involve programming for these motes. I have four of them on my desk acting as an ad-hoc wireless sensor network now.
You think they can't already do this, by tracking your cellphone?
fyi, that's not a mirror, it's a synopsis of the research that was published in PNAS.
I think just the fluctuations in voltage due to a rather insane stereo system could easily offset that. Not to mention the blanket of influence a 500w subwoofer's magnet has on electronics around it. I can warp the display on my monitor from 10 ft away with some of the components I have.
-- AcquaCow
up 12 days, 22:30, 2 users, load averages: 993.20, 994.21, 994.56
*makes note to limit user processes...
I talked extensively with a fellow who was maintaining a red light camera. He was very open, and largely responsible for implementing red light cameras in British Columbia.
There are two sensors about four meters apart, one just before and one just after the stop line. These sensors can identify speed and "weight" (magnetic mass) of the vehicle. The magnetic signature is unique to each vehicle model.
In BC, the sensors activate the camera only when the car actually enters the intersection on a red. If you've entered on a yellow, no camera.
Because of the speed-measuring capability, some Vancouver-district municipalities have also really put the screws to the boy-racer fuckups who are street-racing -- speeds like 140kmh through a red light. Ticket for running the red, ticket for reckless driving, ticket for speeding, etc.
Ticket revenue goes directly back to the red light camera operations. The municipalities and government see no money from this system, thereby eliminating any incentive to generate additional revenue by diddling the yellow-light timing and or going all heavy on the marginal cases. The camera supplier doesn't get money, either (and, thus, were of absolutely no use whatsoever when it came time to provide technical support.)
Camera-monitored red lights in BC are showing significant reductions in accidents. Indeed, the savings generated by not paying out claims is proving to be a remarkable value, and insurance companies across North America are starting to realize that spending money on accident reduction will increase their bottom-line profits.
All in all, it's a hella fine system. Our government-tamed insurance monopoly provides us with basic insurance and uninsured driver insurance, plus is mandated to work with municipalities and the RCMP to increase road safety and decrease accident rates. Our public insurer operates in a not-profitable manner: they're not expected to (and are, in fact, expected to not) produce profits.
All our other insurance (liability, collision, whatever else you dream of having) is provided by third-party insurers, so we really get the best of both worlds: a public insurer who's looking out for our basic needs and safety; and private insurers who compete to provide all else.
Sorry this turned into a bit of an insurance/insurer rant, but the public insurer aspect is crucial to the success of our red light camera system, because it eliminates the urge to use the cameras as a profit center instead of a safety tool.
FYI - Deployable remote sensor systems are not a new concept. The U.S. Army has been using them since about Viet Nam in one form or another. I had the pleasure of visiting Iraq in order to deploy sensors during Desert Storm. The systems in current use can be seen here.
The system in this article appears to potentially be the next generation of this sensor system.
(Any current or former GSR feel free to drop me a line)
There is nothing in the current motes that can not be miniaturized. In three years this demo will be done with a 6" aircraft, and millimeter-scale sensor nodes.
This is all leading to "Smart Dust".