Talking of cellphone triangulation, one could get a much higher sample size by just averaging the movement of the vast sea of mobile phones driving down the road. Almost 100% road coverage without needing shed loads of people to have a GPS enabled phone and the appropriate software.
At Cambridge University us Undergraduate engineers do a similar robotics project in 4 weeks on top of normal lectures. It's manic, but do-able. (Teams of 6 students)
See: http://www.eng.cam.ac.uk/DesignOffice/idp/index.html
Of course, the project is run twice a term so all of the computing equipment and hardware you might need to build the robot is readily to hand - that makes a world of difference to speeding up development.
Slashdot Mirror
Talking of cellphone triangulation, one could get a much higher sample size by just averaging the movement of the vast sea of mobile phones driving down the road. Almost 100% road coverage without needing shed loads of people to have a GPS enabled phone and the appropriate software.
In fact, Google says that Missouri is one step ahead of me:
http://www.engadget.com/2005/10/11/missouri-to-use-cellphone-signals-to-monitor-traffic/
At Cambridge University us Undergraduate engineers do a similar robotics project in 4 weeks on top of normal lectures. It's manic, but do-able. (Teams of 6 students) See: http://www.eng.cam.ac.uk/DesignOffice/idp/index.html Of course, the project is run twice a term so all of the computing equipment and hardware you might need to build the robot is readily to hand - that makes a world of difference to speeding up development.