On Friday I attended the Oregon Transportation Summit.
Being a geek, I sat in on the gadget session. One of the more interesting areas was how to detect vehicles at intersections. The traditional technology is the “loop”, a wire embedded in the pavement that generates a magnetic field that can detect the metal in a vehicle (including bicycles). Alternatives on display, all generally aimed at removing the requirement for wires in the pavement included:
- Detection by video
- A small, wireless, battery-operated cube (about 1.5 inches) that used the same magnetic approach as a loop, but without the wiring. It also had the interesting ability to generate a “magnetic signature” for each vehicle so you could measure vehicle progress across multiple intersections to measure arterial traffic speeds (a measure of congestion).
- Detection by radar
The common factor for all of these was that they were “IP-based” (i.e., Internet Protocol – they could all talk back to your central control center of the Internet).
There was also an internet-controllable street light controller, so you could dim or turn off streetlights from the control center (the standard technology for street lighting just uses a dumb photocell).
7 responses to “Internet Coming to an Intersection Near You”
Did you know that traffic flow can be monitored by tracking your Bluetooth device as it passes different locations? Not the same problem, of course, but I hadn’t given any thought to the device’s unique MAC address or realized the signal strength was this great.
This is pretty cool. I most like the cube idea that a “magnetic signature” can be used somewhat anonymously to track traffic. A lot of people don’t like the “big brother” aspects of car based GPS and the like but this doesn’t seem to get to the level of identification.
Also, those loops are supposed to work on bikes? Never seem to for me. Interestingly however, in Vancouver (my home) I’ve noticed newly paved roads have nice new bike lanes that have their own wire sensor in the bike lane at lights. They DO seem to work for me. Not helpful when trying to make a left from the traffic light though. :-)
Ken: Also, those loops are supposed to work on bikes? Never seem to for me. Interestingly however, in Vancouver (my home) I’ve noticed newly paved roads have nice new bike lanes that have their own wire sensor in the bike lane at lights. They DO seem to work for me. Not helpful when trying to make a left from the traffic light though. :-)
There are a few bike loops scattered around Portland as well. The one I’ve used myself is at the point where NE Sandy, Glisan and 22nd converge.
Yes, there was a vendor there doing the Bluetooth MAC thing as a way to track congestion.
The general-purpose loops for cars should in general also be tuned to detect bicycles, but sometimes you have to be in just the right spot. PBOT sometimes paints a little symbol to show bikes where to be to trip the signal.
Indianapolis airport did a study that tracked people using bluetooth through security check to get an idea of how fast people were moving through. It was merely a study and as far as I know, not a permanent deal, but it is interesting to note.
The common factor for all of these was that they were “IP-based” (i.e., Internet Protocol – they could all talk back to your central control center of the Internet).
The sensor technology (video, induction, detection of RFID or wireless devices) and the network technology (IP, whatever) are, of course, orthogonal issues.
I’d worry if traffic control (and monitoring) devices aren’t being placed on the public Internet, though. (IP-based private networks). While the only “hacker-proof” network technology is a physically isolated (and physically secured) one–and the latter condition is not possible with a network that monitors sensor and control devices located in public places, there’s all sorts of mischief that one must worry about if these things are out on the Intertubes.
The “Internet” in the title was not intended to suggest whether the devices would be on the public internet, private IP-based networks or Virtual Private Networks (how I might engineer it).