Slashdot Mirror
Check Traffic Congestion Online
braddk writes "Looks like traffic helicopters will slowly become unnecessary in Denver, as an ongoing construction project implements online traffic data. The traffic is monitored via "vehicle counters" placed at the onramps and in between interchanges. Although only a 10 mile section is currently monitored, plans are to add more sensors as they complete sections of the larger project. They also have a lighter version for mobile phone users.
Click here to see the Flash version
and to check out the current traffic in Denver. Now I can check whether I really want to head to work in the morning." Kinda like that project in Finland.
Atlanta's traffic Sure it is not in pretty flash, but is is much more extensive.
The reporters can sit in their studios and follow the congestions real-time and will report it with regular intervals on the news. Works fine.
Aren't similar systems in use in lotsof other places?
I'm in a Unix state of mind.
We've had this in the UK for a decade. It's called traffic master. Speed senors are mounted on freeway bridges all around the country and provide constsnt traffic flow information, which can be relayed to a map display on the dash.
Speed sensors give rather better information for this purpose than car-counters. See http://www.trafficmaster.co.uk/
Seattle's had maps like this for years, and they work quite well:
http://www.wsdot.wa.gov/PugetSoundTraffic/
They don't read speed directly, but rather chart traffic density by measuring the changes in inductance of 6 foot metal loops embedded in the highway every 1/2 mile or so. They count cars passing over the loops and measure the time a car spends over it.
From the WSDOT FAQ:
"How are "Stop and Go", "Heavy", "Moderate", and "Wide Open" defined?
Their definitions are based on a measurement called "loop occupancy". This is the percentage of time that a 6 foot square loop sensor is activated, or occupied, by vehicles traveling over it. Loop occupancy is measured by sampling the loop detector at a rate of 60 times per second. Each sample results in a "loop occupied" or "loop not occupied" response. A counter is incremented once for each "loop occupied" response. After 20 seconds, the total number of "loop occupied" responses is divided by 1200 (the total number of samples in a 20-second period) and then multiplied by 100 to get a percentage. The result is known as occupancy. On the map, "Stop and Go" (or BLACK) is defined as occupancy greater than 35%, "Heavy" (or RED) is 22-35% occupancy, "Moderate" (or YELLOW) is 15-22% occupancy, "Wide Open" (or GREEN) is below 15% occupancy. "
Nothing new in France either.
Roads have had sensors for at least a decade.
Used for different purposes
- traffic monitoring (accidents, etc.)
- driver information by huge screens on the road, telling how long to this and this direction ; and I find it really nerve calming to know how long it will take and be able to organize (once it only said how long - in distance - the congestion is, which I don't care about)
- website for 4 years.
Here we have two type of sensors
- simple loops, which only give information about the "coverage rate" (that is, proportion of time there is a vehicle on the loop. Funnily, this figure is heavily correlated with the state of traffic and the speed of the vehicles. 0.1 is heavy traffic and 0.2 is congestion. I do not recall exacly the figures but you get the highest throughput for a magic "coverage rate" which corresponds to around 57 kmph (~37 mph).
- double loops are simple loops 1 meter away ; correlating data from the two gives you the time decay between them and so the speed of vehicles, in a more reliable fashion than just simple loops ; in particular with these you can ajust the nominal traffic model with observed speeds so your model integrates real road conditions (snow, rain, saturday night...) and single loops can then give you very accurate information.
On heavily trafficked roads (eg Boulevard Peripherique in Paris) you have a single loop every 400m and a double every 2km AFAIR.