Bus Data Detects Traffic Snarls

[Randy Burge]

http://www.wired.com/news/technology/0,69991-0.html?tw=wn_tophead_16

Bus Data Detects Traffic Snarls

For commuters, it's a common but aggravating scenario. You speed down the
highway, comforted by radio reports that traffic is flowing smoothly. Then
you get off the exit ramp and find cars jammed to a crawl.

Today, freeway commuters have access to a wealth of up-to-the-minute data
they can use to avoid traffic. But for drivers on major commercial
thoroughfares, gridlock information is limited.

Seattle transportation researchers are trying to make travel on these
arterial routes more predictable by using data collected from city buses.

"We're already tracking all the buses all the time," said Daniel Dailey, an
electrical engineering professor at the University of Washington who leads
the arterial traffic project. "What's left is to differentiate when traffic
is slowing the buses down."

The university's Intelligent Transportation Systems Research Program began
collecting bus data close to seven years ago. The data feeds MyBus, a web
and text-messaging service that notifies commuters about delays.
Seattle-area bus commuters use the service about 5 million times a month,
Dailey said.

Now, the researchers have attached sensors to city buses to detect when they
are moving slowly, as part of a prototype traffic-alert system. During rush
hour, traffic can move as slowly as 10 mph along commuter routes.

Highway traffic speeds are relatively easy to measure because traffic rarely
stops. Many municipalities, including Seattle, use inductive loop detectors
embedded in roadways to record when cars pass by.

But that technology isn't as effective for measuring speeds on routes with
traffic lights, Dailey said, partly because it's unclear whether a car is
slowing because of congestion or a yellow light.

Seattle transit officials said they're interested in including Dailey's bus
data in their traffic map, but haven't yet determined how the information
should be displayed. Another concern is whether data from buses -- which
stop every few blocks to pick up passengers -- is relevant for other
drivers.

"Buses travel in the arterial stream and that's a good thing," said Mark
Hallenbeck, director of the Washington State Transportation Center. "The bad
thing is that buses do things that cars don't, so it's difficult to convert
the bus statistic into a car statistic."

The Seattle project is one of several efforts that use moving-vehicle data
to provide real-time traffic reports. Another program, run by the U.S.
Department of Transportation, is creating vehicle-to-vehicle communication
systems that could monitor traffic.

Private companies are also developing traffic-monitoring systems. Inrix, a
Kirkland, Washington, company founded by former Microsoft executives, uses a
patented statistical-analysis program to predict traffic patterns.
Traffic.com incorporates data from a network of sensors to measure the scale
of traffic jams. And Circumnav Networks in Palo Alto, California, uses GPS
and wireless communications technology to measure traffic.

While it won't eliminate gridlock, Dailey believes using sensor data from
moving vehicles can help drivers avoid congested areas.

That said, Dailey himself has a simpler method for avoiding Seattle
gridlock: He rides his bike.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://redfish.com/pipermail/friam_redfish.com/attachments/20060113/43e61497/attachment.htm