|
Virginia's
Smart Travel Laboratory keeps traffic flowing in metropolitan areas
By
Charlotte Crystal
 |
|
Rebecca
Arrington
|
| Brian
Smith, assistant professor of civil engineering, is co-director
of the Smart Travel Lab, based at U.Va., which interprets signals
from highway video cameras to monitor traffic flow. The lab
works with VDOT's Virginia Beach Smart Traffic Center to help
its staff quickly spread the word about traffic slowdowns or
accidents by notifying drivers via message signs on the road
and highway advisory radio announcements. |
Virginia's
Smart Travel Laboratory
keeps traffic flowing in metropolitan areas By Charlotte Crystal
f Brian Smith has his way, traffic forecasts will someday be as
common as weather forecasts.
Research
under way at Virginia's Smart Travel Laboratory at U.Va. will help
drivers in metropolitan areas better predict traffic patterns and
adjust their travel plans accordingly, whether to ease a daily commute
or avoid holiday traffic, such as that expected during the coming
Labor Day weekend, said Smith, co-director of the lab and a research
assistant professor of civil engineering.
The
lab also expects to help the Virginia Department of Transportation
analyze massive amounts of traffic data piped in from the congested
Washington and Hampton Roads metro areas. More effective interpretation
of the data will help VDOT respond faster to changing traffic conditions
and improve the flow of traffic.
Established
in 1998, the lab conducts cutting-edge research that combines historical
data with traffic-simulation models to create forecasts of traffic
volume and travel times. U.Va. researchers also have helped VDOT
design and upgrade its sensing systems and identify and fix faulty
sensors.
The
lab is directed by Smith and Cathy McGhee, a civil engineer with
the Virginia Transportation Research Council, the research arm of
VDOT. Other U.Va. professors of civil and systems engineering who
specialize in transportation issues and a contingent of undergraduate
and graduate students round out the center's staffing. Funding is
provided primarily by U.Va., VDOT, the Virginia Transportation Research
Council and the U.S. Department of Transportation.
The
lab is currently working with VDOT's Smart Traffic Center in Virginia
Beach, which receives traffic data from 600 vehicle sensors and
features a wall of 38 video monitors linked to cameras set up along
19 miles of the area's most congested roads, Interstates 64 and
264. Smart Traffic Center controllers monitor the camera images
24 hours a day and can respond to traffic slowdowns or accidents
quickly by contacting a Freeway Incident Response Team and notifying
the traveling public of the adverse conditions via variable message
signs and highway advisory radio announcements.
"When
all is said and done, there will be over 280 cameras, 240 variable
message signs, and nearly 2,700 vehicle detection devices along
113 miles of Hampton Roads interstates," said Erika Ricks,
Smart Traffic Center spokeswoman. "At that point, it will be
nearly impossible to monitor traffic flow without the help of the
information reaped from the efforts and technology of the Smart
Travel Lab."
The
U.Va. center receives all of the Hampton Roads vehicle sensor data
and can pick up the signals from any of the highway video cameras,
displaying them on video monitors in Charlottesville. By the end
of September, the center also expects to be connected to the Northern
Virginia Signal System, which will bring in real-time data from
800 vehicle detectors embedded in pavement in Northern Virginia.
A video link with Northern Virginia also is planned.
While
the purpose and potential impact of research conducted at the lab
is straightforward, the underlying mathematics is sophisticated,
Smith said. The math involves optimization models, search techniques,
data mining, data analysis, and simulations.
Unlike
other traffic research projects that stress the use of theoretical
physics in predicting traffic flows, the Charlottesville center
focuses on the analysis of actual, historical data.
"The
mathematical models used by physicists to describe the flow of fluids
are governed by immutable laws of physics," Smith said. "But
individual drivers are not governed by physical laws and may make
decisions that could not be predicted by physics. That's why we
believe it's more useful to base our mathematical models on historical
data. We're looking at what people have actually chosen to do in
particular situations."
Other
research projects include:
-
Automated Condition Monitoring -- Rob Turochy, a Ph.D. candidate
in civil engineering, is looking for ways to quickly determine
when the data stream from vehicle detectors shows something abnormal
and important in the flow of traffic.
-
Short-Term Traffic Conditions -- Kevin Smith, a master's degree
candidate in civil engineering, is working on forecasting algorithms
that will consider travel time -- to answer such questions as
whether a driver would likely be caught in rush-hour traffic two
hours from now if he leaves in half an hour.
Studies
in other parts of the country have shown that traffic management
systems reduce air pollution, fuel consumption and travel times
by up to 25 percent. These systems also cost taxpayers far less
than building new roads.
While
beach traffic this Labor Day weekend may be as slow and heavy as
ever, someday in the not-so-distant future, Virginia drivers may
be able to log onto their home computers, get a traffic forecast
-- good for whatever time they want to leave --‹ and hop into their
cars knowing they'll be taking the road less traveled.
|
