In a Roundabout Way
UMaine engineer takes to the road to improve safety
Ask Per Gårder about traffic safety and
he'll give you a roundabout answer.
The University of Maine professor of civil engineering has been a strong
advocate for modern roundabouts which, he is quick to point out, bear
little resemblance to yesterday's rotaries. His research on accident
rates, personal injuries and traffic flow at 23 intersections in seven
states, including Maine, has convinced him that highway engineers in the
United States can use roundabouts to save lives and reduce congestion.
Roundabouts may look like traditional rotaries to the untrained eye. The
difference is they are designed with specific geometric angles, slowing
entering traffic to 15-25 miles per hour, instead of the 35 miles per
hour or more of rotaries. New roundabouts also offer better visual cues,
such as signs and well-marked islands at exits and entrances that help
drivers maintain a comfortable, orderly flow.
An analogy to flowing water is fitting for the roundabouts advocated by
Gårder, who once considered studying hydraulics in college. Rather than
working with plumbing systems that channel water molecules through
networks of pipes and valves, he focuses on systems of roads and
intersections that safely and efficiently get vehicles from one place to
A bus route
For his master's thesis at the Lund Institute of Technology in Sweden,
Gårder developed new ways to design networks of public bus routes. A
drawback to many public transportation systems, he says, is their
tendency to funnel all riders downtown. If a person needs to go from one
suburb to another, the bus ride can take much longer than necessary.
"When everybody works in the central business district, that's good for
commuting. But nowadays, when people live in one outskirt and work in
another, it is difficult for public transportation to meet people's
needs. That's one of the reasons that people don't use public
transportation in the United States," he says.
Even more important for Gårder is reducing the number of accidents that
take about 42,000 lives and cause about 1.6 million disabling injuries
annually in the U.S. Together with the 25 million other traffic
accidents that don't cause death or injuries, the total bill comes to
more than $100 billion.
At about 16 deaths per 100,000 people, the U.S. has one of the worst
records for traffic fatalities in the developed world, he says. With an
average of 200 traffic-related deaths per year, Maine is near the
"Northern states like Maine do slightly better than Southern states, but
if we compare to the United Kingdom, Norway or Sweden, they are all
slightly below six fatalities per 100,000 people," Gårder says.
Shortly after he came to UMaine in 1992, Gårder made a significant
contribution toward improved safety on U.S. highways. Working with
former civil engineering professor John Alexander, he demonstrated that
highway rumble strips, lines etched into the pavement just outside the
travel lanes, could save lives. The noise of tires hitting rough
pavement can avert one of the most significant problems on interstate
highways — drivers who doze off at the wheel and drift off the road.
Before rumble strips were installed, nearly half of all fatal crashes on
Maine's interstates were caused by people falling asleep.
Gårder and Alexander did not invent the technology. Some states had been
experimenting with rumble strips since the 1950s. However, their project
was the first academic study of this simple life-saving technology in
the country and has been credited with spurring more widespread use.
Their results showed what engineers consider an astounding cost/benefit
ratio of more than 1 to 100, meaning that for every dollar invested,
more than $100 would be saved in property damages, healthcare expenses
and other costs.
"Our first paper was refused by the reviewers because they thought that
no measure would ever have a benefit cost ratio of 100 or above," says
Gårder. However, a new federal report on rumble strips cites the same
results, confirming that Gårder and Alexander were right.
The Maine Department of Transportation published two reports on rumble
strips and began installing them along I-95 in 1994.
"For a couple of years after that, I was getting e-mails and phone calls
once or twice a week from people who had been awakened by the rumble
strips," says Gårder. "They were thanking me for saving their lives."
While the benefits are clear, every technology has some negative
consequences. Rumble strips are no exception. In 2000, a motorcyclist
reported that by moving across a rumble strip to stop along the highway,
she lost control and crashed.
Gårder takes such possibilities seriously. He also points to the
overwhelming evidence of lives saved and property damage reduced. In
fact, since rumble strips were installed, the number of accidents that
have been linked to sleepy drivers on I-95 has gone down about 37
While rumble strips have been well accepted, roundabouts are another
story. Proponents in engineering firms, universities and state
transportation departments are touting the benefits, but the public is
not so sure. Newspaper accounts of public meetings in Florida,
California and Illinois reflect citizen discomfort with roundabouts that
some view as "an accident waiting to happen."
Gårder suggests that the American aversion to roundabouts is the result
of experience with poorly designed rotaries. "In the United States, we
had a lot of circular intersections built in the 1920s and 1930s," he
says. "Some of them worked reasonably well, and some of them were too
big and designed for speeds that are too high."
States began constructing modern roundabouts in the early 1990s. With
some design assistance from Gårder, the Maine Department of
Transportation opened one at the intersection of routes 202 and 237 in
Gorham in 1997. The goal of the project was to reduce the accident rate
and waiting times for the 13,000 vehicles that travel through that
junction on an average day.
Gårder's evaluation after only one year of operation showed that the
roundabout was working as planned. He found that the accident rate had
dropped from 1.36 per million vehicles to 0.8. Even better, personal
injuries were down. Three accidents had occurred at the roundabout, but
none caused injuries. They were all sideswipe accidents. Broadside or
head-on collisions that are more common at traffic lights and stop signs
tend to be far more dangerous.
On the road
The Maine experience is not unique. In Florida, Vermont, Colorado and
other states, roundabouts have performed with similar results. In 2001,
Gårder and colleagues at Ryerson Polytechnic University in Toronto and
the Insurance Institute for Highway Safety published a landmark study
based on accident reports for 23 roundabouts in seven states. They
compared accidents before and after the roundabouts were constructed.
Just as important, they used statistical techniques to account for
variability in traffic flow and what traffic engineers call a randomness
factor. Stated simply, a small number of accidents can happen randomly
anywhere because behavior is unpredictable. Drivers fall asleep, have
heart attacks and drive drunk. Animals wander onto roads. No
intersection or highway is foolproof.
"If you don't take randomness into account, you end up with even better
numbers for the roundabout," says Gårder. "But taking it into account
makes it a more complicated but more honest analysis."
Their results showed that the 23 roundabouts achieved a 40 percent
reduction in all accidents, an 80 percent reduction in injury-causing
accidents and a 90 percent reduction in fatalities. The strength of
their case earned the paper a prestigious award from the Transportation
Research Board, as one of the four best research papers out of 2,000
submitted in 2001.
"My strong belief is that fewer people are seriously injured at a
roundabout than at any traffic signal you will ever have," says Gårder.
"The reason is that the roundabout reduces the speed of almost everyone
to about 15–20 miles per hour. At a traffic signal, a lot of people will
not reduce their speed, especially when the light is yellow. The typical
person believes that traffic signals are perfectly safe because when
it's red, people stop and when it's green, they go."
Despite their demonstrated safety, Gårder wouldn't build roundabouts at
all intersections now controlled by stop signs or signals. However,
where roundabouts would be particularly useful is on rural highways at
the outskirts of small towns and villages.
"We have 45–50 mile per hour highways going through towns without
changing geometric standards," he explains. "We have a lot of accidents
in these small towns, partly because people do not slow down. They
continue to drive at 45 miles per hour, even if the signs say 25 miles
Roundabouts would be an effective way to slow traffic down where rural
highways enter developed areas, Gårder says. They also could offer
drivers equal access to the business district or a downtown bypass.
Giving through traffic the chance to stay out of downtowns would help
reduce the congestion for which some small communities have become
"We should not accept people being killed for reasons that are
avoidable. Traffic fatalities are not fully avoidable, but we could
reduce the number a lot," says Gårder.
by Nick Houtman
for more stories from this issue of UMaine Today Magazine.