Objects May Be Closer Than They Appear
by David Pescovitz
 Printer-friendly
version
UC
Berkeley professor Theodore E. Cohn's research may lead
to better signals at railroad crossings.
|
Each year,
approximately 400 people die trying to beat an oncoming train
at railroad crossings.
More than 1,000 others are injured. Why is it that so many people
misjudge the speed of an oncoming train? That's the question
Theodore E. Cohn, a Berkeley professor of vision science and bioengineering,
hopes to answer. Understanding why people think they can win the
race at railways, Cohn says, may lead to better signals that prevent
drivers from thinking they're faster than a locomotive.
"In 1985, the theory was presented that we underestimate the speed
of large objects," says Cohn, a researcher with PATH (Partners
for Advanced Transit and Highways). "We're finally
testing that idea for the first time."
To conduct their preliminary experiments, Cohn and his students
created a computer-based laboratory test that didn't require
any moving objects. In the study, each subject sees a square,
grey box on a computer screen. The subject is instructed to hit
a button
the moment he or she notices the box begins to expand.
"The expansion of an object in your field of view is a cue your
brain uses to determine how rapidly the object is moving toward
you," Cohn says.
The geometry that links an object's expansion to our estimation
of its speed was first described by astronomer and writer Sir Fred
Hoyle in his 1957 science fiction novel The Black Cloud. As it
turns out, Cohn's experiments revealed that the bigger
an object is when you first see it, the longer it takes you to
notice
it change.
"That makes us think that an object may be approaching much more
slowly than it really is," Cohn says.
Along with devising new signaling systems, the researchers
are conducting experiments to determine where our attention
is focused
when looking at an approaching object.
|
Of course, the rate of expansion is not the only factor humans
use to determine the speed of an oncoming object. Stereopsis,
our binocular perception of depth, also helps us determine how
close
something is to us. The problem, Cohn says, is that stereopsis
isn't very effective at distances of more than 10 meters.
"That's a problem when you're following a vehicle in
traffic," Cohn says. "Interestingly, buses are rear-ended
more often than cars and they're bigger. So we'd
like to see if that's the case with trucks as well."
After the laboratory experiments are complete, the researchers
will begin real-world tests to determine whether it is indeed
a vehicle's large size that causes drivers to misjudge its
speed. For example, Cohn and his students will compare their subjects' ability
to estimate the speed of an approaching train compared to other
smaller vehicles that travel along the railways.
Eventually, Cohn hopes his research could inform the design of
new signal lights for trains. Currently, trains feature a triangle
of headlights on their front ends. The approach is designed to
give the onlooker a sense of the speed of the train based on how
fast the triangle of light seems to be expanding. The irony, Cohn
explains, is that the lights are too bright to look at them for
the length of time necessary for the brain to process the information.
One system the researchers are considering entails nested rings
of lights that are visible but not blinding. The system is similar
to Cohn's Bus Bar, an advanced warning signal optimized to
take advantage of the fastest pathways in a human's visual
nervous system. Beginning at the center, each ring in the train
signal light would flash on sequentially at a speed based on
the velocity of the train.
"The lights would appear to be getting bigger faster than they
should, given what you estimate the speed of the train to be,"
Cohn says. "That way, maybe we can compensate for our misestimation
of the speed of large objects."
Along with devising new signaling systems, the researchers
are conducting experiments to determine where our attention
is focused
when looking at an approaching object.
"This may give us a clue where we might place signals or markings
on vehicles to prevent collisions," Cohn says.
If
You Can See This, You're Too Close by
David Pescovitz (Lab Notes, May/June 2002)
Visual Detection Laboratory
Partners for Advanced
Transit and Highways (PATH)
Lab Notes is
published online by the Public Affairs Office of the UC Berkeley
College of Engineering. The Lab Notes mission is to illuminate groundbreaking
research underway today at the College of Engineering that will
dramatically change our lives tomorrow.
Media contact: Teresa
Moore, Lab Notes editor, Director of Public Affairs
Writer, Researcher: David
Pescovitz
Web Manager: Michele
Foley
Subscribe or send comments to the Engineering Public Affairs
Office: lab-notes@coe.berkeley.edu.
© 2003 UC Regents.
Updated 9/29/03.
|
