|
Archives:
Spring 2000 Table
of Contents
Profile
John Leonard
Imagine
driving down an unfamiliar road and trying to find your way while
simultaneously constructing a map of the area. Now, get rid of the
road and submerge yourself in water, and youll get an idea
of what John Leonard is up against in his research. An assistant
professor in MITs Department of Ocean Engineering, Leonard
is currently focusing on how an underwater robot can at once build
a map of an unknown environment and navigate through that terrain.
"Its an area of research still in its infancy, says Leonard.
"And its relevant not just for underwater vehicles, but
for robots operating in nursing homes, on the surface of Mars, and
in underground mines."
Leonard first got interested in robots
while working in college at a General Motors plant that manufactured
Chevrolets. The plant was introducing robotics that required tearing
up the facility to bury wires in the floor that the robots could
follow. Leonard wondered why the robots couldnt instead navigate
by perceiving their surroundings. That led him to his Ph.D. work
at Oxford, where he studied navigation in land robots. From there,
he came to MIT Sea Grants Autonomous Underwater Vehicles (AUV)
Laboratory, where he focused on using sonar underwater to help vehicles
navigate.
Leonards current project, funded
by MIT Sea Grant, also employs sonar as a tool for underwater sensing.
Part of the inspiration for that, he says, comes from the animal
world. "We know that bats, dolphins and whales use sonar to
find food and presumably to navigate." The challenge, says
Leonard, is to develop ways to use sonar that mimic or copy the
way, say, a dolphin uses sonar.
Along with his research, which includes
a few other projects, Leonard keeps busy with his teaching load.
With Tom Consi, a research engineer in the Department of Ocean Engineering,
he co-teaches a course in which students develop and enhance their
own AUV. "This year the students have added sonar to their
AUV, Autolycus, and their challenge is to find an object in the
pool," he says. The exact object, notes Leonard, is up to the
students, who are learning that design is an iterative process,
demanding many choices.
Such decision-making skills, explains
Leonard, are helpful for robots too. Along with research for underwater
vehicles, he has a land robot that he hopes will one day successfully
navigate its way around MITs campus: a feat not easily mastered
by many a human traveler.
Andrea Cohen, MIT Sea Grant
|