Tony Stentz
Research Professor
Robotics Institute, CMU
Mauldin Auditorium (NSH 1305)
Refreshments 3:15 pm
Talk 3:30 pm
For over 20 years roboticists have been working on the problem of
autonomous ground vehicle navigation in outdoor, off-road environments.
The potential benefits are immense: autonomous machines can venture where
it is too expensive or dangerous to send humans. So why isn't the problem
solved? Why aren't autonomous vehicles everywhere? The answer is that
the problem is very general and rich, and even limited instances of it are
challenging. But steady progress has been made, and we are on the brink
of deployment for a variety of applications.
In this talk, I will define the dimensions of the problem, explain what is
easy and what is difficult, and identify high-impact subproblems. I will
describe the component technologies required to autonomously drive a
wheeled vehicle across outdoor terrain in a safe manner. My talk will
cover perception, including the sensors typically employed and the
algorithms for interpreting sensor data. I will discuss path planning,
including global planning for route selection and local planning for
obstacle avoidance. I will touch on other topics as well, including
position estimation, human/machine interfaces, and system integration.
My talk will feature results from several of my research programs at CMU.
I will conclude with directions for future research in this area.
Tony Stentz is a Research Professor at Carnegie Mellon's Robotics
Institute and Associate Director of the Robotics Institutes National
Robotics Engineering Consortium. He received his Ph.D. in computer science
from Carnegie Mellon in 1989, his M.S. in computer science from CMU in
1984, and his B.S. in physics from Xavier University of Ohio in 1982.
Tony's research expertise includes unmanned vehicle navigation, dynamic
planning, and multi-vehicle coordination in the context of fieldworthy
systems. His current and past projects include automated machines for
agriculture, mining, security and reconnaissance, and planetary
exploration.
The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.