Post by The Sonar Chicken on Sept 3, 2008 3:32:40 GMT
www.livescience.com/technology/080902-helicopter-robot.html
Robot Helicopter Teaches Itself How to Fly
By LiveScience Staff
posted: 02 September 2008 05:56 pm ET
A new artificial intelligence system allows a robotic helicopter to teach itself how to fly and even do challenging stunts, just by watching other helicopters perform the same maneuvers.
The result is an autonomous helicopter than can perform a complete airshow of complex tricks on its own, its inventors say.
The stunts are "by far the most difficult aerobatic maneuvers flown by any computer-controlled helicopter," said Andrew Ng, a Stanford University professor directing the research of graduate students Pieter Abbeel, Adam Coates, Timothy Hunter and Morgan Quigley.
A new video demonstrates the robot's capabilities.
Rather than using software to control flight, the robot learns by observing an expert in what the Stanford team calls "apprenticeship learning." Radio-control pilot Garett Oku operates the 4-foot model helicopter that serves as the expert.
"Garett can pick up any helicopter, even ones he's never seen, and go fly amazing aerobatics. So the question for us is always, why can't computers do things like this?" Coates said.
Well, they can.
The artificial-intelligence helicopter, an off-the-shelf model other than its new brains, can do traveling flips, rolls, loops, stall-turns with pirouettes and more. It can even do the "tic toc," in which the helicopter, while pointed straight up, hovers with a side-to-side motion as if it were the pendulum of an upside down clock.
"I think the range of maneuvers they can do is by far the largest" in the autonomous helicopter field, said Eric Feron, a Georgia Tech aeronautics and astronautics professor who worked on autonomous helicopters while at MIT. "But what's more impressive is the technology that underlies this work. In a way, the machine teaches itself how to do this by watching an expert pilot fly. This is amazing."
Helicopters are not easy to control. Constant input is required to keep one stable.
"The helicopter doesn't want to fly," said Oku. "It always wants to just tip over and crash."
The robotic student is loaded with aftermarket instrumentation, from accelerometers and gyroscopes to magnetometers, which use the Earth's magnetic field to figure out which way the helicopter is pointed.
In the future, such a craft might prove helpful to search for land mines in a war region or to map out wildfire hotspots.
Robot Helicopter Teaches Itself How to Fly
By LiveScience Staff
posted: 02 September 2008 05:56 pm ET
A new artificial intelligence system allows a robotic helicopter to teach itself how to fly and even do challenging stunts, just by watching other helicopters perform the same maneuvers.
The result is an autonomous helicopter than can perform a complete airshow of complex tricks on its own, its inventors say.
The stunts are "by far the most difficult aerobatic maneuvers flown by any computer-controlled helicopter," said Andrew Ng, a Stanford University professor directing the research of graduate students Pieter Abbeel, Adam Coates, Timothy Hunter and Morgan Quigley.
A new video demonstrates the robot's capabilities.
Rather than using software to control flight, the robot learns by observing an expert in what the Stanford team calls "apprenticeship learning." Radio-control pilot Garett Oku operates the 4-foot model helicopter that serves as the expert.
"Garett can pick up any helicopter, even ones he's never seen, and go fly amazing aerobatics. So the question for us is always, why can't computers do things like this?" Coates said.
Well, they can.
The artificial-intelligence helicopter, an off-the-shelf model other than its new brains, can do traveling flips, rolls, loops, stall-turns with pirouettes and more. It can even do the "tic toc," in which the helicopter, while pointed straight up, hovers with a side-to-side motion as if it were the pendulum of an upside down clock.
"I think the range of maneuvers they can do is by far the largest" in the autonomous helicopter field, said Eric Feron, a Georgia Tech aeronautics and astronautics professor who worked on autonomous helicopters while at MIT. "But what's more impressive is the technology that underlies this work. In a way, the machine teaches itself how to do this by watching an expert pilot fly. This is amazing."
Helicopters are not easy to control. Constant input is required to keep one stable.
"The helicopter doesn't want to fly," said Oku. "It always wants to just tip over and crash."
The robotic student is loaded with aftermarket instrumentation, from accelerometers and gyroscopes to magnetometers, which use the Earth's magnetic field to figure out which way the helicopter is pointed.
In the future, such a craft might prove helpful to search for land mines in a war region or to map out wildfire hotspots.
