Robot helicopter 'learns' to fly
Robot helicopter 'learns' to fly
2008-09-03Newsfeed
Computer experts have developed an artificial intelligence system that enables robotic helicopters to teach themselves to fly difficult stunts by watching other aircraft perform the same manoeuvres.
Computer experts have developed an artificial intelligence system that enables robotic helicopters to teach themselves to fly difficult stunts by watching other aircraft perform the same manoeuvres.
The Stanford University team, lead by professor Andrew Ng, said the stunts are "by far the most difficult aerobatics manoeuvres flown by any computer-controlled helicopter".
The technology demonstrates "apprenticeship learning", in which robots learn by observing an expert, rather than by having software engineers type in code from scratch.
The team said the difficulty came because helicopter flight is an "unstable system" that comes unglued without constant input. The researchers said flying a helicopter has been compared to balancing a long pole in the palm of your hand: "If you don't provide feedback, it will crash."
The new technique could be used to allow autonomous helicopters to search for land mines in war-torn areas or to map out the hot spots of California wildfires in real time, allowing firefighters to quickly move toward or away from them.
"In order for us to trust helicopters in these sort of mission-critical applications, it's important that we have very robust, very reliable helicopter controllers that can fly maybe as well as the best human pilots in the world can," Ng said.
Copyright © The Press Association 2008
