Thursday, November 8, 2012

Obstacle avoiding drones developed

Those hobbyists who love flying remotely controlled airplanes and helicopters often experience their machines bumping into trees or building and get destroyed or damaged in these 'unwanted' encounters. Controllers with remote control devices controlling the flight of these flying machines often find it difficult to manoeuvre around their machines and save those form damage or even destruction.

Obstacle Avoiding Drone
Obstacle avoiding drone helicopter

But not anymore. 

Researchers at Cornell University have developed software that, coupled with a camera and hardware that mimics a brain, helps a small drone to dodge obstacles like trees or poles.
The software makes the drone turn an image taken with its camera into a 3D model of the environment. The robotic brain can then determine which objects are obstacles and change its route accordingly. This would solve one of the biggest hurdles towards fully autonomous drones: their lack of good “sense and avoid” systems.
Watch the video below for a demonstration of the outcome of the research that can save the drones/helicopters from damage and save money of the poor RC gadget hobbyists:

The lead researcher, Ashutosh Saxena, hopes to further develop the software capabilities making the drone able to calculate wind patterns and avoid moving objects, like birds.


Post a Comment