An MIT researcher has developed a one-of-a-kind drone that not only zips through the air at the speed of 30 mph, but also dodges trees in the process. "Everyone is building drones these days, but nobody knows how to get them to stop running into things," Andrew Barry, who developed the system as part of his thesis with MIT professor Russ Tedrake, said in a statement.
In order to address the problem, Barry developed a stereo-vision algorithm that enables the drone to detect obstacles and map them in real-time. The drone's action is determined by its response to information captured by a camera built on its wings. Both search through the depth-field at multiple distances so that objects are immediately detected if they are on the drone's path. Watch it in action below:
The details of the drone technology can be accessed through this link. What distinguishes Barry's drone technology is its capacity to achieve its current speed while packing computer hardware that can process the amount of information required for it to dodge obstacles.
"Sensors like lidar are too heavy to put on small aircraft, and creating maps of the environment in advance isn't practical. If we want drones that can fly quickly and navigate in the real world, we need better, faster algorithms," Barry said.
The drone is reportedly easy and cheap to build. Its components can be easily sourced and they only cost $1,700, according to Popular Science. Presently, the drone technology is still experimental but it is expected to contribute in the current drone development, particularly in the areas of safety and maneuverability.
A similar drone delivery system is now promising groceries in 30 minutes or less. To read more, CLICK HERE.
