How an Autonomous Drone Flies With Deep Learning: Page 3 of 3

A team of engineers at Nvidia share how they created a drone capable of fully autonomous flight in the forest.

SLAM picks up on – resulting in huge errors to SLAM algorithms. The solution Smith said was to implement a Semi-dense SLAM specifically for rolling shutter .

Certainly the Nvidia team's project is a long way away from being the sort of fully capable drone we'd imagine in a science fiction film. The in the tests the drone was only about to fly one to two meters off the ground and was limited to a relatively slow speed. Smolyanskiy admitted that the model probably won't work at night right now without special night vision cameras and it may also need additional training to handle rain and other conditions. “The problem was not with the DNN but with the optical flow in low light conditions,” he said.

The team's goal with the project was to keep it as inexpensive as possible with an eye toward miniaturization. Smolyanskiy said given the small form factor of the Jetson TX1 and TX2 development boards it's not hard to imagine a smaller drone handling these tasks already.


Atlantic Design & Manufacturing, New York, 3D Printing, Additive Manufacturing, IoT, IIoT, cyber security, smart manufacturing, smart factorySmart Manufacturing Innovation Summit    at Atlantic Design & Manufacturing.  Designed for industry professionals looking to overcome plant and enterprise-level manufacturing challenges using IT-based solutions. Immerse yourself in the latest developments during the two-day, expert-led Smart Manufacturing Innovation Summit. You'll get the latest on the factory of future including insights into Industrial IoT and IIoT applications, predictive maintenance, intelligent sensors, security, and harmonizing IT/OT. June 13-15, 2017.     Register Today!


Chris Wiltz is the Managing Editor of Design News.


Jerald Cogswell's picture
There are non visual methods of navigation used by animals. Bats can hitting wires stretched across their path using their sonar. Owls can sense the sound of a mouse on the ground in darkness and seize it just by listening. But how can it know to avoid restricted or classified areas? How about avoiding solar concentrator paths? There is much to teach a young drone.

Add new comment

By submitting this form, you accept the Mollom privacy policy.