Autonomy Upgrade with Drone Using Visual Learning

Autonomy is a wonderful thing when it comes to drone technology. While it is often portrayed as something horrific in fiction, in real world, it is the opposite.

The Power of Autonomy

Autonomy in drones means less manpower involved while still getting the job done. Using a program, they can go to their routes and deliver packages, travel distances, and do other tasks without manual input. This frees up an enormous amount of time when using drones. Before, a drone requires manual control during the entire trip. This takes a lot of time, as well as effort from the person navigating. Through the use of GPS, a person can simply monitor the progress of the drone from a control station.

Unlike its portrayal in science fiction movies, drones today are far more useful than robot invaders. While I agree it is far more dull that way, I think we can all agree that it is for the best.

Visual Teach and Repeat

In the University of Toronto’s Institute of Aerospace Studies (UTIAS), a drone learns to fly around with nothing but its vision. Using its camera to map out its surroundings, the drone can, in theory, find its way to its destination without any previous directional input. Yes, the UTIAS’ drone can fly to its target without the use of GPS to mark its location.

Using artificial intelligence on the drone, it can decide for itself where to go, with a bit of help from its sight. This means that in the event its GPS programming fails, it can still fly around without any assistance.

They are calling this drone project the Visual Teach and Repeat (VTR). This program, which can simply be added to drones, can upgrade a drone’s navigation immensely. The idea is that a drone can pick up a trail using its surroundings to follow a path. Using landmarks or visual cues, it can find its way home without any help from a controller.




Leave a Comment