Quadcopter and machine learning

Amazon, DHL, FedEx… All those big logistics companies believes that UAV’s can revolutionize the transportation world, and for a good reason ! All these brands developed prototypes of a particular UAV, the quadcopter. For example, recently, Amazon unveil his Prime Air, a quadcopter who’s having the capacity to deliver individual packages to customers’ doorsteps within 30 minutes of ordering, using only the GPS technology !

These type of UAV’s are more and more coveted because they are having exceptional capacities. The first and the most important ability, is the agility. A quadcopter can move wherever the human wants while remaining stable. That’s why they are very useful in case of a human disaster for example, to find survivor or anything that can help the first-aid worker.

The second quality is the capacity to hover, unlike the sailplane UAV's. This is very useful to take pictures for example. And the last important ability is the athletic power of the quadcopter. The design of the quad allows the UAV to fly by just two rotors instead of four. In addition of the capacity of carrying some heavy objects, the quad represent a very interesting futur solution for industries.

The machine learning field is more and more interested by UAV’s, and in particular quadcopter. A lot components adapted to quadcopter are created to allow machine learning. For example, some sophisticated drone use board computer, laser detector, sensitive camera… Those component enable the drone to navigate in an unknown environnement using just his algorithm to know which direction he should take. They can take decision without the help of any technology like GPS, for example. 

All those abilities allow us to dream about a useful futur for those machines but we still have to improve some important characteristics like the battery life or the legal framework for the UAV’s flight !