Symbiosis between drones and AI: 11500 crashes to learn how to fly

Posted on Posted in Artificial Intelligence, Technology

Hey my friends, this is the article 48 / 1000! For those who don’t know me, my name is Selim Chehimi and I’m an engineering student. I’ve been programming for 7 years now so you can tell that I really enjoy that. Those 7 years of programming lead me to Artificial Intelligence. I really want to be involved in AI so that’s the reason why I’m sharing this article with you. I hope you’ll enjoy reading it as much as I’ve enjoyed writing it. Today, we’ll be talking about an amazing experiment: the symbiosis between AI and drones. Let’s begin.

 CMU (Carnegie Mellon University) researchers are letting a drone learn how to fly by himself using a deep neural network. In this experience, the drone succeeds if… it fails to crash! The drone start at a random location in space and then fly forward. When it crashes, it comes back to the initial position and chooses another destination. By crashing over and over, it’s getting a lot of experience which lead him to know how to fly.

How does it work?

Just before crashing, images of what’s forward are recorded and fed into a deep neural network. Consequently, the drone can know when going forward is good or not. After 11500 crash and 40 hours of flying time, the drone is becoming great and can fly autonomously in any environment. Here is a video of this drone in action:

Thanks so much for reading, that’s the end of Article 48. Please, share it and don’t forget to follow me on social media. You can also subscribe to my newsletter below to never miss any article. Don’t hesitate to tell me what you think of this Article in the comments and stay tuned for another one tomorrow!


Learn The Basics of Artificial Intelligence!

Receive a FREE eBook that brings all the pieces to give you the big picture of Artificial Intelligence.

I will never give away, trade or sell your email address. You can unsubscribe at any time.

Leave a Reply

Your email address will not be published. Required fields are marked *