6
$\begingroup$

I want to make a drone which can follow static and dynamic waypoints. I am a total beginner in the drone field so I can't figure out whether I should use Reinforcement Learning or other learning methods for the drone to make it follow both static and dynamic waypoints. If RL is the best choice for the task, then how would I go about training the model before uploading it to the flight controller. And if RL is not required, then what should I use in order to achieve this task?

Currently, I have no clue about which software to use for the task. I am very interested in this field and I would love to learn more. Please let me know how I should begin with this task.

Improve this question
$\endgroup$
10
  • 1
    $\begingroup$ Hi Ravish, welcome to Drones.SE! As you've currently phrased your question, it encompasses multiple distinct topics, which isn't a great fit for our Q&A model. Please narrow the scope of your question and/or break it up into multiple separate question threads. $\endgroup$
    – ifconfig ♦
    Commented Jun 10, 2020 at 17:57
  • $\begingroup$ I have made the edits that make it directed on the 'waypoints' problem only. I really need help to get started with this. Do let me know if more changes are required, thanks.\ $\endgroup$
    – Ravish Jha
    Commented Jun 10, 2020 at 17:59
  • $\begingroup$ The body text of your question is, unfortunately, missing details. As I read it it isn't clear to me whether you're asking for ground simulation or real-time controller software recommendations. $\endgroup$
    – ifconfig ♦
    Commented Jun 10, 2020 at 18:04
  • $\begingroup$ The question body also contains remnants from your previous title asking about recommendations for a flight controller. Maybe this ask would be more clear if you provide some more information about what you're trying to build? $\endgroup$
    – ifconfig ♦
    Commented Jun 10, 2020 at 18:05
  • 1
    $\begingroup$ There is also an artificial intelligence stack exchange site, which answers questions about reinforcement learning. But you might get an answer here. Let's see! $\endgroup$
    – Nike Dattani
    Commented Jun 10, 2020 at 20:20

1 Answer 1

Reset to default
3
$\begingroup$

Reinforcement Learning can be used for UAV control. This video details a method to use RL for quadcopters.

In summary, you program a model of your UAV that effectively simulates your drone. You then create a policy and value neural network to control your drone and train it using your simulation.

I'm pretty sure similar methods are used in zipline's autonomous drone delivery network that manages 65% of the entire country of Rwanda's blood deliveries.

Improve this answer
$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.