Reward function

This topic contains 3 replies, has 3 voices, and was last updated by Maxim Kuzmin 3 years, 3 months ago.

Viewing 4 posts - 1 through 4 (of 4 total)
  • Author
  • #27655
    Maxim Kuzmin

    Can we get parameters that are used to compute the result of reward function? Is this result the only kind of output, related to duckiebot control quality?



    Hey Maxim,
    Here’s a link to the AIDO objective functions.
    I’m not sure if they’re implemented anywhere yet, but maybe have a look through their baseline code?
    Good luck!

    Julian Zilly

    Hi Maxim,
    Thanks for asking. We will make the code to compute the objective functions available once they are available in full. This will closely follow the performance, traffic law abidance and comfort metrics outlined in the AIDO rules ( There will be reward functions but what is ultimately scored are different cost functions (negative reward essentially).

    Maxim Kuzmin

    Thanks for answers!
    It will be very interesting to see those functions in source code.

Viewing 4 posts - 1 through 4 (of 4 total)

You must be logged in to reply to this topic.