2 comments

  • daww2 days ago
    Author here.<p>I&#x27;ve spent the last six months replicating the paper &quot;Champion-level drone racing using deep reinforcement learning&quot; and now I&#x27;m writing down the blog posts I wish I had along the way.<p>Any feedback is welcome, especially as I&#x27;m a bit unsure if I struck the right balance between being concise and not requiring too many prerequisites.<p>Also if you&#x27;re working on RL and robotics (especially aerial), let&#x27;s connect!
    • avidiax1 hour ago
      I assume you are going to start introducing all the 2nd and 3rd order effects? One big one is ground effect, and another is vortex ring state&#x2F;settling with power and the related translational lift, and the props themselves have p-factor and the dirty air effect for the rear props.
    • dvh1 hour ago
      Isn&#x27;t it just bi-copter?
      • palata28 minutes ago
        No, it&#x27;s a quadcopter setup, but simulated in a 2D world (I guess for simplicity). A bi-copter would require tiltrotors, which is different.
  • kabir_daki1 hour ago
    Physics simulations from scratch are great learning projects. Did you implement your own PID controller for stabilization? That&#x27;s usually where things get interesting — tuning the gains without it oscillating to death.
    • andai48 minutes ago
      Your voice betrays you, Claude!