The Powder Toy

AutoLabSolve by pilotnotflying

pilotnotflying
pilotnotflying
438 / 19
13th Dec 2013
23rd Sep 2014
updated
electronic electronics pstn maze complex slow smartmovingthing robot intelligence sprk

Comments

  • pilotnotflying
    pilotnotflying     18th Dec 2013
    @snkybvr: that's right. it's the only flaw i didn't changed :D
  • Alfa
    Alfa     18th Dec 2013
    edited first path so it went in a straight line up 1 then straight right with a path in the middle it kept going to the end like *lol derp de der*
  • Racheon
    Racheon     18th Dec 2013
    awesome, +1
  • snkybvr
    snkybvr     18th Dec 2013
    It does not solve most mazes, this one is built to allow only straiht lines until it takes a turn... if there was a need to stop in the middle of a path and turn, then it would never finish. Plus one because it's still pretty cool.
  • pilotnotflying
    pilotnotflying     18th Dec 2013
    i actualy ment ntct and ptct - not pscn, nscn :)
  • pilotnotflying
    pilotnotflying     18th Dec 2013
    it is not hardcoded. every time it stops at a wall it checks possible paths by using the sensor inputs
  • pilotnotflying
    pilotnotflying     17th Dec 2013
    @travisw: it took me about a week - first i had to work out the principle ow to "decide" which path to go and then had to compres that boolean function with the quine-mccluskey algorithm
  • pilotnotflying
    pilotnotflying     17th Dec 2013
    the only way i know to make this faster is to use te console. he "decision" time cannot be changed because the nscn and pscn are tooslow to react on fast input variations.
  • megamageiii
    megamageiii     17th Dec 2013
    It follows the rule "Follow the right wall" but with the left wall because it is a hipster and following the right wall it too mainstream.
  • megamageiii
    megamageiii     17th Dec 2013
    @harrison531: You are upset that it doesn't turn when there is no wall..? That is pretty much the whole point. It completes any maze. Well, not any, but most.