Shop OBEX P1 Docs P2 Docs Learn Events
Erco's Peg Puzzle Challenge: Arm Yourself! — Parallax Forums

Erco's Peg Puzzle Challenge: Arm Yourself!

ercoerco Posts: 20,254
edited 2015-10-28 02:04 in Robotics
Figure 8 Challenge? Pishaw.
Oval Challenge? Yawn.
Towers of Hanoi? Gimme a break.

Martin_H, PhiPi, Duane... who will answer the call? The price of entry? $1. That is, you'll find the peg puzzle at the dollar store!

Sorry PhiPi, the solution may not be recursive: http://www.joenord.com/puzzles/peggame/index.html

Comments

  • TtailspinTtailspin Posts: 1,326
    Leave 8 pegs on the board with no more moves, and you're a super duper genius...

  • ercoerco Posts: 20,254
    First responder, you're in, Tommy Tailspin!
  • TtailspinTtailspin Posts: 1,326
    edited 2015-10-29 05:00
    I got tired of drilling the holes for this puzzle board. I must have made 50 of them by now.
    However, maybe just for you, i will drill one more board and turn 14 more pegs, before Xmas..But, the Robot arm is months away..

    I do have some thin stock i put aside for a steamy glue up session.. it's the good stuff from my private reserve,(2000 years old).
    I may not have to sell all my woodworking tools just yet, there may indeed be just one more thing to build... :)


    -Tommy
  • Martin_HMartin_H Posts: 4,051
    An interesting challenge, one which I'm not sure any of my robot arms have the accuracy to complete. They're usually only accurate to within a millimeter or so.
  • MikeDYurMikeDYur Posts: 2,176
    I would like to try something with the servos being incorporated internally, this is a challenge!
  • Martin_HMartin_H Posts: 4,051
    Erco, I went to the local dollar store and they don't have the peg puzzle or golf tees. Neither did the local toy store. So I'm going to have to improvise.
  • ElectrodudeElectrodude Posts: 1,621
    Martin_H wrote: »
    Erco, I went to the local dollar store and they don't have the peg puzzle or golf tees. Neither did the local toy store. So I'm going to have to improvise.

    Have you tried your local Cracker Barrel? :)
  • ercoerco Posts: 20,254
    Martin_H wrote: »
    They're usually only accurate to within a millimeter or so.

    Well now you're just braggin'...

    We'll have to get one of our 3D printing gurus to make a special version of this triangle game just for robotics. Instead of golf tees, a ~3/4" hollow hemisphere with a grasping peg on top, and a triangular base with 15 molded domes to automatically center the piece on top.

  • Phil Pilgrim (PhiPi)Phil Pilgrim (PhiPi) Posts: 23,514
    edited 2015-11-01 21:11
    Instead of golf tees, use steel bearing balls. The base can then consist of 15 heavily countersunk holes that the balls automatically center in.

    The "gripper" can be made from a plastic sleeve bushing, a flat-head machine screw, and a tiny super-magnet the same size as the screw head. The magnet adheres to the screw head, and that assembly slides into the plastic sleeve. By moving the screw in and out of the sleeve via a servo or solenoid, the magnet approaches or retracts from the bottom sleeve opening. When near the bottom, it will attract and pick up the bearing, without the bearing having to touch the magnet. The bearing is then automatically centered against the bottom of the plastic sleeve. When the screw-and-magnet assembly is retracted, the bearing will drop.

    -Phil
  • ercoerco Posts: 20,254
    PhiPi saves the day... AGAIN! Great call as ever, Phil!
  • PublisonPublison Posts: 12,366
    erco wrote: »
    PhiPi saves the day... AGAIN! Great call as ever, Phil!

    Yes, the guy never sleeps! :)
  • ercoerco Posts: 20,254
    Publison wrote: »
    Yes, the guy never sleeps!

    But when he does, he dreams up great stuff!

  • Martin_HMartin_H Posts: 4,051
    Cool idea Phil.

    I was contemplating cutting a dowel into 2" long pieces. On the tops I was going to glue smooth plastic disks so I could pick them up with my suction cup gripper. But smooth plastic balls would work better, and counter sunk holes would center them close enough for the grab to be consistent.

    Now I just need some smooth plastic balls about an inch in diameter.
  • ercoerco Posts: 20,254
    Ping pong balls? Done.
  • Martin_HMartin_H Posts: 4,051
    Challenge accepted.
  • ercoerco Posts: 20,254
    Oops, I just learned it was DOLLAR TREE where I bought my puzzle, if anyone cares. In the toy aisle.

    peg2.jpg
  • WhitWhit Posts: 4,191
    edited 2015-11-11 14:07
    Those "T"s smell a little plastic-y - even online!

    erco - come on! What would Matt say?
  • LoopyBytelooseLoopyByteloose Posts: 12,537
    I thought the whole reason for programing microcontrollers was to have a bigger more challenging puzzle.

    The problem with this one is that Google gives you all the answers.
  • WhitWhit Posts: 4,191
    @Loopy - with a Robotic Arm and Gripper - it is not just solving, but dexterity (and in the proposed case - very fine movements!).

    Having the robot "look" at the puzzle, design a solution and then carry out the solution ARE multiple problems - each a pretty complex task at that.

    I don't think even the most sophisticated robots in existence could just let you put and puzzle in front of them, see it, solve it and return it finished. That would be a amazing accomplishment.
  • ercoerco Posts: 20,254
    Whit wrote: »
    Those "T"s smell a little plastic-y - even online!

    erco - come on! What would Matt say?

    At least the base (the biggest piece and 90% of the puzzle mass) is...

    PLYWOOD!

  • MikeDYurMikeDYur Posts: 2,176
    Whit wrote: »
    @Loopy - with a Robotic Arm and Gripper - it is not just solving, but dexterity (and in the proposed case - very fine movements!).

    Having the robot "look" at the puzzle, design a solution and then carry out the solution ARE multiple problems - each a pretty complex task at that.

    I don't think even the most sophisticated robots in existence could just let you put and puzzle in front of them, see it, solve it and return it finished. That would be a amazing accomplishment.

    IBM Watson must have delicated his life to medicine. With a precision arm, could be programmed with every game in the book.
  • ercoerco Posts: 20,254
    Speaking of arming yourself, here's a newish $200 ER-1 arm for you rich guys: http://www.ebay.com/itm/Evolution-Robotics-ER-1-ER1-ARM1-ARM-1-Personal-Robot-System-Gripper-Arm-/191755139412?

    Martin, how are you coming on this challenge? I'll get back to it after the Christmas crunch.
  • Martin_HMartin_H Posts: 4,051
    edited 2015-12-08 21:31
    I went up a blind alley with the IK code over Thanksgiving weekend, after debugging it I realized the sample code I found on Github was rubbish. It was obvious in retrospect the author never tried it. I then found some new sample code on the Trossen Robotics forum last weekend, it worked, so I got the IK working on Sunday.

    What I have left to do is work out the gcode for the motion script. This is tedious, and the least technically challenging part of the project. It's also the part that's hard to stay motivated on because it is huge amounts of trial and error.

    The reason I need new IK code is this robot is neither a scara arm, nor revolute arm. So I haven't done this sort of IK before.
  • Martin_HMartin_H Posts: 4,051
    (0, 65)
    (-9, 45) (9, 45)
    (-19, 32) (0, 32) (20, 32)
    (-27, 17) (-9, 17) (9, 17) (28, 17)
    (-37, 0) (-19, 0) (0, 0) (19, 0) (38, 0)
Sign In or Register to comment.