Shop OBEX P1 Docs P2 Docs Learn Events
Positioning in 3D — Parallax Forums

Positioning in 3D

Alex SulkowskiAlex Sulkowski Posts: 13
edited 2004-10-04 19:23 in BASIC Stamp
Looking for suggestions.
I am controlling a robotic arm with a stamp.· The first couple sections of the arm are very long (total 35 feet) and do not have very fine control.· The second portion of the arm has very short sections with very fine control.· I am trying to position the arm very accurately to a location in 3D relative to a fixed position.··I first move the first long section of the arm which gets me within the rough location I need to be, and then use the fine section to get to the exact location.· My problem is that I do not know how to identify the exact location that the end of the arm is at (need to be within an 1/8th inch at most) relative to the fixed position.· Does anyone know of any devices that provides very accurate 3D positioning that the stamp could interface with.· The 3D space that the arm is working in is 10ft Wx 60ft Lx 40ft H.· Perhaps a commercial device exists where a radio signal emitter could be located on the end of the arm with several receivers located on the ground·with a computer performing·the triangulation.··Building anything like this is certainly beyong my abilities, but I am not averse to buying·a complete system and interfacing·it to the stamp.·

Thanks in advance for any ideas (or even telling me that this cannot be done)
-Alex

Comments

  • K de JongK de Jong Posts: 154
    edited 2004-09-24 21:40
    RADAR may be something for you, if you don't have too much reflections.

    Your location looks like a warehouse to me from the dimensions you give. Correct?

    Let's see what comes up from the others.

    Regards,

    Kaus
  • Alex SulkowskiAlex Sulkowski Posts: 13
    edited 2004-09-26 01:20
    Thanks for the reply Kaus.· Good guess based on my specs.· Actually not a warehouse, but the side of a house.· I did not think that I could achieve 1/8th inch positioning accuracy using radar, however, I certainly do not know alot about radar.· Do you have any suggestions of existing radar equipment (that I will interface to with the basic stamp).·

    Anybody have any other suggestions for local 3D positioning with very high accuracy (1/8th inch)?

    Thanks,
    Alex
  • ionion Posts: 101
    edited 2004-09-26 02:48
    Alex,

    Depending on your motors to move the arm you·can have any type of precision. In a normal robot arm it is incorporated an encoder. The output of the encoder can be read by stamp and gives you teh exact position of the arm. The only procedure that you have to do is at power up to command the robot to get to a home position. Home means to make some limit switches on each portion of the arm stop. When all the section of the arm are home, you have to reset the count of all the encoders to zero , then you are set. ·To go to a preset postion, you must teach the robot how much to move each arm from zero.

    Once the teaching is done, you can put the robot to repeat the same movement over and over and it will be allways in the same place. It is a long story to write here but this is the basic principle.

    How many joints has your robot. The industrial ones usualy have 5 or 6. 3 for moving the arm in place (x,y,z), and the 2 or 3 for end of arm tool.

    Need more· info, just ask

    Goodluck

    ion
  • Alex SulkowskiAlex Sulkowski Posts: 13
    edited 2004-09-27 00:42
    Hi ion.· I agree that encoders·are the easiest way to typically solve this problem, but I did not think that I could use encoders because the length of the first two segments of the arm is so long (20 and 15ft).· The degree of error (>1/8th inch) in the·positioning at the end of 35 ft would be too large given the resolution of encoders I have seen.· These segments of the arm use hydraulic cylinders, not motors.· If you have any suggestions of encoders that would be appropriate in this situation, I am all ears.

    The arm has 6 degrees of freedom. The arm is on a base that rotates.· The first two segments are powered using hydraulics.· The rest use stepper motors to provide very precise positioning.·· I am using a computer to submit commands to a BS2P which controls (using additional hardware)·the hydraulics and stepper motors (I have not hooked up the stepper motors yet since I am trying to solve the positioning of the hydraulic powered segments of the arm first).

    Thanks,

    -Alex
  • Robert KubichekRobert Kubichek Posts: 343
    edited 2004-09-27 01:02
    Wy don't you use optical positioning,;
    1) a bar coded white/black strip to indicate length of hydraulic cylinder extended. would be limitted by how good the bar/code /optical sensor.
    2) a laser distance mesuring from 2 or 3 preset locations.

    choice 1 would be easier, choice 2 more involved....


    Bob N9LVU smilewinkgrin.gif
  • freq42freq42 Posts: 18
    edited 2004-09-27 13:15
    I work in a factory *** an electrician that has four very large 100+ton presses four extruding aluminum. All the presses are hydraulic and wee use·temposonic transducer's http://www.acshydraulics.com/pages/tempo.html·four knowing wear the end of the ram/pearser and the loader arm's are at. They can be extremely accurate four measurements. You can find them on Ebay or on the web.
    ·yeah.gif
  • Alex SulkowskiAlex Sulkowski Posts: 13
    edited 2004-10-04 00:51
    Thanks for all the replies.· Based on the suggestions, I will try to use encoders to calculate the arm's position.· Thank you all.· As usual, this group always has the answers. -Alex
  • Eric REric R Posts: 225
    edited 2004-10-04 01:56
    Most encoders have between 60-120 counts per revolution. Simple gearing could enhance that number greatly giving more turns to the encoder vs. movement of your arm.
  • KenMKenM Posts: 657
    edited 2004-10-04 19:23
    For high resolution optical encoders, check out

    www.usdigital.com

    Ken
Sign In or Register to comment.