No announcement yet.

Help needed with mecanum autonomous...

  • Filter
  • Time
  • Show
Clear All
new posts

  • Help needed with mecanum autonomous...

    Our team is using mecanum wheels for the first time. We have used the regular 4 inch wheels in the past. Autonomous mode was pretty easy with these wheels but now we have questions if the similar geometry holds for the mecanum wheels as well. If your team has prior experience with these, can you please help answer the following :-

    1) For straight drives does one rotation still equals to 2*pi*r of linear drive for mecanum ?
    2) How does strafing works in autonomous ? e.g. if we need the robot to straf 2 ft. to the left or right what does it translates to in term of wheel diameters and rotation (assuming all wheels rotate by same amount but different directions to achieve strafing) ?

    Any inputs are welcome.
    Thanks in advance.

  • #2
    Unfortunately the answer is - "it depends".

    My programmers just went thru this a bit ago because we are using mecanum for the first time too. They found that at lower speeds going straight was close to the 2*pi*r. The rollers on the wheel didn't really roll that much so the distance was pretty comparable. I think that at higher rates of speed the acceleration was enough that the static friction of the roller to it's shaft was broken and so you have some slippage. For strafing it was roughly a factor of 0.707 (i.e. 1/sqrt(2) ). At least that got it close. So the distances were not 100% accurate but close enough and consistent enough that they just modified the distance they were programming it to go and called it good.

    My suspicion is that if they hit tiles that are a bit different (different batch, more or less worn, etc) then the friction will change a bit and that will throw off this way of doing it.

    I think the best method is to put a couple of free-wheeling omni wheels on the bot with encoders. Those wheels should then be able to accurately measure how far the robot has moved since they won't have the inherent slipping of a mecanum wheel. But at this point I have made the suggestion to the team and there is nobody trying to implement it so I'll just let them figure out on their own if they need a scheme like that.


    • #3
      We used mecanum wheels last year and we have same observations like DanOelkeFTA. We derived empirical formula by noting multiple runs at different speeds esp the speeds we were using in autonomous.


      • #4
        Have you been able to get mecanum control to work by using the human controller? We spent quite a bit of time last year trying to get the mecanum wheels we had to work well. There was just too much slippage with the wheels we had. We tried many techniques for correction and in the end we never built enough confidence to strafe with them in Auto. We aren't using them this year. I believe the wheels we had were the Andy Mark. Searching the forums produced a lot of people having issues with these wheels.


        • #5
          We used AM hd wheels and had same issues. We added dead weights above all four wheels to make it better. Last year travel distance was short we were ok. This year we are not using mecanum as it will be long travel distance and turns


          • #6
            My team is wanting to be able to use mecanum wheels. Does anyone have a basic telope program we could use?