You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When driving PR2's around the building, we've noticed a few of them yawing while commanding forward. It appears to be related to the calibration of the casters. When recalibrated, the direction and magnitude can change.
It'd be nice if the Texai caster calibration controller was ported to the PR2 so we could have really good calibration. We would need to find a way to deal with the "hammer" behavior that jams the caster back and forth.
[wim] I'd be the right person for this, but probably won't get to solve this before Diamondback (if it's not already too late; pr2_controllers might be locked down already). This is a known issue, but AFAIK it is not causing any problems for anyone, except for the inconvenience while joysticking the robot around. Is that true?
The fix is more involved than simply copying the texai calibration controller because the pr2 has a wiggle recovery behavior in the calibration controller to recover when a caster gets stuck.
[watts] Wim and I talked about this, and I talked about it with Eric. For production and support, we'd really like to see this in place before D-turtle. Since it is not an API breaking change, it can go in somewhat late in the process. It should not go in after D-turtle is released, since we don't want any unintended consequences affecting users.
When driving PR2's around the building, we've noticed a few of them yawing while commanding forward. It appears to be related to the calibration of the casters. When recalibrated, the direction and magnitude can change.
It'd be nice if the Texai caster calibration controller was ported to the PR2 so we could have really good calibration. We would need to find a way to deal with the "hammer" behavior that jams the caster back and forth.
trac data:
The text was updated successfully, but these errors were encountered: