Forward walking "drifts" left

Korean company maker of Robot kits and servos designed for of articulated robots. Re-incarnation of Megarobotics.
4 postsPage 1 of 1
4 postsPage 1 of 1

Forward walking "drifts" left

Post by careyg » Wed Aug 11, 2010 8:44 pm

Post by careyg
Wed Aug 11, 2010 8:44 pm

hi Folks,

When huno is walking forward, he drifts maybe 5 degrees to the left per step. I think i have seen similar problems in other folks videos on youtube, etc.

Anyone know if this is a basic-pose issue (i.e. i need to try harder), construction issue around one or more of his joints, or a motion issue (stride for one side longer than the other)?

Either way, I suspect the solution is to modify the stride motion to correct, and i may investigate changes to the stride to allow him to turn whilst walking anyway, but wondered if others had identified (and resolved?) the issue themselves in the past?
hi Folks,

When huno is walking forward, he drifts maybe 5 degrees to the left per step. I think i have seen similar problems in other folks videos on youtube, etc.

Anyone know if this is a basic-pose issue (i.e. i need to try harder), construction issue around one or more of his joints, or a motion issue (stride for one side longer than the other)?

Either way, I suspect the solution is to modify the stride motion to correct, and i may investigate changes to the stride to allow him to turn whilst walking anyway, but wondered if others had identified (and resolved?) the issue themselves in the past?
careyg offline
Robot Builder
Robot Builder
Posts: 21
Joined: Tue Jul 01, 2008 10:20 pm

Post by l3v3rz » Wed Aug 11, 2010 10:52 pm

Post by l3v3rz
Wed Aug 11, 2010 10:52 pm

Given this is all based on dead recogning its amazing it walks in a straight line. Really this needs to auto correct based on an external reference such as using a compass to modify the stride dynamically. I assume thats what Flid is up to with his I2C based compass sensor. I think it would be good idea just give robots a simple couple of CdS sensors connected to A2D port so that you could locate a light source and walk towards it.

cheers
Given this is all based on dead recogning its amazing it walks in a straight line. Really this needs to auto correct based on an external reference such as using a compass to modify the stride dynamically. I assume thats what Flid is up to with his I2C based compass sensor. I think it would be good idea just give robots a simple couple of CdS sensors connected to A2D port so that you could locate a light source and walk towards it.

cheers
l3v3rz offline
Savvy Roboteer
Savvy Roboteer
Posts: 473
Joined: Fri Jul 18, 2008 2:34 pm

Post by careyg » Wed Aug 11, 2010 11:57 pm

Post by careyg
Wed Aug 11, 2010 11:57 pm

:) glad it's not just me then!
:) glad it's not just me then!
careyg offline
Robot Builder
Robot Builder
Posts: 21
Joined: Tue Jul 01, 2008 10:20 pm

Post by Flid » Sat Aug 14, 2010 11:57 am

Post by Flid
Sat Aug 14, 2010 11:57 am

Same here :x

l3v3rz wrote:I assume thats what Flid is up to with his I2C based compass sensor.

Right :wink:


l3v3rz wrote:I think it would be good idea just give robots a simple couple of CdS sensors connected to A2D port so that you could locate a light source and walk towards it.

Or maybe with I2C camera from wiimote like i-Bot hack :wink:
Same here :x

l3v3rz wrote:I assume thats what Flid is up to with his I2C based compass sensor.

Right :wink:


l3v3rz wrote:I think it would be good idea just give robots a simple couple of CdS sensors connected to A2D port so that you could locate a light source and walk towards it.

Or maybe with I2C camera from wiimote like i-Bot hack :wink:
Flid offline
Savvy Roboteer
Savvy Roboteer
Posts: 34
Joined: Sun Jul 11, 2010 11:34 pm


4 postsPage 1 of 1
4 postsPage 1 of 1
cron