[Bug]: WidowX gripper moving in y-direction opposite to the commanded input #239
Labels
Comments
abadithela
added
bug
|
Please enable debug logging on the arm's xs_sdk provide the output when running your script. Please also ensure that the commands you are providing to the arm do not exceed the joint limits and are feasible according to the solver used by the set_ee_cartesian_trajectory method.
I do not think that updating the firmware would resolve this issue but it wouldn't hurt to keep everything up-to-date.
Yes, it is compatible. |
|
What is the command to enable debug logging? We ran the following but it did not seem to do anything: Yes, the commands we are providing to the arm do not exceed joint limits. Every argument to set_ee_cartesian_trajectory is on the order of 0.01 at most. |
|
I didn't catch that you're running on ROS 1 Noetic. Debug logging is not available as a simple command line arg in that distro. You will instead have to configure rosconsole to achieve debug logging. |
|
When I've tried debug logging, it gives me the following output on terminal when I'm seeing this failure: |
|
For reference, here is the debug console output when I successfully command the robot to a certain state. I am curious about the |
|
Hi, We're still facing this issue about 1 in 3 experiments. Is there any way to diagnose this issue? I've plotted the joint states over the course of operation (the relevant part of the plot is after time step 100).
I think the waist joint angle (dark blue / position[0] plot) illustrates this issue. It continues to be positive when the expected behavior should be that it eventually decrease and become negative. Best, |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
What happened?
Hi,
I am using the WidowX 250s on UBuntu 22.04 with ROS Noetic. I have two questions. First, I am using a policy to command changes in gripper position, and I am using the bot.arm.set_ee_cartesian_trajectory(x=dx, y=dy, z=dz, roll=droll, pitch=dpitch, yaw=dyaw) to update the gripper position. We're noticing the following failure mode about once or twice every 10 trials. The gripper moves in a dy position opposite to the commanded dy. For example, the commanded dy position can take a small positive value (which is supposed to shift the gripper to the left), and followed by a larger negative value (which is supposed to shift it to the right) as given below. However, what we observe is that it does not properly execute the commands of step 22, and continues shifting further to the left.
The second question is regarding doing a Dynamixel firmware update. On the dynamixel wizard, all motors have an asterisk next to them, which I understand to mean that a firmware update is available. Would doing a firmware update resolve the above issue? More importantly, is the latest firmware update compatible with the ROS software of interbotix?
Robot Model
wx250s
Operating System
Ubuntu 22.04
ROS Distro
ROS 1 Noetic
Steps To Reproduce
No response
Relevant log output
No response
Additional Info
No response
The text was updated successfully, but these errors were encountered: