Alberto Romay's profile - activity

Have you tried:

source /opt/ros/your-ros-version/setup.bash

before executing gazebo?

I had that issue once and it was because I was not sourcing the correct bash file.

The hand force/torque sensor behaves well until the BDI modes are changed. While in STAND mode the output of the sensor is correct, however, when the mode is changed to MANIPULATE for example, the force/torque sensor start giving a nonsense output.

Hand force sensor while changing BDI modes

The X component from the torque sensor of the hand is buggy. Z and Y components of the torque sensor in the wrist are giving an output while X is not. Discovered that only when the joint limits from the r_arm_mwx joint are reach the X starts giving an output.

Hand torque sensor 1 shows how the Z component is performing well, while the X is giving an output only when it reaches the joint limit (2:40).

Hand torque sensor 2 shows how the X component gives output only in the joint limits (upper 0:31 and lower 3:20).

This pull request addresses the issue of the location from some vehicle parts (steering wheel raised, seat lowered) but it is not clear to me if it also addresses the issue of the fingers passing through the steering wheel like if there is no collision. Example: The hand grabs the steering wheel (surrounds it with the fingers) then the robot pulls the hand and the steering wheel gets out through the fingers. I can make another video if the last one was not clear.

The simulation allows the fingers to pass through the steering wheel, which makes it difficult to hold the hand while trying to turn the wheel.

Sandia collision break through

In 0:35 it can be seen that there is some kind of "transparency" in the collision polygon, or it is in the hand or it is in the steering wheel, but this shows that some collision polygons are not 100% solid.