Full twist command for (omni-directional) Ranger

[nielsvd]

So far, using the ugv_sdk, I have only found instructions to move the Ranger Mini in one of four modes. None of these modes support simultaneous longitudinal, lateral, and angular motion; in other words a full twist command: $(v_x, v_y, \omega_z)$. Theoretically, I believe it is possible to follow such commands using four-wheel independent steering and drive; albeit with practical challenges, such as steering being limited by wiring.

Is it possible to enter such a full twist driving mode? Or alternatively and less preferred, command all eight actuators independently (accepting the risk of accidentally breaking the robot)?

[rdu-weston]

Is this what you're looking for?
https://github.com/westonrobot/ranger_ros/blob/main/ranger_base/src/ranger_messenger.cpp#L377

[nielsvd]

Thank you for the quick reply. Even though it's not precisely what I meant, it could be sufficient. This function is not available in the ros2 driver, is it? I would have to port it.

[rdu-weston]

Thank you for the quick reply. Even though it's not precisely what I meant, it could be sufficient. This function is not available in the ros2 driver, is it? I would have to port it.

Unfortunately, we haven't got time to update the ROS2 driver yet.

As to the full twist command support in the SDK, feel free to propose what you think is better or even create a pull request with your implementation. We will be happy to review it.

[nielsvd]

Thanks for the feedback. I reckon that not all twists are attainable due to the reliance on mode switching: e.g., it is not possible to let the vehicle spin while mid point moves along a straight line. This would be a good argument against implementing this logic into the core of ugv_sdk.

Do you already have an interface in mind for tentative future robots with non-mode-switching omni-drive capabilities? I suppose it makes sense to optionally wrap the Ranger Mini driver into another class that matches this interface. What do you think.

[rdu-weston]

We will work with AgileX to address this issue, as the support of this feature will likely involve both hardware and firmware. But at the moment, we don't have any timeline for it yet.