Basic Moves
- Move the arm to specified joint positions
- Teach trajectories using freedrive mode
- Replay a trajectory:
Move the arm to specified joint positions
Important
rosrun hand_arm move_simple.py [COMMAND] [ARG]
With a simple rosrun command, you can set named arm positions using teach mode.
| Command | Arg | Description |
|---|---|---|
| goto | position_name (str) | Move the arm to a saved position. |
| set | position_name (str) | Set and store the arm’s position. This enable’s freedrive mode so you can push the arm to a desired pose before saving. |
| setnow | position_name (str) | Set and store the arm’s current position. |
| list | n/a | List the names and values of all saved poses |
| listnames | n/a | List the names of all saved poses |
Examples:
rosrun hand_arm move_simple.py goto zero- Move to the zero position (straight up)rosrun hand_arm move_simple.py set home- Set the home positionrosrun hand_arm move_simple.py goto home- Move to the home positionrosrun hand_arm move_simple.py list- List out all the saved poses
Positions get saved in config»arm_poses.yaml, so you can add/edit that file directly if you want.
Teach trajectories using freedrive mode
When using teach mode, the robot will be put into freedrive mode, enabling you to push it around.
Enable freedrive mode without saving anything
rosrun hand_arm teach.py
If you just need to push the arm into a configuration before you start other operations, this is how to do it.
Teach a trajectory from start to finish
rosrun hand_arm teach.py [FILENAME]
Save a trajectory continuously at a rate of 20Hz. This option is meant for debugging, but maybe it’s useful for other things.
Teach a trajectory via individual waypoints:
Important
rosrun hand_arm teach_points.py [FILENAME] [FLAGS]
Teach the robot using waypoints. Waypoints are saved in joint space, and hand moves are saved in pressure space.
- filename: Name of the file to save the trajectory to.
- flags: flags for specific options
- cartesian: Use cartesian end effector poses rather than joint angles. (requires MoveIt! to be running)
- robotiq: Use a robotiq 2F gripper rather than a pressure controlled hand. (requires robotiq_trajectory_control package)
Once running the command, you can use keyboard commands to save trajectory points and control your gripper:
| Command | Function |
|---|---|
| Space | Save the current pose as a waypoint in the trajectory |
| Shift | Toggle continuous capture of poses on/off |
| F | Toggle freedrive mode |
| G | Perform a grasp |
| R | Release the grasp |
| Ctrl + C | End the program and save the trajectory |
Example:
Teach a trajectory called “test_move”, and put it in the “testing” folder:
rosrun hand_arm teach_points.py testing/test_move- Use keyboard commands to add waypoints, grasps, and releases to the trajectory
- Use Ctrl+C to end. The trajectory is then saved:
- “trajectories/.arm/testing/test_move/trajectory.yaml”
- After saving, the trajectory is automatically “planned” (i.e. converted to a “.traj” file) for playback.
Replay a trajectory:
Important
roslaunch hand_arm run-traj-taught.launch config:=arm_hand.yaml \
traj:=[FILENAME] speed_factor:=[SPEED FACTOR] reps:=[NUM REPS]
“speed factor” is a multiplier on the speed (i.e. 0.5 is half speed, 2.0 is double). If you increase the speed, the robot becomes less accurate (i.e. it’s not able to move fast enough to get to every waypoint). Use the speed factor at your own risk
Example:
Run the “testing/test_move” trajectory you taught:
roslaunch hand_arm run-traj-taught.launch config:=arm_hand.yaml \
traj:=testing/test_move