35
Intera 3.2
Training and Managing Tasks
Training a Pick and Place
Baxter can pick up objects blind or by using its cameras.
•
Blind – Baxter picks an object from a fixed location.
•
Visual – Baxter learns the shape of an object, and uses vision to identify it and
pick it up. Objects may or may not be in the same location. See “Training a Pick
Using Vision” on page 36
Note:
Vision is only available for vertical Pick actions.
To train a basic Pick and Place task:
1. In the main button bar, click New.
2. Place the object you would like Baxter to grasp in the fixed location of the Pick.
3. Move the arm to the location, with the gripper’s fingers poised to grasp the
object. (If using a vacuum gripper, place the suction cup on top of the object.)
4. Press the Grasp button on the training cuff. Baxter grasps the object.
5. Move the arm to the location where you want to place the object.
6. Press the Grasp button. Baxter releases the object.
7. Press Back to open the Main Button Bar, then select Reset or Run to perform the
task.
Tip:
If Baxter’s wrist is not pointing straight down when performing an action, meaning
the action is not vertical, a red circle is displayed next to the blue (for Baxter’s right arm)
or green (for Baxter’s left arm) pointer. Non-vertical Picks cannot use vision. If the pointer
is not blue or green but gray, Baxter cannot create an action in that location/pose.
You can change the number of times Baxter will attempt to Pick up an Object before stopping. For
details, see “Change Number of Attempts to Pick an Object” on page 59.
