There are two types of skills in Mirai: Motion skills and Positioning skills. Before continue with this section, make sure you have selected the best skill type for your application: Choosing a MIRAI Skill Type
From the Mirai training App, you will find all your existing skills on the main page under "Skill Overview." To begin with a brand new skill/application, tap the button at the top of this screen "Add New Skill." Once successfully saved, the new skill will be displayed at the top of the list from the Skill Overview page.
Quick Links:
- Skill Selection (Step 1/4)
- Skill Configuration (Step 2/4)
- Camera Settings (Step 3/4)
- Reference Position (Step 4/4)
- Video Tutorial
Skill selection (Step 1/4)
Step 1
To begin the training, tap on the menu button in the upper left corner and select ‘Training’. Ensure that the MIRAI controller is connected to the internet LAN.
Step 2
Tap on ‘Add new skill’ to open the ‘Skill Selection’ form. Set a ‘Skill Name’ in the first field.
Step 3
MIRAI offers two types of skill. For more in-depth information on skill types, check out this article: Choosing a MIRAI Skill Type
Positioning skills:
Use positioning skills for accurate real-time positioning of a tool where the path is a straight, direct motion from the starting point to the final target position.
Examples of use cases where you would use positioning skills are:
- Placing a gripper in a pre-grip pose over an object for picking
- Positioning for insertion tasks (direct path, no obstacles)
- Placing a sensor tool for quality inspection
Motion skills: Use for complex non-linear motion tasks, or where the direct path is not possible, e.g., moving around an obstacle, path following.
Examples of use cases where you would use motion skills are:
- Variance kinds of insertion tasks, for example, cable plugging
- Path or contour following tasks
- Picking or placing objects from and to a moving conveyor
- Quality inspection tasks (positioning of sensors or cameras)
- Finishing tasks like polishing or gluing
- Tool positioning where there are obstacles and the direct path from the starting point to the goal is not feasible
Step 4
Choose the number of cameras for your skill. For a dual camera setup, make sure that both cameras are connected to the USB port of the MIRAI controller. In case you change a camera and connect a new one, please make sure to restart the MIRAI controller so that the change in devices is correctly detected by MIRAI.
Skill configuration (Step 2/4)
Step 1
Choose the robot for this skill from the ‘Robot’ drop-down menu.
Step 2
Proceed to the ‘Device Configuration’ section. In case the devices are not auto-selected, assign devices for the camera(s) and F/T sensor. Choose from the drop-down options for ‘color_vision_1’ (cameras) and ‘force_torque_1’ and select the required devices. Ensure that all the devices are installed, correctly connected, and powered on before this step or they might not appear in the respective selection. In this case, go back and retry.
Note: In case you change a camera and connect a new one, please make sure to restart the MIRAI controller so the change in devices is correctly detected by MIRAI.
Step 3
Please specify how the camera(s) is (are) mounted. Select the ‘static-mounted’ option only if your camera is not mounted on the robot wrist but is placed independently on the working space. (This is only applicable for motion skills.)
Note: Selecting static-mounted will have significant implications on how rotations behave and should only be used if necessary. This selection would imply that the rotations and actions defined in the next section would be with reference to the fixed base frame (coordinate system) of the robot and not with reference to the relative action frame of the TCP, which is the standard case for wrist mounted setups.
Step 4
You can train robot movements on all the three translational axes i.e., movement along the x, y or z-axis. De-selecting one or more axes will prevent the robot to move along that axis. Please do this only when the range of the robot’s movement needs to be restricted.
It is not possible to disable all the 3 axes because this would cause the robot to remain stationary. (This feature is disabled for positioning skills.)
Step 5
You can also train robot movements using rotations (Note: For motion skills only one rotational axis can be enabled; for positioning skills you can enable all three axes). This will allow you to move and rotate the robot arm with reference to the action frame that was specified in the step before (i.e., the dynamic TCP action frame or the static robot base frame). It is highly recommended to enable all axes when it is really required, to limit the degrees of freedom and ensure the trained trajectory is stable.
When rotations are enabled, you are also required to specify the respective tool in use. From the Tool dropdown, select an existing tool, or tap ‘Go to tool configuration’ to set up a new tool.
Step 6
Certain tasks require to displace the center of rotations to the TCP or some other reference point. This could be because the tool (see image below) is offset on all 3 axes or attached at a different angle, meaning that it is not parallel to the z-axis of tool flange.
To enable this feature, select the checkbox ‘TCP displacement from center of tool flange’. Enter the x, y and z displacement values. Select ‘degrees’ or ‘radians’ and then enter the RX, RY and RZ angular values to specify the displacement of the rotation center point in reference to the center of the tool flange. You can also use the “TCP wizard” in the UR teach pendant to get these values and use then here.
Step 7
Press ‘Next’ to finalize these settings and move to the third step of the skill creation process.
When adjusting the gain and exposure value in this screen, make sure to also adjust the focus and aperture physically on the camera itself. Aperture ranges from 1.4 to 16; a value of 8 is recommended. Adjust the focus of the lens so that in the camera feed the tip of the end effector is in focus. Note that these settings are global and should not be changed for all the skills that you will create.
WARNING: Make sure to not change these manual settings, as this will break all the skills that were trained using these settings!
Camera settings (Step 3/4)
Step 1
Once the settings on the camera lens are done, adjust the camera exposure time and gain in the training app to minimize over/under exposure. Guide the robot through the trajectories that the robot will likely take to check and adjust the camera settings to values that offer a good lighting range and sufficient contrast. Tap ‘Next’.
Tap ‘Next’ to go to the last step.
Reference position (Step 4/4)
Step 1
You can set a safe and meaningful reference position for the tool center point by guiding the robot to the desired position and tapping ‘Save Position’. This is the position the robot can be moved back to when starting to record new episodes or test skills. Then tap on ‘Save Reference Position’.
Note: For skills that work in relative positions rather than in absolute ones, a reference position is not really needed but good to have for safety reasons.
Skill configuration is now complete. Proceed to the next steps by tapping ‘Next’ to start recording your training episodes.
As of version 8.0.0, it is possible to change the reference position as desired. Select the skill under the menu (three dots) then sub-item "Set reference position".
Comments
0 comments
Article is closed for comments.