Robot Pick Calibration#

This page explains how to connect vision coordinates with robot coordinates to allow accurate component picking.

What is Robot Pick?
The Robot Pick function calculates the offset between the coordinates detected by FlexiVision One and the real robot coordinates, allowing the robot to pick components in the correct position.

Danger

Robot coordinates are mandatory

This phase REQUIRES the X, Y, and Rz coordinates saved during the physical setup preparation, Step 1 of the Clearances section.

Without these coordinates, calibration cannot be completed. If they were lost or forgotten, the entire physical robot preparation must be repeated.


Robot Pick interface overview#

After clicking Next on the Clearances page, the Robot Model Pick page opens.

Robot Pick page

Section

Parameter

Function

Enable

Enable Robot Pick

Activates robot calibration

Vision Result

X cord

X coordinate detected by vision

Vision Result

Y cord

Y coordinate detected by vision

Vision Result

RZ cord

Z rotation detected by vision

Insert Robot Coordinate

X cord

Robot X coordinate to be entered manually

Insert Robot Coordinate

Y cord

Robot Y coordinate to be entered manually

Insert Robot Coordinate

RZ cord

Robot Z rotation to be entered manually

Function

Description

Find Object

Detects the component and displays vision coordinates

Picking Offset

Calculates the offset required for correct picking


Step 1: Activate and detect the component#

1

Click Enable Robot Pick

2

Click ../../../_images/tasto_FIND_OBJECT1.png:

  • The system detects the reference component

  • The coordinates appear in the Vision Result section

Note

Vision Result These are the coordinates FlexiVision One “sees” in the image. They are not yet linked to the robot coordinate system.

Step 2: Enter robot coordinates and calculate the offset#

3

In the Insert Robot Coordinates box, enter the coordinates saved during model creation:

Danger

Use the coordinates saved during model setup. Without these coordinates, calibration will be wrong.
Coordinates must be entered with maximum precision:

  • Copy the values exactly as written, including decimals

  • DO NOT round them, for example 450.23 is not the same as 450.2 or 450

  • Verify that X and Y have not been swapped

  • Check the sign of each coordinate, positive or negative

Errors at this stage cause completely incorrect robot offsets, resulting in pick attempts in the wrong positions, even with errors of several centimeters. Failing to respect these points may also cause robot collisions and damage to the FlexiBowl, components, or the robot itself.

4

Click ../../../_images/tasto_GRIPPER_OFFSET.png

  • The system automatically calculates the transformation between vision coordinates and robot coordinates

  • This offset will be applied to all future detections


How the Gripper Offset works

The system compares:

  • Vision Coordinates: where FlexiVision One sees the component origin

  • Robot Coordinates: where the robot actually grasped the component

It calculates the difference and stores it as an offset. This offset is then applied to all components detected in the future, ensuring that the robot always picks in the correct position.


Step 3: Finalization and saving#

5

By clicking ../../../_images/tasto_next.png, the workflow returns to the recipes page ../../../_images/tasto_recipes.png

6

Click ../../../_images/tasto_save_recipes.png to save the full configuration

Complete save

The save includes:

  • ✓ created model

  • ✓ working area, ROI

  • ✓ tolerances, Accept Threshold

  • ✓ configured Clearances

  • ✓ robot calibration, Gripper Offset


Multiple models - add additional models#

Step 4: Additional models, optional#

7

To create additional models in the same recipe:

  • Return to ../../../_images/tasto_edit_recipes.png

  • Select a new model slot that has not yet been configured

  • Repeat the entire procedure starting from Model Creation

Tip

Each model inside the recipe can have different settings, such as ROI, Clearance, and offset, allowing management of components with different characteristics within the same application.


Final verification#

Before considering the recipe complete, continue with:

See also


Back To Top