Introduction to Vision-Guided Robotics with AOIs

Pixel-to-World Coordinate Translation

The Cognex In-Sight camera handles spatial processing onboard using its internal Trans Pattern to World function:

1. The camera captures an image of the target object (e.g., specific die faces in a dice-sorting cell).

2. It detects the target's pixel offset and converts those raw pixel numbers into real-world physical coordinates.

3. These world coordinates output directly as $X$, $Y$, and $R$ (rotation) position values aligned with the robot's Tool Center Point (TCP).

Connecting Cognex AOI to Yamaha Robot AOI

To execute vision guidance inside Studio 5000:

1. Rung 1 (Vision Capture): The Cognex AOI triggers the camera job file, parses the inspection data, and extracts the calculated $X$, $Y$, and $R$ world position values.

2. Rung 2 (Robot Motion): A Synchronous Copy or direct tag assignment feeds those $X$, $Y$, and $R$ variables straight into the Yamaha MOVE_MM instruction block.

3. The robot instantly moves to the exact spatial coordinate of the identified part without requiring external middleware or complex string manipulation.

Application Logic: Sorting & Re-rolling Workflow

In the live demonstration cell, the PLC manages an automated sequence searching for specific dice numbers (1 through 6):

• Part Inspection: The camera scans the tray for the targeted die face.

• Re-roll Trigger: If the required number is not found, the logic executes a physical re-roll mechanism to reshuffle the dice.

• Unload Sequence: Once a matching number (e.g., 1, 2, or 3) is identified and its coordinates are passed to the Yamaha controller, the robot executes the pick and unloads the part.