Design and Vision-Based Calibration of a Five-Axis Precision Dispensing Machine

Area of Science:

• Mechanical Engineering
• Robotics
• Metrology

Background:

• Five-axis precision dispensing machines are crucial for semiconductor packaging.
• Geometric errors in these machines significantly impact dispensing accuracy.
• Existing calibration methods can be complex and time-consuming.

Purpose of the Study:

• To propose a vision-based measurement (VBM) system for identifying and calibrating geometric errors in five-axis dispensing machines.
• To develop an automated solution for kinematic calibration and workpiece detection.
• To enhance the overall accuracy and reliability of semiconductor packaging processes.

Main Methods:

• Established a kinematic model using screw theory's product-of-exponential formulation.
• Designed a geometric error identification algorithm incorporating position-independent geometric errors (PIGEs) and position-dependent geometric errors (PDGEs).
• Employed singular value decomposition and regularization to solve overdetermined equations and identify PIGEs in the A and C axes.

Main Results:

• The VBM system successfully identified eight PIGEs in the A and C axes.
• Measured mean spatial position error of 59.9 μm and mean orientation error of 160 arcsec for the end-effector.
• Verified the proposed approach through comprehensive error measurements.

Conclusions:

• The VBM system offers a feasible and automated calibration solution for five-axis precision dispensing machines.
• The developed method effectively quantifies geometric errors, crucial for high-precision applications.
• This approach contributes to improved accuracy in semiconductor packaging.