A Bayesian approach to automated optical inspection for solder jet ball joint defects in the head gimbal assembly process | |
| Author | Wai, Mak Chee |
| Call Number | AIT Thesis no.ISE-13-31 |
| Subject(s) | Bayesian statistical decision theory Assembly-line methods |
| Note | A thesis submitted in partial fulfillment of the requirements for the Degree of Master of Science in Microelectronics and Embedded Systems, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Thesis ; no. ISE-13-31 |
| Abstract | The storage industry continues to grow at a 40% compounded annual growth rate (CAGR) due to the continued growth in digital content consumption and creation.Automation or selective automation is adopted as a solution to most productivity problems in the manufacturing of hard disk drives (HDD). As part of the automation solution, an automated production line for manufacturing Head-Gimbal-Assembly (HGA)has been developed. In the automated HGA production line, a solder jet ball (SJB) solderingstation connects the suspension circuit to the slider body. Although a vision system is installed at this station, the capability to inspect the quality of the solder joint is lacking.This thesis proposes a (constrained)Bayesian approach to automated optical inspection(AOI)of SJB jointin the HGA process. Prominent characteristicsextracted from a captured image (by the imaging system)will be used to establish a model that accurately detectsdefectsin the SJB joint. This model classifies the quality of the SJB joint into “good”, “bridge” and “burnt”. The elicited probabilistic model is implemented together with image processing software and used to inspect the quality of SJB joints in-situ.The algorithm is further enhanced with a result check to improve accuracy. Thefinal attained accuracy for classification (at parts level) is 91.52%using TAN-BN with 3-parameter enhance check. From the enhance check routine, the number of mismatched decisions is monitored, which in turn becomes a triggering platform for model re-learningusing statistical P-Chart method. The implementation is then extended to a learning framework, enabling machine learning in-situ. |
| Year | 2013 |
| Corresponding Series Added Entry | Asian Institute of Technology. Thesis ; no. ISE-13-31 |
| Type | Thesis |
| School | School of Engineering and Technology (SET) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Afzulpurkar, Nitin V. |
| Examination Committee(s) | Dailey, Mathew N.;Bernard, Philip |
| Scholarship Donor(s) | Western Digital;NECTEC;Asian Institute of Technology Fellowship |
| Degree | Thesis (M. Eng.) - Asian Institute of Technology, 2013 |