Initial process-parameters setting of transfer moulding in microchip encapsulation: a case-based reasoning approach

Microchip encapsulation based on transfer moulding is one of the important processes of semiconductor manufacturing. This process involves numerous el ements working together in order to produce a quality moulded product. Amon g all of these elements, process-parameters setting of transfer moulding in microchip encapsulation plays a vital role. Determination of the initial p rocess-parameters for transfer moulding in microchip encapsulation is a hig hly skilled job, based on a skilled operators "know-how" and intuitive sens e acquired through long-term experience rather than on a theoretical and an alytical approach. Faced with global competition, the current trial-and-err or practice in the determination of the initial process-parameters becomes inadequate. In this paper, an artificial intelligence technique, case-based reasoning, is introduced in the process-parameters setting of transfer mou lding in microchip encapsulation, from which a case-based system has been d eveloped. The system aims to suggest the proper initial process-parameters of transfer moulding and improve its own moulding know-how through a learni ng process. The system was implemented and validated in collaboration with a semiconductor equipment manufacturer. From the validation results, it was found that the process-parameters obtained from the case-based system are very close to those obtained from the company. (C) 2001 Elsevier Science B. V. All rights reserved.