Managing yield by lot splitting in a serial production line with learning, rework and scrap

Abstract

Wright's (1936) learning curve (WLC) depicts that the time required to accomplish a repetitive operation decreases with each subsequent repetition. The WLC model assumes that the components produced in an operation are all of perfect quality. On the other hand, in many production operations, some of the components require rework. Some components are even scrapped if they cannot be reworked. We employ the WLC model by considering the learning process in the time to produce and the time to rework a given lot. This composite learning curve was developed by Jaber and Guiffrida (2004). The impact of splitting a production lot into batches of equal size is studied through this composite learning curve. The objective of the study is to maximize a combination of performance of average processing time and process yield with respect to the number of batches. The effect of varying the learning curve parameters in production and in rework is studied for the developed model.