A dynamic learning model for on-line qualitycontrol usingthe taguchi approach

An important philosophy of quality control due to Taguchi rests on the belief that, once quality is designed into the product and the process, very little inspection is necessary.Taguchi evaluates quality in terms of costs incurred, including inspection and adjustmentcosts apart from quality loss. Using these costs, he determines an optimal interval for periodic inspection of products. At each inspection, the first item on the process line is checkedfor defects. If no defects are found, the process is assumed to be under control. On the otherhand, if the first item turns out to be defective, the process is immediately stopped andadjusted. The underlying assumption here is that, once a defect is observed, it is construedthat the process has drifted and all subsequent items are defectiveThis article challenges the underlying assumption by recognizing that the observation of one defective item does not necessarily imply that the process is out of control. Noisy signalsmay occur as a result of nonprocess disturbances such as measurement error or operatorerror. The result of the relaxation of Taguchi 's assumption is that the decision of whether tostop the process would have to be made dynamically. The decision model developed in thispaper is based on the concept of state-space search and incorporates Bayesian andregression-learning mechanisms for parameter estimation. Subsequently, a framework for anexpert controller for automating the system is developed. In this framework, knowledge isidentified in the form of three components: assessment knowledge, simulation knowledge, andprocess knowledge. The problem studied has wide applications in the implementation of theTaguchi approach to quality control in manufacturing environments. Possible extensions anddirections for-future research are also presented.