In manufacturing processes where material is processed through a series of stages and at each stage there are many parallel equipment units that can be used, a major difficulty is assessing whether or not any of the large number of equipment units is defective and is causing a problem with the final product. A procedure is introduced in this paper to analyze historical data from such processes to address this question. The method is based on the use of indicator variables and latent variable models. The assessment of the performance of the equipment units is easily done using simple plots available from the resulting latent variable model. The methodology is illustrated through its very successful application to a complex hollow fiber membrane manufacturing process.