Affiliations: Department of Statistics, Pondicherry University, S.V. Nagar, Kalapet, Puducherry – 605 014, India. E-mail: [email protected] | GfK Custom Research North America, 8401 Golden Valley Rd, Minneapolis, MN, USA
Abstract: The modern manufacturing technology requires the effective use of statistical tools for analyzing quality problems and controlling, monitoring, maintaining and improving the performance of manufacturing process. When dealing with measurable quality characteristics, it is necessary to monitor the behaviour of the process average of the quality characteristics and its variability. Control charts are widely used to take a proper and a meaningful decision such as: How long the present process can be allowed for production? When do we interfere into the process for correction / modification / rectification? It is important to take a proper decision so as to improve the quality and its productivity. The present paper deals with the use of diagonal systematic sampling in industrial applications for on line process control. The relative performance of diagonal systematic sampling with that of systematic sampling and the simple random sampling is assessed for a certain natural population obtained from an automobile ancillary industry with the presence of linear trend among the observed values due to the tool wear. The efficiency of the different sampling schemes is measured for estimating the mean of quality characteristics; for controlling and monitoring the manufacturing process and also for the effective computation of process capability indices. Consequently it has been recommended to use the diagonal systematic sampling for the selection of random samples to assess the quality characteristics of the components produced; to take a proper and meaningful decision whether the present manufacturing process can be allowed further or not in a short production run of engineering industries in the presence of systematic assignable causes.
Keywords: Diagonal systematic sampling, systematic sampling, simple random sampling, linear trend, industrial applications, control charts, process capability analysis and quality characteristics
