4 Proc. of the Fifth International Conference on Advances in Civil, Structural and Environmental Engineering - ACSEE 2017. Copyright © Institute of Research Engineers and Doctors. All rights reserved. ISBN: 978-1-63248-122-1 doi: 10.15224/ 978-1-63248-122-1-08 Application of Multivariate EWMA for performance monitoring of RMC Keval Dabasia 1 and Debasis Sarkar 2 Abstract— Statistical process control is applied in order to monitor and control a process. Statistical process control procedures can be widely applied to understand and improve the production process of construction materials. In performance monitoring of Ready Mix Concrete plant, there are always more than one quality characteristics which affect the performance of concrete and are usually correlated. In this paper, Multivariate exponentially weighted moving average control chart (MEWMA) is applied to monitor the performance of concrete in Ready Mix Concrete plant. The different parameters considered are strength and density at different age of concrete. One difficulty found with multivariate control charts is the interpretation of out-of-control points. The univariate control charts are used to obtain a rough estimate of the sources of Multivariate out-of-control points. EWMA and MEWMA methods are compared for performance monitoring of concrete. The result shows that MEWMA charts are more relevant than EWMA charts. Keywords— Statistical Process Control, Multivariate EWMA, Univariate EWMA, Ready mix Concrete I. Introduction Ready-mixed concrete (RMC) was patented in Germany in 1903, but it was only after the development of the modern truck mixer in the 1950s that the supply of RMC became commercially viable [1]. There is tremendous growth in the use of ready mixed concrete (RMC) for construction in the developing countries. The RMC industry is not regulated or monitored properly, as a result it has contributed to a general disregard for the basics of good quality of RMC [2]. To rectify the situation and to improve the performance of RMC producers, systematic monitoring and inspection is required. Keval Dabasia 1 Pandit Deendayal Petroleum University India Dr. Debasis Sarkar 2 Head of Department Civil Engineering Department Pandit Deendayal Petroleum University India The SPC tools are very useful in identifying the defects in the process and developing the solutions for eliminating them to improve the quality of the products. The standardization of the system will definitely improve the quality of the products manufactured as in this case it is concrete. Statistical process control consists of a number of powerful tools for problem solving and improvement of quality control by reducing variability in manufacturing processes. At present, quality control charts for monitoring a continuous quality characteristic is done using Shewhart, Cumulative Sums (CUSUM), exponentially weighted moving average (EWMA), etc. In some cases it may also be necessary to control simultaneously two or more related quality characteristics, which can be done with the multivariate SPC using multivariate charts such as Hotelling’s T 2 control charts, multivariate EWMA (MEWMA), etc. [3] Most of the models developed were dealing primarily with only one quality parameter or characteristic which is mainly strength of the grade of concrete monitored. But higher degree of quality monitoring can be carried out if multiple quality characteristics can be monitored simultaneously. Thereby, Sarkar and Bhattacharjee (2014) made an attempt to develop a multivariate CUSUM methodology (MVCUSUM) where the other parameters like slump, density and temperature of fresh concrete which affects the quality of RMC can be monitored. Later Sarkar and Panchal (2017) did an attempt to explore the potentiality of application of Quality Function Development and particularly House of Quality which primarily focuses on customer satisfaction, for performance monitoring of RMC. To improve efficiency in the processes with small changes, univariate EWMA and multivariate EWMA (MEWMA) control charts were developed. Their advantage is that they take into account the present and the past information of the process. Therefore they are more powerful to detect small changes than Hotelling’s T 2 control charts [4]. These control charts be used as daily or weekly monitoring tool, which can improve the efficiency of the production process [5]. Thus this paper is an attempt to explore the MEWMA control chart as a tool for the process monitoring of more than one variable which can affect the quality of the RMC. II. Monitoring Multivariate Processes using Control Chart The multivariate charts can be separated into two categories, i.e., the directionally invariant schemes and the direction specific schemes. Typically, most symmetric two- sided univariate control charts are directionally invariant, whereas multiple univariate schemes used for a multivariate