APPLICATION OF TIME SERIES CONTROL CHARTS IN CHIP WOOD PRODUCTION: A CASE OF HAWASSA CHIP WOOD FACTORY

Abstract

Application of time series control charts in chip wood production, a product of Hawassa Chip Wood Factory, is investigated in this paper. Currently, the factory is implementing the ISO 16895:2016 specifications of the International Standards Organization and the company’s quality standard for the quality variables of its product. The main aim to conduct this study was to overcome the problem which faces the quality of products during manufacturing process. The data on the quality variables were collected from the quality control department of the factory. The main quality variables in the manufacturing of chip wood product are the amount of moisture of chips in before blender core, before blender surface, after blender core, after blender surface, and material distribution test by using density of chip wood. Descriptive statistics were used to check whether the data follow normal distribution, the data for all the quality variables are stationary, Box-Jenkins methodology was performed in order to fit the model. For the variables before blender core and before blender surface ARIMA (2, 0, 1) was the best model identified and for after blender core, after blender surface and material distribution test (density) ARIMA (1, 0, 1) was the best model identified. From ARIMACUSUM and ARIMAEWMA charts, all the observations were between LCL and UCL, which indicates that the process is in control. But there was a shift which started decreasing from one observation to the other and also it started increasing from one observation to the other for the quality variables except in material distribution test of density, in ARIMAEWMA chart the 1st observation is out of control. There were also other minor shifts in the process. These shifts in the process are minor which are due to random causes which can’t be eliminated. Further, no assignable causes are present and hence the process is in statistical control.