Data preparation and analysis for modelling of gas consumption of production of canned meat

Abstract

We have been looking for the possibility to build a computer model of costly gas utilization in the course of heat treatment of canned meat. The aim of our long-term research is to set up a model to be able to accomplish minimized costs by synchronizing heat treatment processes running side by side and by production programming to define completion time of the cans. First, because of our multiple data (paper based, text file, image file, Excel file), we have arranged them under a common user interface, while defining the best one from the data with different precision appearing in several sources. We could define the set-out time most difficultly where the data from the image file name, the evaluation of the image content and the data of the heat treatment specifications were also needed. Based on the accurate data we did correlation and regression analysis to define the data influencing the gas consumption and to see whether our present data was enough to develop a model. Since the furnace has to ensure the steam supply of the heat treatment equipment and the heating of the buildings too, we defined the gas consumption of the furnace which is dependent on the environmental temperature. According to our examinations it is not practical to reduce the costs by keeping the quantity of tied-up gas on a minimum level because by the use over this limit we can lose the multiple cost margin won by the minimizing. However we need the product programming and the heat treatment synchronization to avoid the penalty kind costs of stepping over the tied-up level of gas consumption.