ENBIS-18 in Nancy

2 – 25 September 2018; Ecoles des Mines, Nancy (France) Abstract submission: 20 December 2017 – 4 June 2018

An Analysis for Industry 4.0: Change Point Detection of the Effects of Industrial Revolutions on Performance Indicators

4 September 2018, 11:40 – 12:00


Submitted by
Ozgun Sarikulak
Ozgun Sarikulak (ASELSAN A.Ş.), Murat Caner Testik (Hacettepe University)
Industrial revolutions have huge impacts on human history when those revolutions compare with other historic moments. Alterations of energy regime, improvements of production processes and changes in the ways of communication and transportation affect the social and economic life.

In this study, performance indicators of England, France, Netherlands, Germany and USA are studied retrospectively by using cumulative sum (CUSUM) control charts to estimate the change points. Hence, reflections of industrial revolutions on the performance indicators are evaluated. The data used in the analysis include income per capita and energy consumption per capita from the 18th to 21st century. Autoregressive Integrated Moving Average models are used to represent autocorrelation structures of the time series of indicators. CUSUM control charts are implemented for the residuals of the models, where non-normal residuals are transformed to normal distribution. The last reset time corresponding to a CUSUM signal is taken as the estimate of a change in the performance indicator and this is analyzed for the effects of a war or an economic crisis besides an industrial revolution. If an effect of war or a crisis is found, corresponding out-of-control observations are omitted from the dataset and missing point imputation is performed.

According to the study results, reflections of the industrial revolutions’ effects can be observed from performance indicators. This study is also consistent with the years that are accepted in the literature for industrial revolutions. In addition, this study shows that the era of industrial revolutions also altered the patterns of income per capita and energy consumption per capita.

Return to programme