Predicting motor oil condition using artificial neural networks and principal component analysis

Condition monitoring of engines’ oil is a strategic area in the maintenance management field. Replacing the oil too early represents unnecessary unavailability, as well a financial and environmental costs which could be spared. Replacing it too late can impair the oil’s ability to protect the engine, therefore increasing the chances of damage and premature ageing of the engine, or even the risk of causing accidents which can endanger people, equipments or vehicles in urban environments. The use of modern tools from data mining and Artificial Intelligence (AI) can contribute to help make the right decision at the right time, thus protecting the environment, the companies’ profits and the safety of people and property. The present paper discusses a methodology to create models to facilitate the process of oil analysis, tested with a dataset for oil of Diesel engines, from urban passenger buses. Preliminary work was already done [14], using Artificial Neural Networks (ANN). In the present research, the neural models were improved and the results are compared with analysis using multivariate systems, namely Principal Component Analysis (PCA). PCA showed the relevance of each variable is different, and some of the variables may even have a negative impact on the predictive power of the ANN. Data used for the experiments come from two passenger bus companies. Each company provided a dataset, containing results of laboratory analysis of the oils and their classification, according to human experts of a specialized oil analysis company. Data were mined and neural models were created, for both datasets separated and combined. The remainder of the paper is organized as follows. Section 2 presents a summary of the state of the art. Section 3 describes the datasets used. Section 4 describes the neural networks. Section 5 describes the analysis performed using multivariate systems. Section 7 presents a comparative and critical analysis of the results obtained. Section 8 highlights the main contributions of the present research. Section 9 presents some conclusions and outlines future work.