Application of Artificial Intelligence for Maintenance Modelling of Critical Machines in Solid Tire Manufacturing

Machine maintenance is a challenging task in the manufacturing industry. The reliability and maintenance scheduling of continuously running machines are essential for the performance of manufacturing plants. In this paper, an artificial intelligence-based machine maintenance management system is proposed for a tire manufacturing plant. The proposed system consists of two main subsystems: dynamically updating maintenance scheduler and machine troubleshooter. The maintenance schedular is implemented using an Artificial Neural Network (ANN) whereas the machine troubleshooter is based on an expert system. The ANN-based maintenance schedular provides the optimum time frame to plan the preventive maintenance of critical machines based on the condition monitoring data and production data. The ANN is validated using validation performance charts and regression state charts obtained from the Matlab runtime environment. It is found that the R-squared value of the ANN is 0.998. On the other hand, a rule-based inference system is used in the machine troubleshooter. The expert system is validated by evaluating the maturity of the knowledge base. The percentage maturity of the expert system is reached to a level of 90% within 3 months.