2022 Volume 71 Issue 6 Pages 180-182
In case of RBM（Risk Based Maintenace） assets risk evaluation, metal corrosion mechanism has been decided by the experts. But, the perfect estimation of corrosion mechanism to too many corrosion environments is difficult and the population of such experts will decrease in near future. As the countermeasure, AI system is proposed here and continues to reach more reliable RBM. This AI system has 2 Python program codes. One is RBS（Rule Based System）. The other is Decision Tree Analysis. RBS has 172 metal damage mechanisms classified to fatigue, Creep, Wet/Dry Corrosion, Metal degradation and others. RBS and Decision Tree Analysis can predict effectively each asset corrosion mechanism using small amount of data such as kind of factory, kind of equipment, material, chemical environment, and precisely predict the risk for RBM using a lot of data such as chemical concentration. The issue to resolve is that there are too many natural language characters in data reducing the reliability of estimation of corrosion mechanism.
