Use of Artificial Intelligence and big data analysis in control and diagnosis of complex systems

Artificial intelligence (AI) and Machine Learning (ML) have transformed many industries profoundly in the last several years. Surface coating systems and processes are complex systems difficult to be controlled and diagnosed using traditional techniques. A coating process consists of all aspects of customer requirements, recipe development, coating test and validation. Thermal spray (TS) is one main surface coating technology that provides several benefits to the coated parts such as improve thermal protection, high hardness, wear resistance, and increased life-time. However, TS is a complex process with several thermodynamics, fluid mechanics, electromagnetics phenomena happen simultaneously. Also there are several tens of parameters influencing coating, some of which are not even measured.

In this talk, the use of AI for three aspects of surface coating system is presented:
(1) Coating recipe development: Neural network is used to model the coating process. The model shows acceptable accuracy on predicting coating quality with test recipes
(2) Adaptive coating quality control. In order to control any system, a model of the system is needed. It is difficult to gain overall insight of the whole process and dependencies between different inputs and outputs using mathematical and CFD analysis, due to very complex and interconnected nature of the thermal spray process. It is shown that the AI model can be used successfully to control the coating quality in an adaptive manner.
(3) Coating machine diagnosis: AI data analysis is performed on the collected machine data to achieve the following goals: Alert creation, Anomaly prediction, Hardware life-time prediction