Sputter coating is an advanced manufacturing process that uses a vacuum chamber to deposit thin metallic and oxide layers on glass surfaces for applications such as architectural and anti-reflective glass, solar panels, and automotive coatings.

The process is equipped with hundreds of sensors to monitor every aspect of the environment and the machines. Managing the sputter coating process requires precise control of numerous variables, including pump settings, conveyor speeds, and other operational parameters.

However, the sheer number of controllable factors and the complexity of the process often results in a high rate of defects, such as scratches or chemical weaknesses in the coatings, requiring enhanced process control strategies.

First, the goal must be set to target the scrap rate reduction. This is easy to achieve if the outcomes of the different parts (defect or not-defect) are systematically documented in the data along with all other process parameters.

Next, the raw historical process data of the targeted sputter coating line is fed into the AI engine that powers Process Booster, aivis®. The AI performs a root cause analysis based on the predefined objective to identify the key factors and their characteristic behavior influencing the defect rate. Such insights enable process engineers to further develop a comprehensive understanding of the underlying issues.

The AI also creates a process control model live-predicting the defect risk and recommending appropriate adjustments of controllable parameters to prevent defects.