Design of a Database-Driven Quality Predictor for Painting Systems

Abstract

In recent years, painting robots (Fig.1) have been introduced in the painting field of excavator due to the labor shortage caused by the declining birthrate and aging population. The object is painted by a painting machine attached to the tip of a 6-axis industrial robot. The quality of the painting is evaluated by whether painting defects occur or not. There are multiple types of painting defects such as curtaining and orange peel. Since the rotational speed of the bell cup greatly affects the painting quality and it is important to suitably determine it just before painting. However, the appropriate rotational speed of the bell cup is highly dependent on the painting environment and operating conditions, and there is no recommended value for it. Therefore, it is currently set based on the operator’s knowhow. The rotational speed of bell cups, once set, is not changed frequently during the day’s painting process, although the painting environment may change significantly with changes in temperature and humidity. For this reason, painting defects often exist. To prevent the occurrence of painting defects, the rotational speed of the bell cup should be adjusted online based on the painting quality. However, the operator visually monitors the painting quality, and includes a considerable time-delay between the painting process and the evaluation of painting quality.