As industry progresses toward intelligent production and development, on-site workers that perform repetitive tasks will be replaced by intelligent machines. Currently, automation applications still have the following problems: (1) on-site personnel are required to line up the workpieces before a robot arm can pick it up; (2) the trajectory generated by offline programming software must be adjusted by on-site personnel in accordance with the processing results; and (3) because of workpiece positioning errors and tool wear, achieving acceptable processing results is difficult. This study developed intelligent application modules that solve the aforementioned automation application problems. These modules predict processing quality, generate trajectory, enable robot arms to load and unload randomly positioned workpieces, and automatically calibrate the system. An automatic gear edge grinding system was developed by integrating each module; the system increases the processing efficiency and solves the current problem of manual grinding being required after gear processing.

No abstract