Developing of Automatic Gasket Gluing Robot for a Tractor Company

Humans are still the primary laborer to apply gasket glue on the brake cases and inspect the tractor's assembly quality in the tractor factory, Siam Kubota Corporation Company Limited. Because working with manual labor still has human errors, the researchers desire to improve the process by using image processing technology and automation to build a machine vision to control the gluing quality on the brake cases and the assembly inspection. For the automatic glue extrusion, a cartesian robot was chosen. The image processing technology was also used to indicate an initial point to start applying glue, and G-code was used to control the robot to move along the designed path on the brake case. Then the glue extrusion robot was tested with two experiments: applying the glue on the sample, which has a surface like the real brake case called Sample 1, to evaluate the performance of the image processing and using the glue on the printed image of the brake case called Sample 2 to determine the performance of the robot in terms of accuracy and precision. This has resulted in the deviation of the gluing distance error process at a 95% confidence interval got; the mean dimension on the actual surface Sample 1 is 0.73±0.04 mm width and 0.72±0.27 height, and the performance of the robot within the exact surface dimension Sample 2 has a mean distance of 1.04±0.1 width and 1.06±0.20 height.