| |
Automatic surface inspection of painted plastic parts |
| |
| Starting point / Task definition |
| |
The challenge of the research and development task entrusted by a car manufacturer to Rosenheim Polytechnical Institute consisted in the complexity and variety of the visual factors to be examined in the surface inspection of painted plastic parts, which was to be implemented in a fully automatic process. It was thus required for a camera to be able to recognize on a reference component streaking, sink marks, scratching, differences in luster, deformations produced on removal from the mold, the mark of the weld line, torching, over-injection, etc., all with the same quality as from a manual visual check. In addition, there was also a requirement for saving the fault data, making a statistical evaluation and for incorporation into a QA system / PDA system.
Furthermore, the customer set high demands on the system itself. Even within the constraints of the research assignment it was necessary for the criteria applied to industrial series production to be fulfilled. This made the task more comprehensive, adding such factors as robustness, suitability for series production with regard to hardware and software, the ability of the system to handle several, possibly differing, components, ease of use for the operator and installer, little programming work in the event of conversion, and compatibility with peripheral devices.
In order to guarantee optimal output from the machine, it was necessary for the component checking to be carried out in the non-productive times of the injection molding machine.
Closely related to this project was a similar application for the Fakuma 2002 trade fair, in which the quality management process for the surface inspection is visualized and demonstrated in a practical example. |
|
| |
| Implementation / Solution |
| |
 | | Automatic surface inspection of painted plastic parts |  |
The robot then turns the part on its sixth axis so that all surfaces can be inspected by the image processing system. The image processing system is able to pick out differences or inconsistencies in color with the aid of previously defined limits set in its measuring window, and thus evaluates whether the part has passed the quality check. The results of the surface inspection are also transferred in parallel to a monitor outside the cell.
As well as the surface-only examination, the image processing system can also carry out a completeness check. As part of this test, the system checks for the conditions expected at the previously defined reference points, such as contours or openings. If a part should deviate at these points from the tolerance values of the guide measurements, it is rejected as faulty.
The interaction between the controllers of the robot and the measuring system takes place via handshaking. The robot conveys the part into the first measuring position and waits for the “OK” signal of the measurement system. After the inspection of the part has been passed, the robot places the part in position two, etc. In the event that a part fails the inspection, the robot interrupts the process and automatically discards the part. |
|
| |
| System components / Scope of supply |
| |
| All the components of the system were included in the scope of the supply. Selection of the imaging systems, the light source, the programming of the measuring processes and the robot controller including handshaking between the machine, the jointed-arm robot and the measuring system. |
|
| |
| Results / Success |
| |
The original demonstration system was successfully implemented into operations. The demand of the industry for an automatic optical inspection process for painted and unpainted plastic components to minimize the defect quota was realized in accordance with the requirements of production operations and within the constraints of college research.
This example shows the process capabilities of innovative technologies, which are contributing to “zero-defect production”. It can also be seen as an example of the functional effectiveness of the ever closer cooperation between producers of machines, automation firms, end customers and research and development institutions. |
|
| |
|
| |
|
|
| Industry |
| |
Automotive manufacturers
Rubber, plastics
Research, education |
|
| |
| Application |
| |
Handling, loading and unloading
Measuring, testing and inspection
Other handling operations |
|
| |
| Customer |
| |
| University of Applied Sciences, Rosenheim, Germany |
|
| |
|
|