VISUAL METAL PANEL QUALITY DETECTION BASED ON CUTTING EDGE

Abstract

The invention is directed at a method for determining the quality of a foamed unit (1), wherein the foamed unit (1) is produced by forming solidified foam (2), wherein a foamed unit edge (5) is formed by cutting through the foamed unit (1), wherein a camera (7) captures an image (8) of the foamed unit edge (5), wherein the image (8) is analyzed to detect defects (9) in the foamed unit edge (5) and wherein quality information data (11) describing the detected defects (9) is generated based on the analysis of the image (8). The method is characterized in that the foamed unit (1) is produced by a production apparatus based on production parameters (17), which production parameters (17) comprise production variables (18) measured by production instruments (20) during the production of the foamed unit (1) and/or production settings (19) input to the production apparatus, wherein based on the quality information data (11) updated production settings (21) are generated, preferably, that the updated production settings (21) are input to the production apparatus, wherein the updated production settings (21) are generated by applying the quality information data (11) and the production parameters (17) to a calculation model (22), which calculation model (22) provides a computational relationship between the production parameters (17) and the quality information data (11), preferably, that a computer system (10) generates the updated production settings (21) by applying the quality information data (11) and the production parameters (17) to the calculation model (22), in particular, that the calculation model (22) is saved on the computer system (10).The invention is also directed at a corresponding system for determining the quality of a foamed unit (1).

Claims

1-13. (canceled)

14. Method for determining the quality of a foamed unit, wherein the foamed unit is produced by forming solidified foam, wherein a foamed unit edge is formed by cutting through the foamed unit, wherein a camera captures an image of the foamed unit edge, wherein the image is analyzed to detect defects in the foamed unit edge and wherein quality information data describing the detected defects is generated based on the analysis of the image, wherein, the foamed unit is produced by a production apparatus based on production parameters, which production parameters comprise production variables measured by production instruments during the production of the foamed unit and/or production settings input to the production apparatus, wherein based on the quality information data updated production settings are generated, wherein the updated production settings are generated by applying the quality information data and the production parameters to a calculation model, which calculation model provides a computational relationship between the production parameters and the quality information data.

15. Method according to claim 14, wherein the foamed unit is produced by forming solidified foam on at least one solid sheet.

16. Method according to claim 15, wherein the foamed unit is produced by feeding the at least one sheet in a substantially continuous feed and that the foamed unit is separated from the feed by cutting through the feed.

17. Method according to claim 14, wherein the solidified foam comprises polyurethane.

18. Method according to claim 14, wherein the solidified foam is formed by mixing, in particular injecting, materials configured to react and create the solidified foam.

19. Method according to claim 14, wherein the detected defects are classified into defect categories, wherein the defect categories comprise bubbles within the foam, cracks within the foam, voids within the foam, or overrolling marks within the foam.

20. Method according to claim 14, wherein the quality information data is compared to predefined alarm criteria and that if the quality information data meets the pre-defined alarm criteria an alarm signal is generated.

21. Method according to claim 14, wherein generating the updated production settings comprises comparing the quality information data with a pre-defined rule set for generating the updated production settings.

22. Method according to claim 21, wherein the production settings comprise a preheat temperature of a production line, a pressure during forming the solidified foam and/or moving speed of the production line.

23. Method according to claim 21, wherein the production variables comprise a temperature of the solidified foam and/or an ambient humidity when forming the solidified foam.

24. Method according to claim 14, wherein a further foamed unit is produced based on the updated panel production settings, that the camera captures a further image of a foamed unit edge of the further foamed unit, that the further image is analyzed to detect defects in the foamed unit edge of the further foamed unit, that further quality information data describing the defects of the foamed unit edge of the further foamed unit is generated and that the calculation model is updated based on a comparison of the further quality information data and a predicted quality information generated by applying the updated production settings to the calculation model.

25. Method according to claim 14, wherein the quality information data is generated according to an analysis algorithm, that a thermal insulation of the foamed unit is measured to obtain a thermal insulation measurement result and that based on a comparison between the thermal insulation measurement and the quality information data the analysis algorithm is updated.

26. System for determining the quality of a foamed unit with a production apparatus for producing the foamed unit by forming solidified foam and for forming a foamed unit edge by cutting through the foamed unit, with a camera for capturing an image of the foamed unit edge and with a computer system for analyzing the image in order to detect defects in the foamed unit edge and for generating quality information data describing the detected defects based on the analysis of the image, wherein the system is configured such that the foamed unit is produced by the production apparatus based on production parameters, which production parameters comprise production variables measured by production instruments during the production of the foamed unit and/or production settings input to the production apparatus, wherein based on the quality information data updated production settings are generated.

Description

[0036] Further advantages and preferred features are discussed in the following description with respect to the Figures. In the following it is shown in

[0037] FIG. 1 a schematic view of a first embodiment of a system according to the invention for carrying out the method according to the invention and

[0038] FIG. 2 a schematic view of a second embodiment of a system according to the invention for carrying out the method according to the invention.

[0039] The foamed unit 1 shown in FIG. 1 is a metal panel. This metal panel is produced by forming solidified foam 2 between solid sheets 3a, b, which are here both made from metal, arranged opposed to each other. The solidified foam 2 substantially consists of polyurethane, which was formed by mixing an isocyanate, a polyol, a blowing agent, activators and catalysts. This mixing was carried out by a production apparatus of the system according to the invention. A production line 4 of that system is shown in FIG. 1. The production line 1 feeds the sheets 3a, b in a substantially feed 6. The foamed unit 1 shown is separated from the feed 6 by cutting through the feed 6, which cutting results in a foamed unit edge 5.

[0040] A camera 7 of the system according to the invention is arranged to capture an image 8 of the foamed unit edge 5. This image 8, which is shown in FIG. 1 as an image file, is analyzed by a computer system 10 of the system according to the invention to detect defects 9 in the foamed unit edge 5. As a result of this analysis, the computer system 10 generates quality information data 11 in which the detected defects 9 are described. In particular, the quality information data 11 comprises a classification of the detected defects into bubbles 12 within the foam 2, cracks 13 within the foam 2, voids 14 within the foam and overrolling marks 15 within the foam 2. The quality information data 11 further comprises a count of each class of detected defect 9 as well as the respective size and position of each detected defect 9, i.e. their distance both to the sheets 3a, b as well as to a lateral edge 16 of the foamed unit 1, and a severity classification. Thus, each detected defect 9 is classified as being of low severity, medium severity or high severity.

[0041] The production process of the foamed unit 1 may be described by production parameters 17, which in turn comprise production variables 18 measured by production instruments 20 of the production apparatus and production settings 19 input to the production apparatus and in particular to the production line 4.

[0042] The production parameters 17 are also provided to the computer system 10, which then generates updated production settings 21 by applying the quality information data 11 as well as the production parameters 17 to a calculation model 22, which calculation model 22 is saved on the computer system 10. This calculation model 22 is a software to simulate the production process. In particular, the calculation model 22 permits to predict the occurrence of defects 9 based on the production parameters 17 used in the production of a foamed unit 1, which is here a panel. Based on the same data set, the calculation model 22 further permits to arrive at improved production settings 19 based on the quality information data 11 describing detected defects 9 of a foamed unit and the production parameters 17 used in the production of that foamed unit 1. In other words, the calculation model 22 is able to determine which of the production settings 18 needs to be adjusted in order to avoid those defects 9 that were detected according to the quality information data 11 obtained from the foamed unit 1 with the defects 9.

[0043] These updated production settings 21 are then applied to the production system for the production of a further foamed unit, which further foamed unit is not shown here. This further foamed unit is then subjected to the same analysis as the foamed unit 1. Based on this analysis and the generation of a further quality information data from the further foamed unit, the calculation model 22 is updated by the computer system 10 in order to more closely align the defects that were actually detected from the further foamed unit as described in the further quality information data with the defects that were expected according to the calculation model 22 based on the updated production settings 21. In this way, the calculation model 22 can be successively made more accurate.

[0044] Specifically, an analysis algorithm 23 carried out on the computer system 10 generated the quality information data 11 from the image 8. This analysis algorithm 23 also takes into account for each class of defect 9 its relevance for the thermal insulation of the foamed unit 1. To assess the accuracy with which the analysis algorithm 23 considers each class of defect 9 relevant, the foamed unit 1 is assembled with further foamed units to form a building component and this building component is subjected to a thermal insulation test, resulting in a thermal insulation measurement. The thermal insulation measurement can then be used to update both the analysis algorithm 23 as well as the calculation model 22.

[0045] The embodiment of the system according to the invention in FIG. 2 is a simplified variant, with only the differences to the embodiment of FIG. 1 being described in the following. In this embodiment, the quality information data 11 is compared to pre-defined alarm criteria 24. If the specified conditions of the pre-defined alarm criteria 24 are met, which means that the panel 1 has defects 9 such that it fails to comply with quality demands, an alarm signal 25, which here is a visual output is generated. Further, the updated panel production settings 21 are here generated by comparing the quality information data 11 with a pre-defined rule set 26.