Method for checking, marking and sorting individual objects

Abstract

The invention relates to a method for checking, marking and sorting individual objects for combination with other individual objects in at least one component. The present invention makes it possible to simplify and refine this combination significantly. All the component parts and each assembly up to the end product (for example, a vehicle) has its own unique digital twin.

Description

[0001] Methods for checking, marking and sorting individual objects

[0002] The invention relates to a method for testing, marking and sorting individual objects for assembly with other individual objects in which at least one component is provided.

[0003] From the applicant's own German patent DE 10 2019 122 578 Bl, a method for checking, marking and sorting individual objects is already known, comprising the following steps:

[0004] Method for checking, marking and sorting individual objects, comprising the following steps:

[0005] - Providing a coated, tempered, uncoated or untempered Metal I strip / tape, characterized by

[0006] - Applying an individual marking to sections of the tape / strip,

[0007] - Punching and / or lasering and / or bending and / or forming sections into individual objects,

[0008] - Checking the individual objects,

[0009] - Linking and saving the respective marking with the respective test result for each individual object,

[0010] - Creation of an individual test certificate for each individual object,

[0011] - Making the test results available to third parties.

[0012] These are, in particular, parts made of metal, for example, by punching from a coated or uncoated, tempered or untempered metal strip or metal II strip. The present invention takes up the basic idea of ​​this initial invention and develops it further, independent of the material (e.g., instead of metal, the part can also be made of plastic, wood, textile, or any other material) and / or the manufacturing process (e.g., instead of producing the part using a punching or forming process, it can also be cast, injected, printed, extruded, or manufactured using any other manufacturing process).

[0013] According to the invention, at least one component is provided, wherein the following steps are subsequently carried out:

[0014] - Applying an individual marking to the component

[0015] - Checking the individual object

[0016] - Linking and saving the respective marking with the respective test result for each individual object,

[0017] - Creation of an individual test certificate for each individual object

[0018] - Comparison of the certificates of different individual objects intended for assembly,

[0019] - Selection of individual objects with suitable test certificates and,

[0020] - virtual assembly of the appropriate individual objects to form a complete object and

[0021] - Release for physical assembly.

[0022] Furthermore, it is planned that the data of all individual objects (6, 6', 6"-6 n ) are compared with each other and, based on the ideal measured values, those individual objects that best match each other are assigned to each other.

[0023] In the design, it is provided that after the virtual assignment, the physical assignment and assembly of the individual objects (6, 6', 6"-6 n ) can be triggered.

[0024] The invention also includes a method for capturing feedback from physical assembly, comprising confirmation of proper assembly, assembly effort, e.g. assembly time and pressing and friction forces, test values ​​of the functions of the finished assembly, describing the function of the finished assembly (performance), e.g. the force required to actuate it, the spring forces acting during actuation as well as the speed, noise and safety of the assembly function, as well as a computer program comprising instructions for carrying out these methods and a computer program product that can be loaded directly into the internal memory of a digital computer and includes software code sections with which all steps are carried out when the product is running on a computer.

[0025] In particular, a computer program product is provided, stored on a computer-usable readable medium, wherein at least one component is provided to which an individual marking 11 is applied, wherein the individual object is tested and the respective marking 11 is associated with the respective test result for each individual object 6, 6', 6"-6 n linked and saved, an individual test certificate for each individual object 6, 6', 6"-6 n created and a comparison of the certificates of various individual objects that are intended for assembly is carried out, the individual objects with suitable test certificates are selected and virtual assembly the appropriate individual objects are virtually assembled into an overall object and then a release for physical assembly takes place.

[0026] According to the invention, before the individual components are physically assembled, suitable software first generates a virtual assembly of all individual components, each individually labeled. All component data for the individual components is compared with each other. This can include, among other things, specific measured values.

[0027] For example, an assembly can consist of three components AC. In a first example, component A consists of a stamped part, e.g., made of stainless steel, with a hole. Component B represents a plastic clip with a mushroom head that is to be installed through the hole in stamped part A. Component C is a rubber plug that is slipped over the mushroom head after the plastic clip B is installed.

[0028] According to current practice, the components are manufactured individually according to component drawings. The individual features of the components are manufactured according to the specific manufacturing possibilities within a specific tolerance range.

[0029] To assemble the module, one component each, A, B and C, is taken from a container and assembled together.

[0030] The quality of the component connection is thus determined randomly and uncontrolled within the possible tolerance combination.

[0031] The present invention offers numerous advantages. All individual components or parts are individually labeled. Before physical assembly begins, the software can compare the respective data of the components to be installed. This allows the ideal part combinations to be selected.

[0032] The criteria for selecting which part combinations fit together best can be optionally selected (e.g. best tolerance level, best matching to utilize all possible part combinations from a delivery to avoid waste, lowest CO2 emissions).

[0033] For example, a custom component B with a correspondingly large expansion head can be selected to match a large hole diameter in component A. Based on the individual markings of the individual components, the software decides which of the available components B enables an optimal tolerance combination. Based on the data of the now virtually created subassembly A + B, the software decides which optimal component C fits this.

[0034] Only after the virtual assembly, i.e. the selection of the optimal individual components A, B and C, is the release for physical assembly given.

[0035] Thus, the quality of the component connection is optimized and controlled, a selection is made within the best possible tolerance combination, e.g. within the best possible tolerance combination (mating).

[0036] This results in conscious and advantageous assembly of the components. The selection is carried out using software trained by artificial intelligence (AI). This software is self-learning, as each automated piece of information about the assembly (e.g., press-in forces and test results) improves the software's understanding of the selection criteria. In addition to the primary selection criteria, such as diameter tolerances, all other product data is processed based on the individual markings, i.e., a digital twin created in this way. The software thus continues to learn which components fit together optimally. This can also be done using so-called pattern analysis.

[0037] The CO2 footprint of the assembly is automatically calculated. This is determined by adding the components plus the consumption data from the assembly process feedback. Here, too, the software can continuously optimize the process by selecting a component combination that generates the smallest overall CO2 footprint.

[0038] This also prevents the scrap or surplus of components that arises from the indiscriminate consumption of individual parts, which corresponds to the current state of the art, always leaving parts that do not match each other in their respective tolerances and are therefore discarded. In addition to the individual components, the assembly also receives a unique part ID, thus creating a digital twin for this assembly IT system.

[0039] With this digital twin of the BC assembly, additional application data can be collected throughout the component's lifecycle. This allows potential failures to be directly compared with the assembly and individual component data.

[0040] If it is discovered in the field that, for example, a specific assembly configuration, a specific component characteristic, alone or only in combination with another component characteristic in a specific form or a combination of several characteristics, leads to errors, the affected components can be clearly identified immediately.

[0041] The process also captures feedback from the physical assembly process. This can include, among other things:

[0042] • Confirmation that the installation worked perfectly

[0043] • Assembly effort (e.g. pressing forces, friction forces, process time)

[0044] • Test values ​​of the function of the finished assembly, which describe the performance of the assembly function (e.g. force required for actuation, spring forces during actuation, speed, noise, safety of the assembly function)

[0045] Based on these findings, the process uses AI-based methods to analyze which parameters are relevant for each individual component in combination with another component in order to achieve certain performance values. After specifying a target performance value, the process can not only find the best possible part combination based on predefined parameters, but also independently determines which parameters are relevant for each component in combination with the parameters of the other components. This allows the process to optimize its selection of component combinations.

[0046] A possible example scenario will be illustrated using automobile production as an example: A vehicle manufacturer or an original equipment manufacturer (OEM) identifies a problem in the field. The subsequent failure analysis reveals that all installed individual parts conform to the drawings. However, it turns out that a specific component, e.g., a stamped part, with a specific material batch (e.g., a manganese content in a steel batch at the upper end, with simultaneous tensile strength at the lower end) causes a problem if it was manufactured at the upper end for a specific component characteristic (e.g., bend angle).

[0047] However, this problem only occurs in the field if such a component has been installed with a rivet whose diameter is at the upper end of the tolerance. Therefore, the defect is localized to the "stamped part" component type in combination with the rivet.

[0048] In previous practice, it has been determined that 20,000 components of a stamped part were manufactured from a given batch of material. The records of the sample inspection during production only allow the conclusion that possibly up to 5% of the manufactured parts have the bend angle at the upper end. However, an exact number of parts cannot be determined. So far, the possible occurrence of the defect has been limited to the aforementioned 20,000 components.

[0049] During assembly production, the installation of these components can be limited to one batch within 35,000 possible assemblies.

[0050] The manufacturer or OEM then initiates a recall of all 35,000 products in which these assemblies were incorporated. This creates enormous time and expense and upsets numerous customers and buyers.

[0051] The present invention significantly simplifies and limits this process. All component parts and subassemblies, right up to the final product (e.g., a vehicle), have their own unique digital twin.

[0052] The manufacturer or OEM has all the data and can therefore digitally determine the exact number of components (e.g. stamped parts) that have the specific feature (e.g. only 3,000 of the total 20,000 pieces).

[0053] The individual part IDs of the affected parts can be identified, it can be read out in which assemblies they have been installed, and these assemblies can be clearly listed with their individual part IDs.

[0054] For each of the affected assemblies, it can be analyzed whether rivets have been installed that have the diameter at the upper end (e.g. only 500 pieces).

[0055] This allows the manufacturer or OEM to focus its recall on precisely those few motor vehicles where this specific combination of defects occurs, for example, stamped parts from a specific material batch with a bending angle at the upper end of the tolerance in combination with a rivet whose diameter is at the upper end of the tolerance.

[0056] The advantages are clear:

[0057] The manufacturer or OEM has the certainty that all parts affected by the defect have been identified. At the same time, the recall campaign is limited to only those vehicles that are actually technically affected, thus avoiding the enormous effort and unnecessary costs previously incurred. The speed of fault isolation and correction is increased, and the risk of vulnerable vehicles remaining on the market is reduced or even eliminated. Overall, the effort of the entire recall campaign is reduced to a fraction of the current batch screening and costs.

[0058] The invention provides a method for testing, marking, sorting individual objects 6 for assembly with other individual objects in at least one component, which comprises the following steps:

[0059] - Applying an individual marking to the component

[0060] - Checking the individual object

[0061] - Linking and saving the respective marking with the respective test result for each individual object,

[0062] - Creation of an individual test certificate for each individual object,

[0063] - Comparison of the certificates of different individual objects intended for assembly,

[0064] - Selection of individual objects with suitable test certificates and,

[0065] - virtual assembly of the appropriate individual objects to form a complete object and

[0066] - Release for physical assembly.

[0067] The invention is explained in more detail below with reference to the drawing, which shows

[0068] Fig. 1 shows a schematic overview of the sequence of the production method according to the invention and the arrangement according to the invention,

[0069] Fig. 2 a sketch of the initial situation,

[0070] Fig. 3 the 1st step of a conscious selection part A2, Fig. 4 the 2nd step of a conscious selection part B3,

[0071] Fig. 5 the 3rd step of a conscious selection part CI and in

[0072] Fig. 6 the summary of the selected parts A2, B3 and CI for

[0073] Further processing.

[0074] A metal strip or band 4 is fed from a coil 3 wound on a reel 2 to a production machine, generally designated 1, for example a punching machine. In a coding station 5, a marking or individual identification 11 is applied to sections of the strip / band 4, which correspond to the subsequent individual objects 6. Processing then takes place in the machine 1, for example, punching or stamping sections of the strip / band 4 into individual objects 6 by a punching device 12. The applied code 11 is then read out by a reading device 10. The individual objects 6 are simultaneously measured and inspected, for example by means of two camera systems 7 and 8, after they are individually separated and deposited or collected, for example in a collecting container 9.If necessary, further processing, such as heat treatment, may also be performed. Packaging, which is also not shown in detail, follows. The test result(s) for each individual object 6 are linked to the respective marking 11 for the same individual object 6 and saved. From this, an individual test certificate is created for each individual object 6, which can be made available to third parties, such as buyers and / or customers.

[0075] 1.

[0076] Initial situation: A quantity X of three different components each to be assembled into an assembly A+B+C: Each component of each type is individually labeled, and there is a digital twin for each component, a so-called PlockVatar. This not only provides the assembly company with the information that there are X parts of type A, B, or C, but also identifies each component as a product-specific item thanks to the individual labeling.

[0077] 2.

[0078] Targeted selection of a component of type A, according to the optimal selection

[0079] Before physical assembly, the process virtually determines the best possible combination of all components for assembly A+B+C. The process therefore selectively selects, for example, component A2 for the initial assembly.

[0080] 3.

[0081] Targeted selection of a component of type B, according to the optimal selection

[0082] The process then selectively selects, for example, component B3 for the first assembly.

[0083] 4.

[0084] Targeted selection of a component of type B, according to the optimal selection

[0085] The process then selectively selects, for example, the component CI for the first assembly.

[0086] The assembly A2+B3+C1 is tested, measured and evaluated.

[0087] The results of the physical assembly are compared with the virtual assembly prediction. The method automatically optimizes the selection and weighting parameters for calculating the optimal part combination for assembly A+B+C. Of course, the invention is not limited to the illustrated embodiments. Further embodiments are possible without departing from the basic concept.

[0088] List of reference symbols:

[0089] 1 production machine

[0090] 2 reels

[0091] 3 coils

[0092] 4 metal bands / strips

[0093] 5 coding stations

[0094] 6 individual objects

[0095] 7 camera systems

[0096] 8 camera systems

[0097] 9 collection containers

[0098] 10 Reading device

[0099] 11 Marking / Labeling

[0100] 12 Punching device

Claims

Patent claims:

1. Method for checking, marking, sorting individual objects (6, 6', 6"-6 n ), for assembly with other individual objects (6, 6', 6"-6 n ), which includes the following steps: - Provision of at least one component characterized by - Applying an individual marking (11) to the component - Checking the individual object - Linking and saving the respective marking (11) with the respective test result for each individual object (6, 6', 6"-6 n ), - Creation of an individual test certificate for each individual object (6, 6', 6"-6 n ) - Comparison of the certificates of different individual objects intended for assembly, - Selection of individual objects with suitable test certificates and, - virtual assembly of the appropriate individual objects to form a complete object and - Release for physical assembly. Method according to claim 1, characterized in that the data of all individual objects (6, 6', 6"-6 n ) are compared with each other and, based on the ideal measured values, those individual objects that best match each other are assigned to each other.

3. Method according to claim 2, characterized in that after the virtual assignment, the physical assignment and the assembly of the individual objects (6, 6', 6"-6 n ) can be triggered.

4. Method for recording feedback from the physical assembly, comprising a confirmation of the correct assembly, its assembly effort, e.g. the assembly time and the pressing and friction forces, the test values of the functions of the finished assembly, describing the function of the finished assembly (performance), e.g. the force required for its actuation, the spring forces acting during actuation as well as the speed, noise and safety of the assembly function.

5. Computer program comprising instructions for executing the method according to at least one of claims 1-4.

6. A computer program product which can be loaded directly into the internal memory of a digital computer and which comprises software code sections which carry out the steps according to method claims 1 to 4 when the product is run on a computer.

7. A computer program product stored on a computer-usable readable medium, comprising

Images