Design support device and method

Abstract

To provide a design support device and a method that facilitate and automate evaluation of a man-hour, difficulty, and a necessary tool required to extract a specific component or a subassembly from a three-dimensional CAD model.SOLUTION: A design support device 100 for supporting design of a design object includes: an input unit that receives a three-dimensional CAD model of the design object; a fastening shape recognition unit that recognizes a fastening shape including a shape of a fastening portion of a component constituting the design object, an attribute of the fastening portion of the component, and a fastening relationship between the component and another component from shape information included in the three-dimensional CAD model; an important component selection unit that selects the component to be analyzed from the component constituting the design object; a disassembly path generation unit that generates a disassembly path including a disassembly method for the selected component using the fastening relationship; an evaluation value calculation unit that calculates an evaluation value indicating an evaluation of the disassembly method included in the disassembly path by using the attribute of the fastening portion of the component; and an output unit that outputs at least one of the disassembly path and the evaluation value.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a technology for assisting the design of design targets such as products and assemblies.

Background Art

[0002] In recent years, in order to reduce the greenhouse gas emissions and resource waste of products, there has been a movement to improve the maintainability, reparability, and recyclability of products and disclose such information. In order to improve maintainability, reparability, and recyclability, it is necessary to grasp, and more preferably improve, the ease of disassembly, which is the ease of accessing the target parts and sub-assemblies. The ease of disassembly is determined by the configuration and fastening method of parts and modules defined during design. Therefore, it is required to grasp the ease of disassembly at the design stage in order to improve the ease of disassembly.

[0003] Here, in 3D computer-aided design (hereinafter referred to as 3D CAD (Computer Aided Design)), which is currently widespread as a design technology, a designer creates a 3D shape of a product on a computer by techniques including solid modeling and parametric modeling. In many 3D CAD software, the 3D shape is represented by BREP (Boundary REPresentation) that describes the solid, surface, edge, point of the shape and its topological information. Hereinafter, the 3D shape created by the aforementioned 3D CAD software is referred to as a CAD model.

[0004] In 3D CAD, 3D annotations are used. These 3D annotations are annotations containing information such as tolerances, welding, and surface finishes, which are given to the shapes including solids, surfaces, edges, points, etc. in the CAD model, or attribute information to which various forms of information including part types and specifications are given. And the 3D annotations are mainly used to describe product requirements, manufacturing requirements, and manufacturing instructions. Furthermore, by associating these attribute information with work instructions such as manufacturing, assembly, and disassembly procedures as included in the BOP (Bill of Process), it can be utilized in downstream manufacturing and maintenance. Also, based on the specified assembly and disassembly procedures, the feasibility, man-hours, difficulty level, etc. of the work can be evaluated at the design stage.

[0005] On the other hand, in order to evaluate disassemblability, etc. at the design stage, it is necessary to carefully check the CAD shape from various angles, compare it with the disassembly procedure, and estimate man-hours, etc., which requires a certain amount of labor.

[0006] Therefore, in Patent Document 1, information on work contents (assembly, fastening, wiring, etc.) is defined in advance for each part type, component data constituting the product is extracted from the input CAD data, and the user rearranges the work contents corresponding to each component. By doing this, Patent Document 1 has proposed an apparatus that defines the entire work content and calculates man-hours and difficulty level.

Prior Art Documents

Patent Documents

[0007]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0008] However, the technology described in Patent Document 1 has the following four problems. First, in Patent Document 1, it is necessary to define procedures for all part types in advance, which requires a lot of man-hours. Also, since the procedures are linked for each part type, it is impossible to handle new parts or parts with design changes, and it is necessary to define new patterns.

[0009] In the technology described in Patent Document 1, like fastening methods such as snap fits, the workability may vary greatly depending on the partial shape included in the part. For this reason, the method of obtaining information based on the parts in Patent Document 1 cannot handle cases where the assembly and disassembly properties change according to the partial shape in this way.

[0010] Also, in the technology described in Patent Document 1, the assembly and disassembly procedures are defined by manually rearranging them, but depending on the order of assembly and disassembly, part interference, available tools, etc. also differ. For this reason, in Patent Document 1, when defining a disassembly order where it is difficult for the disassembly tool to access in the manual definition, it is difficult to notice the mistake. Also, when the procedure itself is complicated or there are many steps, rearranging itself also requires a lot of man-hours.

[0011] Furthermore, in the technology described in Patent Document 1, mainly assuming assembly, the user rearranges the work content, but the maintenance and disassembly procedures are not necessarily the reverse of the assembly procedure, and it is necessary to consider the method of accessing each target part in the shortest way. For this reason, in Patent Document 1, when the number of target parts increases, it is necessary to define the work order for each part, and the evaluation man-hours increase.

Means for Solving the Problems

[0012] In order to solve the above problems, in the present invention, using the 3D CAD model of the design object, the fastening shapes of the respective components constituting the design object are recognized, and using this recognition result, the disassemblability of the components is analyzed. This analysis of disassemblability includes identifying at least one of a disassembly path indicating a disassembly method and an evaluation value indicating an evaluation of the disassembly. Note that the components of the present invention include, in addition to individual components, sub-assemblies composed of combinations of a plurality of components and parts. Also, design objects include products, assemblies, devices, equipment, and systems.

[0013] A more specific configuration of the present invention is a design support device for supporting the design of a design object, including an input unit that receives the 3D CAD model of the design object, a recognition unit that recognizes the shape of the fastening part of the component constituting the design object, the attribute of the fastening part, and the fastening shape including the fastening relationship between the component and other components from the shape information included in the 3D CAD model, a component selection unit that selects a component to be an analysis target from the components constituting the design object, a path generation unit that generates a disassembly path including a disassembly method of the selected component using the fastening relationship, an evaluation value calculation unit that calculates an evaluation value indicating an evaluation of the disassembly method included in the disassembly path using the attribute of the fastening part, and an output unit that outputs at least one of the disassembly path and the evaluation value. The present invention also includes a design support method by this design support device.

Advantages of the Invention

[0014] According to the present invention, it becomes possible to more easily grasp the disassemblability of components.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5A

Figure 5B

Figure 6

Figure 7

Figure 8

Mode for Carrying Out the Invention

[0016] Hereinafter, an embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a functional block diagram showing an example of the configuration of the design support apparatus 100 in the present embodiment. In FIG. 1, the design support apparatus 100 includes an input unit 101, a fastening shape recognition unit 102, an important part selection unit 103, a disassembly path generation unit 104, an evaluation value calculation unit 105, an output unit 106, a storage unit 107, and a document generation unit 111. Hereinafter, each of these units will be described. First, the input unit 101 receives a 3D CAD model of a design target such as a product. The input unit 101 receives operations of the user on the design support apparatus 100 such as selection of a local file via a user interface and selection of a file on a server.

[0017] In addition, the input unit 101 receives a 3D CAD model through instructions by a program or server communication via an API (Application Protocol Interface). The 3D CAD model includes a BREP, shape information including a tessellation mesh, and annotations such as notes and attribute information including manufacturing information.

[0018] In addition, based on the 3D CAD model received by the input unit 101, the fastening shape recognition unit 102 performs recognition of the fastening shapes in each part to be designed, that is, shape recognition. Here, shape recognition means recognition of the shape of the fastening part, the attributes of the fastening part, and the fastening relationship between each part, that is, identification of these. Here, the fastening part includes, for example, fastening parts used for fastening such as bolts and nuts, and fastening parts of parts related to fastening such as snap fits and welded parts. In addition, the attributes of the fastening part include attribute information related to the content including whether the fastening part can be disassembled, the necessary tools, assembly, and disassembly difficulty. It is desirable to use the shape of the fastening part and the attributes of the fastening part as the shape of the fastening part and the attributes of the fastening part.

[0019] In addition, the fastening relationship indicates which parts among the fastening parts are fastened to each other in the assembly shown by the design target, for example, the 3D CAD model. The fastening shape recognition unit 102 only needs to be able to recognize the fastening shape of each part, and can also be expressed as a recognition unit. Furthermore, the fastening shape recognition unit 102 may be configured to recognize the fastening shape of the parts selected by the important part selection unit 103 described later.

[0020] In addition, the important part selection unit 103 selects one or more parts whose disassemblability is to be analyzed from the 3D CAD model received by the input unit 101. This selection may be made manually, that is, according to a designation from the user, or automatically. When performed automatically, it can be selected based on certain criteria based on the shape of the parts shown in the received 3D CAD model, annotation information, and related external information including the failure probability. In this way, the important part selection unit 103 only needs to be able to select the parts to be analyzed based on any criteria or methods, and can also be expressed as a part selection unit.

[0021] Further, the disassembly path generation unit 104 generates a disassembly path including a disassembly method for the components selected by the important component selection unit 103. For this purpose, the disassembly path generation unit 104 uses the shape of the fastening part, which is the fastening situation / aspect / form recognized by the fastening shape recognition unit 102 for the selected components, the attributes of the fastening part, and the fastening relationships between the components. Then, the disassembly path generation unit 104 generates a disassembly path including a disassembly method for taking out each of the selected components. The disassembly methods included in the disassembly path include the order of use of the selected components, the working order including the order of using tools, and the geometric path when taking out the components. Note that the disassembly path generation unit 104 may have a function of generating an assembly path for the design target. For this reason, the disassembly path generation unit 104 can also be expressed as a path generation unit.

[0022] Also, the evaluation value calculation unit 105 calculates an evaluation value for evaluating the generated disassembly path, particularly the disassembly method included therein, using the attributes of the fastening part (particularly, the attributes of the fastening components). At least one of the feasibility, difficulty level, man-hours, etc. of the generated disassembly path can be used for this evaluation value. This evaluation value may be calculated in parallel with the generation of the disassembly path, or may be calculated at a different timing. Further, the disassembly path generation unit 104 may newly generate a disassembly path in order to search for a more suitable disassembly method when a predetermined condition is satisfied based on the evaluation value in the evaluation value calculation unit 105.

[0023] Also, the evaluation value calculation unit 105 outputs the calculated evaluation value, which is the evaluation result, to the disassembly path generation unit 104, the output unit 106, and the document generation unit 111. Note that in the calculation of the evaluation value, the results obtained upstream in the processing of the design support apparatus 100, such as the input unit 101, the fastening shape recognition unit 102, the important component selection unit 103, and the disassembly path generation unit 104, are used. Further, the evaluation result includes, in addition to the numerical value of the evaluation value, gradient information, reliability, and a probability distribution including errors.

[0024] The output unit 106 outputs the disassembly path and / or its evaluation value calculated by the evaluation value calculation unit 105. The output unit 106 can be realized by a communication device using a network or a display device that displays information. As a result, it becomes possible to present these to the user. For presenting to the user, the document generated by the document generation unit 111 can be used. The document includes information for display on a graphical user interface and information in a machine-readable structured data format such as JSON (JavaScript Object Notation, JavaScript is a registered trademark). Further, the presented document includes disassembly information regarding the disassembly method indicated by the generated disassembly path. This disassembly information can be configured to include the generated disassembly path and / or the calculated evaluation value. As a more specific example, it includes the work order of the disassembly path, the movement paths of parts and tools corresponding to each work, information in the form of animation of these, and a plurality of evaluation values calculated by the evaluation value calculation unit 105.

[0025] Further, the storage unit 107 stores the CAD shape DB 108, the fastening attribute DB 109, and the failure probability DB 110. First, the CAD shape DB 108 includes 3D CAD shape data (including shape information) that may be used in the shape recognition algorithm used in the recognition process by the fastening shape recognition unit 102. Further, the fastening attribute DB 109 includes data on the attributes of the fastening shape including the fastening part site that may be used in the shape recognition algorithm by the fastening shape recognition unit 102. Examples of the shape recognition algorithm include a method based on similar shape search and a method of performing shape recognition based on statistical processing of performance data. In these methods, for example, a similar shape search using the CAD shape of the CAD shape DB 108 is performed, and further, the fastening attribute information associated with the CAD shape is referred from the fastening attribute DB 109 and can be used for fastening shape recognition.

[0026] In addition, the failure probability DB 110 stores evaluation indexes including the functional importance of each component and the failure probability of that component. These evaluation indexes can be utilized when selecting components to be assembled or disassembled by the important component selection unit 103, or when calculating the evaluation value by the evaluation value calculation unit 105. For example, the important component selection unit 103 can select important components according to the threshold values of functional importance and failure rate. Alternatively, when calculating the evaluation value of the disassembly path, the evaluation value calculation unit 105 can use the failure rate information to calculate an evaluation value including the disassembly man-hours generated per period.

[0027] In addition, when the design support device 100 is realized by a server or the like and a terminal device is provided separately, the input unit 101 and the output unit 106 can be omitted. With the above, the description of the configuration of the design support device 100 of the present embodiment is completed, but the design support device 100 can be realized by a computer. This realization example will be described later with reference to FIG. 8.

[0028] Next, the fastening shape used in this embodiment will be described. The fastening shape in this embodiment indicates the state of fastening of the components to be designed, and is a combination of the components to be fastened and the fastening components. For example, it includes the shape of the fastening part, the attributes of the fastening part, and the fastening relationship of each component.

[0029] Here, FIG. 2 is a diagram for explaining the fastening shape in this embodiment. The fastening shape recognition unit 102 recognizes the attribute information related to fastening including the fastening components involved in the fastening, the partial shape of the fastening part in the component, the disassemblability, necessary tools, disassembly direction, disassembly difficulty, and man-hours, which are included in the fastening shape. Further, the fastening shape recognition unit 102 recognizes the fastening relationship of the components included in the fastening shape. In FIG. 2, 201 is an example of a CAD model including fastening by bolts, nuts, and screws. The fastening shape recognition unit 102 uses such a CAD model to perform component recognition of bolts, nuts, screws, or washers attached to the bolts, and attribute recognition including their respective bolt diameters, tools, directions, etc.

[0030] In addition, 202 shows an example of a CAD model including fastening by snap fit or clip. In the fastening shape recognition unit 102, after recognizing partial shapes such as snap fits and clips included in the solids in this CAD model, attribute information such as disassemblability, necessary tools, and force required for disassembly is recognized.

[0031] In addition, 203 shows an example of a CAD model including components joined by welding. In the fastening shape recognition unit 102, based on such a CAD model, the annotation of the welding is recognized, or partial shapes such as fillet weld locations and groove weld locations are recognized based on shape recognition, and sides and surfaces that are partial shapes corresponding to the welded locations are recognized. Also, in the fastening shape recognition unit 102, attribute information such as welding method, type, and thickness is recognized for the recognized welded locations. Furthermore, in the fastening shape recognition unit 102, it is recognized as a fastening (joining) method that is basically non-disassemblable for the welded locations.

[0032] Furthermore, 204 shows an example of the result recognized as a fastening shape. Specifically, 204 shows an example of recognizing the recognized fastening attribute information and the fastening relationship between components as a graph. The graph of the fastening relationship is a graph in which components are nodes and edges are defined when components are in contact with or fastened to each other. Regarding fastening components such as bolts and screws, or partial shapes such as snap fits and welding, fastening attributes in key-value format are recognized as in 204.

[0033] Next, the disassembly analysis process in this embodiment will be described. First, the display for executing this analysis process will be described. FIG. 3 is a diagram showing an example of a graphical interface 301 (display screen) for selecting components to be analyzed in this embodiment. The important component selection unit 103 does not necessarily need to select components in response to an input to the graphical interface 301, and selection by a program such as an API is also possible. Therefore, in this embodiment, the graphical interface 301 shown in FIG. 3 can be omitted.

[0034] Here, in the graphical interface 301 of FIG. 3, a selection area 302 for selecting parts to be disassembled and a 3D CAD model 303 to be designed are displayed. The user designates one or more parts to be disassembled, that is, the parts to be analyzed, from the tree (list) of "parts to be disassembled" (components) in the selection area 302. That is, "part0002" is designated from "part0002" and "part0003" in the figure. Alternatively, the user designates a part from the 3D CAD model 303. In this case (the shaded part in the figure). In response to this, the important part selection unit 103 selects the designated part. Also, the graphical interface 301 has a reference information display area 304. The reference information display area 304 includes failure rates, functional importance, selling price, cost, etc., which are the criteria for selecting parts to be disassembled. When selecting parts by a program such as an API, the selection is executed based on reference information, part names, their attributes, etc.

[0035] Also, when the selection button 305 of the graphical interface 301 is pressed by the user, the calculation of the disassembly path of the selected part and the calculation of the evaluation value are executed. For this purpose, the disassembly path generation unit 104 and the evaluation value calculation unit 105 perform processing according to the designation of the selection button 305. Note that when the selection button 305 is designated, either the calculation of the disassembly path or the calculation of the evaluation value may be performed. With the above, the description of FIG. 3 is completed. First, the fastening shape recognition process in the fastening shape recognition unit 102 will be described. FIG. 4 is a flowchart showing an example of the processing flow of the fastening recognition process of the fastening shape recognition unit 102 in the present embodiment. First, the fastening shape recognition unit 102 reads the 3D CAD model received by the input unit 101 (step S401). Next, the fastening shape recognition unit 102 extracts shape information, for example, BREP, the corresponding tessellation mesh, and annotation data from the read 3D CAD model (step S402).

[0036] Next, the fastening shape recognition unit 102 extracts fastening information from the extracted shape information (step S403). For example, from the annotation information, annotation and attribute information related to fastening attached to parts and partial shapes are extracted. Examples of this fastening information such as annotation and attribute information include part attributes such as the model number, diameter, and length of the bolts being used, and annotation information of the welding method attached to partial shapes such as the annotation of the welded part.

[0037] Next, the fastening shape recognition unit 102 uses the extracted fastening information to recognize the fastening shape (step S404). For example, the fastening shape recognition unit 102 performs the recognition of the fastening shape based on the BREP information, mesh shape information, and attribute information obtained in step S403. Here, the recognition of the fastening shape includes the recognition of fastening parts that are examples of fastening sites, or the partial shapes of fastening parts in the parts.

[0038] These recognition methods include the following methods (1) to (5). (1) A feature recognition method by graph search using a face adjacency graph representing the face-edge adjacency relationship included in the BREP and the geometric information of each face and edge (2) A method of recognizing a shape by a statistical method including a graph neural network or the like with the graph of (1) as an input (3) A method of recognizing a fastening shape by a neural network or pattern matching method using a mesh shape or a corresponding point cloud, rendering image, signed distance field, voxel, etc. as inputs Note that as outputs of the methods (1) and (2), by calculating the vector representation of each part and partial shape, the difference in vectors such as the cosine similarity with the 3D CAD model in the CAD shape DB108 is calculated to perform a similar shape search. And there is a method of recognizing the fastening shape and attributes by outputting the fastening information corresponding to the CAD model with the highest similarity.

[0039] (4) A method of recognizing parts by a classifier that predicts by a neural network or the like what kind of fastening parts each part is A method of recognizing a partial shape by segmentation that statistically predicts which type of fastening method is to be established for each partial shape such as a BREP surface, a mesh-shaped surface, a point of a point cloud, etc. As described above, for each of the fastening parts and partial shapes recognized by the classifier and segmentation, the fastening attributes are estimated from the BREP shape surface side information and mesh information based on rules or statistically.

[0040] Then, the fastening shape recognition unit 102 generates a component-to-component fastening relationship graph indicating the component-to-component fastening relationship (step S405). For this purpose, the fastening shape recognition unit 102 calculates the distance relationship between components from the BREP and mesh shapes, and combines the fastening shape data to generate a component-to-component fastening relationship graph. This component-to-component fastening relationship graph is a graph having components as shown in 204 of FIG. 2 as nodes and the fastening relationship between components as edges. To create the above-described graph, the fastening shape recognition unit 102 calculates the distance relationship between components. To calculate the distance relationship between components, first, a bounding box is created for each of the mesh and the BREP surface. Next, the bounding boxes are hierarchically organized using a BVH (Bounding Box Hierarchy), a Kd tree, etc. to speed up the interference determination between the bounding boxes. Next, when the distance between components is equal to or less than a threshold value, the fastening shape recognition unit 102 calculates the distance of the bounding box.

[0041] Then, the fastening shape recognition unit 102 makes a high-speed determination based on the aforementioned tree structure. As a result, for parts whose bounding boxes are close within a predetermined range, the fastening shape recognition unit 102 calculates information including the distance between adjacent parts, the contact area, the contact angle, and the attributes of the fastening shape in contact. For this purpose, interference determination between meshes, distance determination, distance determination based on a signed distance field estimated from meshes, or distance calculation by Bézier clipping between BREP surfaces is used. Alternatively, a method including a ray method in which a ray is emitted from one part in a direction perpendicular to the surface and the distance until the ray intersects the surface of another part is measured is used. Then, based on the information calculated in this way, the edges between the parts included in the inter-part fastening relationship graph are defined. As a result, an inter-part fastening relationship graph is generated. Note that step S405, that is, the generation of the inter-part fastening relationship graph may be omitted.

[0042] Next, the processing in the disassembly path generation unit 104 and the evaluation value calculation unit 105 will be described. First, FIG. 5A is a flowchart showing the disassembly path generation process in the present embodiment. The disassembly path generation unit 104 reads the 3D CAD model received by the input unit 101, the recognition result of the fastening shape recognition unit 102, and one or more important parts selected by the important part selection unit 103 (step S501). Here, when the disassembly path generation unit 104 selects a plurality of important parts in step S501, steps S502 to S505 are repeated the number of times of the selected important parts.

[0043] Also, the disassembly path generation unit 104 determines whether the selected important part can be disassembled using the 3D CAD model (step S502). Further, the evaluation value calculation unit 105 calculates an evaluation value for evaluating the disassemblability, the difficulty level of disassembly, the man-hours, the contribution to the disassembly of the target part, etc. for each part of the design target (step S503). Note that in step S503, the disassembly path generation unit 104 can cause the evaluation value calculation unit 105 to calculate the evaluation value.

[0044] Also, based on the calculated evaluation values, the disassembly path generation unit 104 samples a plurality of disassemblable parts (step S504). For this purpose, the disassembly path generation unit 104 extracts a predetermined number of disassemblable parts according to a predetermined rule or randomly. For example, as sampling of disassembled parts, sampling can be performed by methods such as random search or Monte Carlo tree search. Furthermore, the calculation process of the evaluation value and sampling can be implemented by a differentiable method such as a neural network, and it is also possible to obtain the evaluation value and perform sampling by a reinforcement learning method such as policy gradient.

[0045] Also, the disassembly path generation unit 104 creates an excluded design target, which is a configuration with each of the sampled parts removed from the design target, and generates a disassembly path for the corresponding important parts. Then, the disassembly path generation unit 104 determines whether the excluded design target meets the disassembly completion criterion (step S505). For this purpose, the disassembly path generation unit 104 determines whether the evaluation value in step S503 is less than the threshold value. As a result, if it is less than the threshold value and meets the disassembly completion criterion (YES), the process proceeds to step S506. Also, if it is greater than or equal to the threshold value and does not meet the disassembly completion criterion (NO), the process proceeds to step S502, and the subsequent processing is repeated. As a result, this is done until all important parts are extracted from the design target. As a result, it is also possible to repeatedly execute until the disassembly paths until each part becomes independent during sampling are evaluated. Note that the disassembly path generation unit 104 can generate an assembly path (assembly) procedure by reversing the obtained disassembly path. Note that the disassembly path generation unit 104 may generate the disassembly path according to at least one of the disassembly purposes such as repair and recycling and the disassembly entity such as a disassembly contractor.

[0046] In addition, the disassembly path generation unit 104 identifies one or more disassembly paths that meet a predetermined criterion among the disassembly paths of the important parts for which the design target has been excluded. Here, the disassembly path that meets the predetermined criterion includes a disassembly path with a good evaluation value (for example, a disassembly path within a certain top number or above a threshold value). Then, the disassembly path generation unit 104 outputs the identified disassembly path to the evaluation value calculation unit 105 and the output unit 106 (step S506). At this time, it is desirable for the disassembly path generation unit 104 to also output the identified disassembly path to the document generation unit 111. In this case, the output of the disassembly path to the output unit 106 can be omitted. Also, when the output unit 106 is realized by a display device, the disassembly path will be displayed on the display device. Further, when the output unit 106 is realized by a communication device, the disassembly path will be notified to the terminal device and displayed on the terminal device.

[0047] With the above, the description of the disassembly path generation process in Fig. 5A is completed. Subsequently, the evaluation value calculation process will be described. Fig. 5B is a flowchart showing an example of the process flow of the evaluation value calculation process in this embodiment. First, the evaluation value calculation unit 105 receives the 3D CAD model, the fastening shape which is the recognition result of the fastening shape recognition unit 102, the selected important part, and the disassembly path generated in Fig. 5A (step S507).

[0048] In addition, the evaluation value calculation unit 105 determines whether the received disassembly path is physically achievable (step S508). For this purpose, the evaluation value calculation unit 105 can determine by means of a simple evaluation using only the recognized fastening attributes, or by evaluating whether there is a component movement path for removing components from the design target by a method including interference determination and mechanism analysis using the 3D CAD model shape and tool shape. In addition, for interference determination, global interference determination is performed by BVH, and detailed interference determination is performed by a method including intersection determination between meshes, interference determination combining convex hull decomposition and the GJK method, and interference determination using SDF. In addition, mechanism analysis is performed in combination with interference determination by a method including the penalty method, position-based dynamics, and the sequential impulse method. When it is necessary to consider component deformation, it is considered by means of software analysis using position-based dynamics or relaxation of interference determination criteria. For the determination of the component movement path for the determination of removing components, a path generation algorithm including RRT connect, etc. can be used.

[0049] As a result of step S508, if possible, the process proceeds to step S509. If not possible, this process ends. In this case, it is desirable for the evaluation value calculation unit 105 to output this determination result via the output unit 106. As a result, the determination result can be presented to the user.

[0050] In addition, the evaluation value calculation unit 105 calculates an evaluation value indicating an evaluation of the disassembly method included in the generated disassembly path using the fastening shape, particularly the attributes of the fastening part (step S509). This evaluation value includes, for example, evaluation values including disassembly difficulty, disassembly difficulty, number of tool switches, number of required tools, etc., and a comprehensive score value for use in sampling criteria, taking into account the disassembly man-hours predicted in subsequent disassembly steps in addition to the above-mentioned evaluation values. Here, the disassembly difficulty, disassembly difficulty, number of tool switches, number of required tools, etc. can be calculated from the disassembly path, the fastening attributes recognized by the fastening shape recognition unit, the component fastening relationship, and the component path analysis result described later.

[0051] Then, the evaluation value calculation unit 105 outputs the calculated evaluation value to the disassembly path generation unit 104 and the output unit 106 (step S510). As a result, the disassembly path generation unit 104 can execute step S503. Also, it is desirable for the evaluation value calculation unit 105 to output the calculated evaluation value to the document generation unit 111 as well. In this case, the output of the evaluation value to the output unit 106 can be omitted. Also, when the output unit 106 is realized by a display device, the evaluation value will be displayed on the display device. Also, when the output unit 106 is realized by a communication device, the evaluation value will be notified to the terminal device and displayed on the terminal device. Note that at least one of the output of the disassembly path in step S506 and the output of the evaluation value in this step may be performed. That is, at least one of the disassembly path and the evaluation value may be output. Further, the storage unit 107 stores order information indicating the assembly order and the disassembly procedure order for each fastening shape model. Then, the disassembly path generation unit 104 may generate a disassembly path and an assembly path using a learning model that has learned the relationship between the shape indicated by the fastening shape model and the order information. Note that the assembly path can be generated by reversing the disassembly steps of the disassembly path. Further, when disassembling, reassembly is required, so the disassembly path and the assembly path may be treated together as a disassembly path.

[0052] With the above, the description of the disassembly path generation process and the evaluation value calculation process is completed. Next, the display of the analysis results such as the disassembly path and the evaluation value will be described. FIG. 6 is a diagram showing an example of a graphical interface 601 (display screen) that displays the analysis results of important parts in the present embodiment. Note that the analysis results can also be output in a machine-readable format such as JSON via an API without going through the graphical interface 601.

[0053] Hereinafter, the content of FIG. 6 will be described. In FIG. 6, a disassembly path display area 602 for displaying the disassembly path generated by the disassembly path generation unit 104 is displayed on the graphical interface 601. In the disassembly path display area 602, one or more generated disassembly paths are listed in order based on the evaluation value, and have an evaluation value display area 603 and a disassembly step area 604.

[0054] And in the evaluation value display area 603, a plurality of evaluation values for each disassembly path are displayed. Also, in the disassembly step area 604, the disassembly steps (disassembly procedures) of the generated disassembly paths are displayed. That is, in the disassembly step area 604, the work contents for disassembly are listed. And when a disassembly step is selected from the disassembly step area 604, the following displays are made in the model display area 605 of the graphical interface 601. · The 3D CAD model of the design object being disassembled with a part corresponding to the corresponding disassembly step removed and 3D display · Emphasis display of the part to be disassembled in the selected disassembly step · Display of part movement animation based on the generated part movement path · The removal direction of the part.

[0055] Also, the graphical interface 601 has a disassembly step evaluation value display area 606. In the disassembly step evaluation value display area 606, the evaluation values related to the selected disassembly step are displayed. Note that in this embodiment, when the same disassembly steps such as removing bolts continue, a plurality of disassembly steps can be merged and displayed in the disassembly path display area 602.

[0056] The above concludes the description of FIG. 6, and the document generation process in the document generation unit 111 will be described. This document is generated using the aforementioned disassembly paths and evaluation values and presented to the user as an analysis result. More preferably, the document includes a maintenance manual or the like that describes how to execute each procedure of the generated disassembly path in a video, image, and natural language. Note that the presentation to the user may be the disassembly paths and evaluation values themselves, and this document generation process may be omitted.

[0057] FIG. 7 is a flowchart showing an example of the processing flow of the document generation process in this embodiment. The document generation unit 111 first acquires the 3D CAD model, disassembly path, and evaluation value corresponding to the disassembly path (step S701).

[0058] Also, the document generation unit 111 renders a plurality of component movement animations corresponding to the acquired disassembly path using the 3DCAD model (step S702). Also, the document generation unit 111 generates a sentence explaining the disassembly step based on the animation video corresponding to the disassembly step of the disassembly path, the necessary tools corresponding to each disassembly step, and information on fastening attributes such as the extraction direction (step S703). Here, the document to be generated can be realized in natural language and includes structured sentences such as markdown. Also, the generation of the sentence in this step can be performed using a neural network of a large language model composed of Transformer or the like. As a result, using a multimodal pre-training method for querying images and videos such as Blip, an image and fastening attributes or the like can be input to generate a sentence.

[0059] Also, the document generation unit 111 combines the generated sentence with the video and image corresponding to the disassembly path to create a document in the form of a sentence including an image, a 3D animation with a caption, or a video. Then, the document generation unit 111 outputs this to the output unit 106 (step S704). When the output unit 106 is realized by a display device, the document will be displayed on the display device. Also, when the output unit 106 is realized by a communication device, the document will be notified to the terminal device and displayed on the terminal device.

[0060] With the above, the description of FIG. 7 is completed, and an implementation example of the design support device 100 in the present embodiment will be described. Hereinafter, a design support system 1 in which the functions of the design support device 100 are implemented on a server will be described as an example. In this case, it can be realized by a so-called cloud system or on-premises, but the design support device 100 may be implemented on a terminal device other than the server.

[0061] FIG. 8 is a system configuration diagram of a design support system 1 which is an example of implementation in the present embodiment. The design support system 1 includes a design support apparatus 100, a database 200, and a terminal apparatus 300, which are connected to each other via a network 400. The design support apparatus 100 and the terminal apparatus 300 can be realized by computers that execute processes according to programs. However, they may be realized by dedicated hardware or the like instead of programs.

[0062] First, the design support apparatus 100 includes a communication device 11, a processing device 12, a main memory device 13, and an auxiliary storage device 14. These are connected to each other via an internal communication path such as a bus. Hereinafter, each component will be described.

[0063] First, the communication device 11 has a function of connecting various components of the design support apparatus 100 to other devices via the network 400. The network 400 only needs to have a communication function, and its type such as the Internet is not limited. And the communication device 11 can realize the functions of the input unit 101 and the output unit 106 in FIG. 1.

[0064] Also, the processing device 12 is realized by a processor such as a CPU. That is, the above-described analysis process is executed according to each program stored or expanded in the main memory device 13.

[0065] Also, the main memory device 13 is realized by a storage medium such as a memory. And the main memory device 13 stores programs stored in the auxiliary storage device 14 and expands each program for processing in the processing device 12. In this way, it is desirable that each program is usually stored in another storage device or storage medium such as the auxiliary storage device 14.

[0066] In addition, the auxiliary storage device 14 stores various types of information and programs. The various types of information include the CAD shape DB 108, the fastening attribute DB 109, and the failure probability DB 110 described in FIG. 1. Although not shown, the auxiliary storage device 14 may store the generated documents, disassembly paths, and evaluation values. Further, the various types of information may be stored in a database 200 other than the design support device 100. In FIG. 8, the CAD shape DB 108 is stored in the database 200, but it is not limited thereto. Also, the database 200 may be omitted.

[0067] Furthermore, the program stored in the auxiliary storage device 14 is the design support program 15. This is a program for executing the above-described various processes. Therefore, the design support program 15 includes a fastening shape recognition module 151, a key component selection module 152, a disassembly path generation module 153, an evaluation value calculation module 154, and a document generation module 155. As a result, the processing device 12 can execute the processes of the fastening shape recognition unit 102, the key component selection unit 103, the disassembly path generation unit 104, the evaluation value calculation unit 105, and the document generation unit 111 according to the design support program 15. Note that these modules may be realized by independent programs.

[0068] Also, the auxiliary storage device 14 can be realized by a hard disk drive (HDD), a solid state drive (SSD), various optical disks, or the like.

[0069] Next, the terminal device 300 is used by the user and has the functions of the input unit 101 and the output unit 106. The input unit 101 can be realized by an input device such as a keyboard or a pointing device, and accepts the user's operations. Also, the output unit 106 can be realized by a display device that displays the processing results of the processing device 12 such as documents.

[0070] With the above, the description of this embodiment ends. In this embodiment, since it is not necessary to define and associate work instructions such as disassembly procedures for various parts in advance, the man-hours for this part can be reduced. Also, as a method of shape recognition, by adopting a fuzzy method such as measuring the similarity to a template shape and statistically estimating part types, etc. from the shape, it is possible to estimate even for new parts and parts with design changes.

[0071] Moreover, according to this embodiment, not only parts but also recognition of partial shapes is carried out. Therefore, it is possible to recognize parts related to fastening such as snap fits included in the parts and evaluate the disassemblability. Also, in this embodiment, based on the fastening relationship and interference determination of the parts, a disassembly path is generated for each disassembly target part. Thus, a disassembly path can be obtained without manual order definition, and it is possible to output a disassembly path whose validity has been verified based on interference determination, etc.

[0072] Furthermore, in this embodiment, since a disassembly path is generated for each set disassembly target part, it is possible to generate and present a path such that the number of disassembly step procedures is reduced for each target part. Note that in this embodiment, the evaluation of easy disassembly necessary for evaluating the maintainability and repairability of the product is automated by the recognition of parts, the fastening method and disassembly method by partial shape recognition, and the generation of disassembly paths for each selected disassembly target part using the results. In order to evaluate KPIs such as disassembly difficulty based on the recognition results, the evaluation criteria for easy disassembly are standardized. The creation of documents such as disassembly manuals is supported based on the generated path. Furthermore, by using the recognition technology, the man-hours for preparations for automation such as data preparation of the pair of parts of the disassembly method and disassembly procedures can be reduced.

Explanation of Signs

[0073] 100… Design support device, 101… Input unit, 102… Fastening shape recognition unit, 103… Important part selection unit, 104… Disassembly path generation unit, 105… Evaluation value calculation unit, 106… Output unit, 107… Storage unit, 108… CAD shape DB, 109… Fastening attribute DB, 110… Failure probability DB, 111… Document generation unit

Claims

1. 1. A design support device for supporting the design of a design object, comprising: an input unit for receiving a 3D CAD model of the design object; a recognition unit that recognizes a fastening shape including a shape of a fastening part of a component constituting the design object, an attribute of the fastening part, and a fastening relationship between the component and other components from shape information included in the 3D CAD model; a component selection unit for selecting a component to be analyzed from among components constituting the design object; a path generating unit that generates a disassembly path including a disassembly method for the selected part by using the fastening relationship; an evaluation value calculation unit that calculates an evaluation value indicating an evaluation of a dismantling method included in the dismantling path by using an attribute of the fastening unit; The design support device has an output unit that outputs at least one of the disassembly path and the evaluation value.

2. 2. The design support apparatus according to claim 1, Further, a storage unit is provided for storing fastening attribute information indicating attributes of a fastening shape, The recognition unit is a design support device that recognizes the fastening relationship using a learning model that statistically processes the relationship between the fastening shape and fastening attribute information.

3. 2. The design support apparatus according to claim 1, The component selection unit is a design support device that selects a component to be analyzed based on a failure probability DB that includes at least one of the failure rate and the failure impact degree of the component that constitutes the design object.

4. 2. The design support apparatus according to claim 1, a document generation unit that generates a document including a disassembly procedure for the selected part based on the disassembly path, The output unit outputs the document, A design support device that is capable of presenting the document to a user of the design support device.

5. 2. The design support apparatus according to claim 1, Further, a storage unit is provided for storing sequence information indicating an assembly sequence and a disassembly sequence for each fastening shape model, The path generation unit is a design support device that generates the disassembly path including an assembly order of selected parts and an assembly path including a disassembly order using a learning model that has learned the relationship between the shape indicated by the fastening shape model and the order information.

6. 2. The design support apparatus according to claim 1, The path generating unit is a design support device that generates the dismantling path in accordance with at least one of a dismantling purpose and a dismantling subject of a selected part.

7. A design support method for supporting design of a design object by a design support apparatus, comprising: An input unit receives a 3D CAD model of the design object; A recognition unit recognizes a fastening shape including a shape of a fastening part of a part constituting the design object, an attribute of the fastening part, and a fastening relationship between the part and another part from shape information included in the 3D CAD model; A parts selection unit selects a part to be analyzed from the parts constituting the design object, A path generating unit generates a disassembly path including a disassembly method for the selected part using the fastening relationship; an evaluation value calculation unit calculates an evaluation value indicating an evaluation of a dismantling method included in the dismantling path by using an attribute of the fastening unit; An output unit outputs at least one of the disassembly path and the evaluation value.

8. The design support method according to claim 7, The design support method further includes a storage unit that stores fastening attribute information indicating attributes of a fastening shape, A design support method in which the recognition unit recognizes the fastening relationship using a learning model that statistically processes the relationship between the fastening shape and fastening attribute information.

9. The design support method according to claim 7, The design support method includes selecting a part to be analyzed based on a failure probability DB including at least one of the failure rate and the failure impact degree of the part constituting the design object by the part selection unit.

10. The design support method according to claim 7, Furthermore, a document generation unit generates a document including a disassembly procedure for the selected part based on the disassembly path; The output unit outputs the document, The design support method enables the document to be presented to a user of the design support device.

11. The design support method according to claim 7, The design support device further includes a storage unit configured to store sequence information indicating an assembly sequence and a disassembly sequence for each fastening shape model, A design support method in which the path generation unit generates the disassembly path including an assembly order of selected parts and an assembly path including a disassembly order using a learning model that has learned the relationship between the shape indicated by the fastening shape model and the order information.

12. The design support method according to claim 7, The design support method further comprises generating the dismantling path in accordance with at least one of a dismantling purpose and a dismantling subject of the selected part by the path generating unit.

Images