Patent Application
20240338014
This application claims priority under 35 U.S.C. § 119 or 365 European Patent Application No. 23305499.8 filed on Apr. 6, 2023. The entire contents of the above application are incorporated herein by reference.
The disclosure relates to the field of computer programs and systems, and more specifically to a method, system and program for selecting materials of components of a product to be manufactured.
A number of systems and programs are offered on the market for the design, the engineering and the manufacturing of objects. CAD is an acronym for Computer-Aided Design, e.g., it relates to software solutions for designing an object. CAE is an acronym for Computer-Aided Engineering, e.g., it relates to software solutions for simulating the physical behavior of a future product. CAM is an acronym for Computer-Aided Manufacturing, e.g., it relates to software solutions for defining manufacturing processes and operations. In such computer-aided design systems, the graphical user interface plays an important role as regards the efficiency of the technique. These techniques may be embedded within Product Lifecycle Management (PLM) systems. PLM refers to a business strategy that helps companies to share product data, apply common processes, and leverage corporate knowledge for the development of products from conception to the end of their life, across the concept of extended enterprise. The PLM solutions provided by Dassault Systemes (under the trademarks CATIA, ENOVIA and DELMIA) provide an Engineering Hub, which organizes product engineering knowledge, a Manufacturing Hub, which manages manufacturing engineering knowledge, and an Enterprise Hub which enables enterprise integrations and connections into both the Engineering and Manufacturing Hubs. All together the system delivers an open object model linking products, processes, resources to enable dynamic, knowledge-based product creation and decision support that drives optimized product definition, manufacturing preparation, production and service.
Within this context, there is a need for improved solutions for selecting materials for a product to be manufactured.
It is therefore provided a computer-implemented method for selecting materials of components of a product to be manufactured. The method comprises providing a set of materials for the components of the product. The method also comprises providing, for each component, a set of use and/or manufacturing and/or material constraints for the component. The method also comprises providing, for the product, a set of use and/or manufacturing constraints for the product. The method also comprises providing, for the product, one or more material compatibility constraints. The method also comprises providing specifications. The specifications indicate an extent of compatibility of one or more reference materials within the set of materials with the constraints. The method also comprises determining, for each component of the product, at least one optimal material, with respect to compatibility with the constraints and based on the provided specifications, for manufacturing the component.
The method may comprise one or more of the following features:
It is further provided a computer program comprising instructions for performing the method.
It is further provided a computer-readable data storage medium having recorded thereon the computer program.
It is further provided a computer system comprising a processor coupled to a memory (and optionally a graphical user interface), the memory having recorded thereon the computer program.
Non-limiting examples will now be described in reference to the accompanying drawings, where:
Described is a computer-implemented method for selecting materials of components of a product to be manufactured. The method comprises providing a set of materials for the components of the product. The method also comprises providing, for each component, a set of use and/or manufacturing and/or material constraints for the component. The method also comprises providing, for the product, a set of use and/or manufacturing constraints for the product. The method also comprises providing, for the product, one or more material compatibility constraints. The method also comprises providing specifications. The specifications indicate an extent of compatibility of one or more reference materials within the set of materials with the constraints. The method also comprises determining, for each component of the product, at least one optimal material, with respect to compatibility with the constraints and based on the provided specifications, for manufacturing the component.
This constitutes an improved selection for material selection for a product to be manufactured.
Indeed, given a product to be manufactured and formed by several components, and given an input set of materials (i.e., the provided set of materials) and constraints related to the product and its components (i.e., the provided constraints concerning manufacturing, use, and/or material), the method finds, for each component of the product, at least one (e.g., one) material of the set which is optimal, with respect to satisfaction of/compatibility with the constraints and the provided specifications, for manufacturing the component (that is, given the constraints and the provided specifications, the determined material is found to be the optimal one for manufacturing the component, and thus the product). This means that the output is made of materials (i.e., at least one material, for example one, for each component) selected within the provided set of materials and found to be optimal for manufacturing the components (i.e., with the determined optimal material for each component) and thus the product, where optimality is with respect to satisfaction of the provided constraints and specifications. The output materials may for example be or include a list of these materials, for example with each material associated to the component for which the material was found optimal, or a list of pairs each formed by one component and its associated optimal material(s). The method thus allows the objective selection of multiple materials for multiple components forming a product, the materials thereby selected being optimal, given the constraints, for manufacturing the components and thus the product.
The constraints provided by the method, which forms inputs of the method and of the determination of the optimal materials, comprise constraints related to the product and/or components thereof during use (referred to as use constraints), constraints related to the manufacturing of the product and/or components thereof (referred to as manufacturing constraints), and/or constraints related to materials forming the components (referred to as material constraints). The constraints also include constraints constraining compatibilities and incompatibilities of the materials of the components (referred to as material compatibility constraints). In other words, some materials of the provided set of materials are compatible with constraints (for example related use or manufacturing) of some component(s), while other materials are not but are compatible for other components. Some materials may satisfy constraints related to the product and other may not. Some materials may be incompatible for being part of the same product, even if individually at the components level, they are suited for some components. The method thus finds, within the provided materials, those which are accepted for manufacturing the components and thus the product given their compatibility with all the provided constraints.
As outlined above, the method considers constraints both at the level of the components and at the level of the product, that is constraints that concern the components individually and constraints that concern the whole product. The method notably accounts for material compatibility constraints. All of this improves the material selection, as materials may for example well be accepted individually for some components but found to be incompatible when used together in a same product. The method thus improves the selection of materials for manufacturing a product formed by several components, since considering the constraints at both the product level and the component level allows to obtain the optimal materials not only for the product, but for each component as well. The material selection provided by the method has thus improved reliability, which allows to manufacture the product (using the determined optimal materials) with improved reliability as well, and so that the reliability of the product is as well improved.
Finding the appropriate components materials given the constraints, that is finding the best components materials to meet the constraints, may in practice be complicated or even infeasible for the user/designer. Such selection problem may indeed in practice depend on many parameters (e.g., best theoretical material vs material availability in practice, compatibility of materials with the available manufacturing machines), and thus may require performing an extensive objective classification task which is not feasible by hand or mentally. The method however objectively finds these components materials which are optimal with respect to compatibility with the constraints, and thereby credibly and objectively assists the designer for selecting the appropriate materials for manufacturing the components and thus the product given the constraints.
Furthermore, the determination of the optimal components materials is based on the provided specifications, which indicate an extent of compatibility of one or more reference materials (i.e., which are within the provided set of materials) with the constraints. In other words, the provided specifications guide the determination of the optimal components materials by providing reference data regarding the materials and constraints compatibility. This allows a user, or several users, e.g., the designer(s)/engineer(s) that work on the product, to provide their specifications, thereby feeding the method with objective guiding data for selecting the best materials. The method determines the optimal materials that meets at least to some extent these user-provided inputs. The method thus provides a solution to objectively and credibly assist and guide a designer or a group of designers through the components material selection problem for manufacturing the product.
The output of the method comprises the materials determined as optimal for manufacturing the components and thus the product, for example a list of these materials in association with their corresponding components or a list of pairs components-associated determined optimal material(s), as discussed above. These materials may for example by outputted with a label “accepted”. The output of the method is thus data describing the components materials determined as optimal for manufacturing the components. The output of the method may thus be used for manufacturing the components with their optimal materials, and thus for manufacturing the product.
The output of the method thus forms the basis of an implied further use of the output, which is the real-world manufacturing of the components, and thus of the product, using the determined optimal materials, where each component is manufactured according to its corresponding determined optimal material. The method may thus form or be part of a design and/or manufacturing process. The method may for example form a material selection step that is performed after a design process and before a manufacturing process where manufacturing is performed in accordance with the optimal materials determined by the method. The method may for example further comprise (i.e., further to the determination of the optimal materials) using the determined optimal materials for manufacturing the components (and optionally the product), i.e., where for each component, the material determined by the method to be optimal for the component is used for manufacturing the component, and optionally where the components are suitably assembled to form the product. Using the optimal materials for manufacturing the product designates any actions or series of actions that is/are involved in/participate to the manufacturing of the product via the manufacturing of its components using the optimal materials. Using the optimal materials for manufacturing the components may comprise one or more of the following:
The method is for selecting materials of components of a product to be manufactured. This means that the method is a method for performing a selection of materials so that, for each component of the product (which comprises several components), at least one material is selected for the component. As previously discussed, this materials selection may be a step of a design and/or manufacturing process, e.g., may follow a design step of the process and/or precede a manufacturing step of the process.
Materials selection relates to the selection the most appropriate materials for engineering applications. Material selection is an important function in the design and development of products. Indeed, several dimensions of the design problem are influenced by the used materials including the product central function, the customer satisfaction, the production systems used for manufacturing, the product life cycle, who is going to use or produce it, the usability, and finally the costs (as discussed in Ali Jahan, Kevin L. Edwards, Marjan Bahraminasab, “Materials selection in the context of design problem-solving”, Butterworth-Heinemann, 2016, Pages 25-40, which is incorporated herein by reference). Materials selection can be carried out either to choose alternative materials for changes to the design of an existing product in order to reduce say cost or weight, meet new legal requirements, overcome failure occurrence, and satisfy different market demands or to choose materials for the design of a completely new product. As stated in Jahan, Ali, and Kevin L Edwards, “Multi-Criteria Decision Analysis for Supporting the Selection of Engineering Materials in Product Design”, Oxford: Elsevier Science & Technology, 2013. Web, which is incorporated herein by reference, it has been estimated that the relative cost of a design changes after manufacture is 10,000 times more than at the conceptual stage of design, thus, an intelligent decision support tool such as the method of the present disclosure is relevant for materials selection in order to reduce costs. Material selection within the product design process may be either cost driven or performance driven. For example, a performance-driven could be a bespoke biomedical prosthesis/implant or a tennis racquet and an example of cost-driven could be a mass-produced motorcar or beverage can/bottle. Furthermore, when it comes to a multi-component product to be manufactured, such as the one contemplated in the method, changing the material of one component may entail several changes in the neighboring components. For example, replacing steel with aluminum may cause galvanic corrosion with the neighboring steel components (as discussed in Farag, M. M., 2002, “Quantitative methods of materials selection”. In: Kutz, M. (Ed.), Handbook of Materials Selection. London, John Wiley & Sons, which is incorporated herein by reference). A component-by-component materials selection procedure might lead to a suboptimal solution, therefore, selecting the optimal material mix for the different components of a product is not trivial (as discussed in reference Leite, M., Silva, A., Henriques, E. et al., “Materials selection for a set of multiple parts considering manufacturing costs and weight reduction with structural isoperformance using direct multisearch optimization”, Struct Multidisc Optim 52, 635-644 (2015), which is incorporated herein by reference).
The method comprises providing, as inputs to the determining step, the set of materials, the constraints (for each component and for the product), and the specifications. These inputs are now discussed.
The provided set of materials consists in data (i.e., numerical data) describing a set of one or more real-world materials. In other words, the provided set of materials is a numerical set of materials, that is a numerical description of the real-world set of materials. For the sake of efficiency, the set of one or more real-world materials and the corresponding provided (numerical) set of materials may be used interchangeably in this disclosure. The set of materials may comprise available and/or usable materials for a factory, or several factories, in charge with the manufacturing of the product and/or its components and/or materials known to the designer(s)/engineer(s) using the method and/or other materials known in the industry (for example stemming from a material database). The set of materials may be a list of materials, e.g., a numbered list of materials. The set of materials may be provided by one or more users, e.g., one or more designers/engineers using the method. Providing the set of materials may for example comprise, by the user(s), selecting the materials in a material database. The method may alternatively provide automatically a fixed (non-modifiable) set of materials, e.g., that stems from a material database. Yet alternatively, the method may provide automatically a fixed subset of materials which a user, or several, completes with its (resp. their) own material selection. Providing the set of materials may comprise, providing, for each material and each constraint, a score (also referred to as “evaluation”) of the material on the constraint. This may be done by one or more users, or the materials may be provided automatically with these scores fixed and non-modifiable (e.g., the scores may stem from a material database). The provided set of materials may be a large dataset. As mentioned in references Dieter, G. E., 1997, “Overview of the Materials Selection Process”, ASM Metals Handbook, Materials Selection and Design. ASM International, Materials Park, O H, and Jahan, Ali & Edwards, Kevin & Bahraminasab, Marjan, (2016), “The importance of decision support in materials selection”, 10.1016/B978-0-08-100536-1.00001-1, which are both incorporated herein by reference, there are an enormous number of materials available, each with a range of different properties and behaviors. In case where a user oriented support system is provided, a real time selection is demanding, which makes the computation time demanding.
The constraints include constraints related to the component (which may be referred to as “component-oriented constraints”) and constraints related to the product (which may be referred to as “product-oriented constraints”). The constraints may also be referred to as “requirements”.
For each component, the constraints include a set of use and/or manufacturing and/or material constraints (i.e., a set of one or more of such constraints) for the component. A use constraint for a component is a constraint during use of the component once fabricated, such as a certain required stiffness, or a certain required heat conduction capacity. A manufacturing constraint for a component is a constraint related to the fabrication of the component, such as a certain ability to be bent or a certain drying time. A material constraint for a component is a constraint on the material to be used for the component, such as a mandatory choice of material or type of material (e.g., metal) or a desired material property (e.g., being waterproof). The constraints may also be referred to as “attributes”. The set of constraints may be divided into rigid constraints, which necessarily have to be met by the optimal material for the component, and soft constraints, which do not have to be necessarily met by the optimal material. The determination of the optimal materials by the method may require satisfaction of the rigid constraints and maximization of the satisfaction of the soft constraints. The set of constraints may further include one or more economic constraints. In examples, the set of constraints does not include these economic constraints.
For the product, the constraints include a set of use and/or manufacturing constraints for the product (i.e., a set of one or more of such constraints). A use constraint for a product is a constraint during use of the product once fabricated, such as a certain required stiffness, or a certain required heat conduction capacity. A manufacturing constraint for a product is a constraint on/related to a fabrication of the product. The constraints may also be referred to as “attributes”. The set of constraints may be divided into rigid constraints, which necessarily have to be met by the materials of the component forming the product, and soft constraints, which do not have to be necessarily met by these materials. The determination of the optimal materials by the method may require satisfaction of the rigid constraints and maximization of the satisfaction of the soft constraints. The set of constraints may further include one or more economic constraints. In examples, the set of constraints does not include these economic constraints.
For the product, the constraints further include one or more material compatibility constraints. A material compatibility constraint is a constraint that describes that two (or more) materials must or can, or must not or cannot, be used for manufacturing two (or more) components of the product and/or two (or more) interacting components. Such constraint thus enforces rejection or the acceptation of such combination of materials for these components of the product by the step of determining the optimal materials. Such constraint may be specific to the product (i.e., the combination is excluded or accepted for reasons, e.g., use or manufacturing reasons, related to the product) or specific to the materials (i.e., the combination is simply excluded or accepted per se because the materials can/must or cannot/must not be used together and/or in proximity to one another). The material compatibility constraint(s) may include or consist in a compatibility matrix between materials and the product constraints (e.g., budget constraint, interdependencies between materials, and/or logic constraints). An example of two materials that are not compatible in a knee prosthesis is the use of aluminum and titanium in the articulating components. While both metals are commonly used in prosthetic manufacturing, the use of aluminum and titanium in direct contact in the articulating components can cause a galvanic reaction, leading to corrosion.
One or more constraints (i.e., for the components and/or for the product) may be latent constraints. This means that the method does not know in advance whether such a constraint is to be maximized, minimized, or minimized below or above a peak, or maximized above or below a peak. The latent constraints may include constraints of one or more of the following three types of latent constraints (also referred to as latent requirements preferences): monotonous latent constraints, single-peaked constraints, and single-valley constraints. Monotone latent constraint is the case where the preference directions (maximization or minimization) on the constraint are not known (e.g., optical properties). Single-peaked constraint is interpreted as a constraint to be maximized bellow a peak a and as a constraint to be minimized above the peak α. Single-valley constraint can be interpreted as a constraint to be minimized bellow a peak P and as a constraint to be maximized above the peak (e.g., thermal and radiation properties). The method handles the latent constraints through the determination of the optimal materials using the provided specifications: the method determines the optimal materials using the provided specifications and, while determining the optimal materials, finds for each latent constraint, whether the constraint is to be maximized, minimized, or minimized below or above a peak, or maximized above or below a peak. The method may for that learn one or more NCS models as discussed hereinafter, wherein the method learns the specificities of the latent constraints while the method learns the parameters of the NCS model(s). An example of a relevant criterion in material selection for which it is not clear whether to maximize or minimize is the hardness of the material. Hardness is an important physical property that can affect the strength and durability of a material. In the manufacture of dental prostheses, hardness is an important criterion for the strength and durability of the material. However, excessive hardness can also make the material difficult to work with and can cause damage to the tools used to machine it.
The set of constraints may include numerical constraints, which express the absolute measure of a material property or behavior with respect to a component. Examples of numerical requirements are density, modulus, strength and thermal conductivity. The set of constraints may additionally or alternatively include nonnumerical constraints, which regroup material properties that are not expressed numerically such that resistance to corrosion which can be represented as linguistic variable (e.g., poor, adequate, good). The method thus handles both constraints.
The set of constraints (i.e., for the components and/or for the product) may be provided by one or more users of the method, e.g., one or more designers or engineering. For example, the user(s) may select the provided constraints by graphical interaction with a displayed list of constraints, for example by checking (e.g., with a mouse click or a touch) displayed checkboxes each corresponding to respective constraints. This may include selecting one or more latent constraints, i.e., constraints selected without knowing whether they are to be minimized or maximized for example. The method may comprise, further to the providing of the constraints, displaying the evaluations/scores of the materials on the provided constraints, for example as an array-like/table display of all the scores (which may be referred to as a “the performance table”), with the materials as lines (resp. columns) and the constraints as columns (resp. lines). There may be one of such table per component and one for the product, or one table for all the constraints of the product and all the components. In examples at least some scores may be modified by the user(s). In other examples, the scores are fixed and non-modifiable.
The inputs also include the specifications, which may also be referred to as “preferences”. The specifications indicate an extent of compatibility of one or more reference materials with the constraints. The specifications thus form data indicating how compatible is each reference material with the constraints. The specifications may comprise for example a list of the reference materials, and for each material, an overall score/classification of compatibility with the all the constraints, for example among the scores/classifications “Accepted” (which means that the material is considered compatible with the constraints) and “Rejected” (which means that the material is considered not compatible with the constraints. Each score/classification may be a string representation of one of these scores “Accepted” and “Rejected”. The specifications may alternatively comprise solely a list of one or more reference materials considered as compatible with the constraints. For each reference material, or at least a part thereof, the score/classification (for example “Accepted” or “Rejected”) may be associated with a component or a list of components, which means that the score is only relative to this component or list of components (e.g., the material is accepted or rejected for a given component only or for some given components only). This means that the preferences may concern the product, the components or at least some of them, or both. The reference materials are material within the provided set of materials, i.e., they belong to the provided set of materials. This means that, for at least some materials of the provided set, i.e., the reference materials, the determination of the optimal materials may use both the individuals scores of the materials on the constraints and the overall score/classification (e.g., “Accepted” or “Rejected”) of compatibility of the material with respect to all the constraints.
The specifications may be provided by one or more users, e.g., one or more designers/engineers using the method. This allows a user, or more, to guide the method towards the determination of optimal components materials that meets or tends to meet, or at least accounts for, their expertise regarding compatibility with the constraints and regarding the manufacturing. This guides the method towards achieving an even more meaningful result, and thus makes the method in the meantime a solution to credibly assist and guide the user(s) for material selection for manufacturing: the method responds to the user specifications by providing optimal components materials that accounts for their expertise. The user(s) may also thereby react to what the method provides, and then provide different input specifications to get different results that is more suited to their technical expertise and expectations. The method thus forms a solution for a continuous and guided user-machine interaction process for assisting the user in the material selection problem for a multi-component product to be manufactured.
The inputs have been discussed. For any input herein (i.e., constraint or material or specification), providing the input may be carried out by a user (or more) defining the input (e.g., defining a constraint, selecting a material, defining a specification, as previously discussed). Alternatively, providing the input may comprise obtaining/retrieving (e.g., downloading) the input, for example from a (e.g., distant) memory or database or server.
Further to the providing of the inputs, the method comprises determining, for each component of the product, at least one optimal material with respect to compatibility with the constraints and based on the provided specifications, for manufacturing the component. This means that the method processes the provided inputs to output, using the providing constraints, the materials found to be optimal, with respect to compatibility with/satisfaction of the constraints, for manufacturing the components. The determining of these optimal materials for the components is now discussed.
The determining of the optimal materials searches, within the provided set of components, and for each component, a material (e.g., one material) that is optimal, with respect for manufacturing the component. This may consist in processing the provided scores of each material on the constraints, as previously discussed, and the provided specifications (for example specifying the reference materials as compatible/accepted or non-compatible/rejected) to extract from the provided set of material the optimal combination of materials where each material is assigned to, and optimal (with respect to the constraints) for a component, and the combination is altogether optimal with respect to the constraints. The specifications may thus guide the processing of the material scores, the specification serving as reference of which material is compatible and which are not, i.e., serving as a reference to learn from. Processing the scores using the specifications allows the method to understand which scores on which constraints provide for a classification as compatible/accepted and which provide for a classification as non-compatible/rejected. In other words the determination of the optimal materials aims at both satisfying the constraints and following the provided specifications.
The determining of the optimal materials may be carried out automatically further to the providing of the inputs. For example, the user(s) using the method may, further to the providing of the input, launch the automatic determining of the optimal materials, e.g., by clicking on or touching a displayed icon adapted for this launch. The determining of the optimal materials may in examples proceed in two steps: first, the materials are classified as accepted or rejected with respect to each component, i.e., a determination is made at the component level, and then only those materials classified as accepted are considered in a second, product level, determination which finds those accepted materials which are optimal for the product constraints. These two steps may be carried out sequentially (for example by learning two NCS models as discussed hereinafter) or in a same determination process (for example using a single NCS model). The determining of the optimal materials may be performed several times, for example with a partial modification of the inputs between each determining. This may lead to more than one material considered to be optimal, at least for some component(s).
The determining of the optimal materials for the components may comprise learning a Multiple Criteria Decision Aiding sorting model. The model is leaned to (i.e., configured to) take as input the set of materials and to output the optimal materials for the components. The learning is based on the set of materials, on the constraints, and on the specifications. The concept of Multiple Criteria Decision Aiding sorting model is known per se. The sorting model comprises parameters which are inferred through the learning. The learning in other words comprises inferring the parameters, this inference being the result of the optimal components materials determination. In other words, the learning infers the parameters through the optimal materials determination that is done for determining the optimal materials, where the specifications guide the learning. The model, once learnt, is configured to take as input the set of materials and to output the optimal materials for the components.
Multi-criteria decision-Aiding (MCDA) aims at developing decision-support models explicitly based on the construction of a set of criteria reflecting the relevant aspects of the decision-making problem. These n criteria (N={1, 2 . . . , n} with n≥2) evaluate a set of alternatives A={a, b, c, . . . } under consideration with respect to different viewpoints. The purpose of MCDA methodologies is to support the decision maker (DM) by providing methods and frameworks to take decisions regarding the considered decision situation. Decision problems considered in MCDA are of different types and three reference problems are encountered in practice (as discussed in reference B. Roy, “Multicriteria Methodology for Decision Aiding”, Kluwer Academic, Dordrecht, 1996, which is incorporated herein by reference):
The specifications may form a learning set of the model, also referred to as a set of determined choices, which is data that the parameters must match, or at least tend to match, when updated. In other words, the learning may comprise inferring the parameters from the specifications so that the inferred parameters are consistent with the specifications. It may be in other words said that the learning extends the specifications, since ultimately the learned model (i.e., the model with its parameter inferred) takes as input the set of materials and outputs the components optimal (i.e., with respect to compatibility with the constraints) materials therein and that are altogether consistent with the provided specifications (i.e., the compatibility of the optimal materials with the constraints is consistent with the compatibility data provided by the specifications). In yet other words, the learning of the model takes as input a set of assignment examples (i.e., the specifications), and learns a model that is consistent with this preference data. In yet other words, the learning learns the model parameters that match (e.g., perfectly) the assignment examples.
The model may include one or more Non-compensatory Sorting models, also referred to as “NCS” model. For example, the method may learn a single NCS model that performs the determination of the optimal materials both at the component level and at the product level, as previously discussed. Alternatively, the method may learn two models, one for each level. The learning of the model may thus include learning these one or more NCS models. The NCS model(s) take as input the set of materials and outputs the optimal materials. In other words, the NCS model outputs the materials found to be optimal with respect to their compatibility with the constraints. The learning of the NCS model may comprise inferring the parameters of the NCS model based on the constraints and using the specifications, so that the NCS model parameters are such that the NCS model extends the specifications as previously discussed. For example, the learning may search for parameters of the NCS model that best match the specifications. The specifications may thus form a learning set of the NCS model. In other words, the output of the model is consistent with the specifications.
Learning a NCS model allows to consider interactions between the constraints, which may be dependent from each other, and allows to account for the alternative evaluations in an ordinal perspective, thereby avoiding compensation and enabling to deal meaningfully with the constraints. In real-world decision-aiding situations and in particular in the material selection problem, interactions between requirements/constraints may occur as there may be requirements conflicting. The concept of NCS model is known and discussed in references Bouyssou, D., Marchant, T., “An axiomatic approach to noncompensatory sorting methods in MCDM, I: the case of two categories”, European Journal of Operational Research 178(1), 217-245 (2007), and Bouyssou, D., Marchant, T., “An axiomatic approach to noncompensatory sorting methods in MCDM, II: more than two categories”, European Journal of Operational Research 178(1), 246-276 (2007), which are both incorporated herein by reference. NCS originates from the Electre TRI method discussed in reference Yu, W., “Aide multicritère à la décision dans le cadre de la problimatique du tri: methodes et applications”, Ph.D. thesis, LAMSADE, Universite Paris Dauphine, Paris (1992), which is incorporated herein by reference.
The NCS model of the present disclosure and its learning may be implemented as follows (in other words implementations of the learning of the NCS model are now discussed). As previously discussed, the method in these implementations learns to assign the materials in the categories Accepted and Rejected. NCS may be formulated according to the following: a material is assigned to the category Accepted if:
Given the input set of materials A, each material a∈A may be evaluated according to a finite set of properties (constraints) N with n=|N|, let (a1, a2, . . . , an) denote the evaluation vector of a∈A with aj denoting its evaluation on property (constraint) j. The providing of the set of materials may comprise providing the set already evaluated, or providing the evaluations besides providing the set, as previously discussed. Without loss of generality, it is assumed that the preference on each property increases with the evaluation. Considering both predefined and ordered categories Rejected Accepted, multiple criteria sorting aims at assigning materials to one and only one category.
The assignment rule of the Non-Compensatory Sorting model operates on two types of parameters:
The classification rule of the NCS model is defined as follows:
∀a∈A, a∈Accepted⇔{j∈N:aj≥bj}∈T
The method may implement a learning paradigm in which a set of reference materials assignments (the learning set, i.e., the provided specifications) is provided by the product designer; the aim is to extend this learning set using an NCS model. To do so, the learning of the NCS model searched for the NCS parameters that best match the learning set. The method may, at the providing of the specifications, collect explicitly or implicitly these reference assignments. Explicitly means for example that the method comprises asking the product designer to classify a subset of materials in categories RejectedAccepted, and sets this set as the learning set. Implicitly means for example that the method may consider a history of examples for which previous decisions have been taken and validated by the product designer or product designers. Consequently, the method sets this history as the learning set. In some situations, the learning set may be inconsistent (i.e., the product designer preferences cannot be restored by some multicriteria sorting model). To cope with this, a preliminary phase (interactive procedure for building consensus) may be applied to revise the learning set, as discussed hereinafter. As previously discussed, the method focuses on the evaluation of materials on latent, single peaked and single valley properties. Thus, when learning the NCS model that evaluates the set of materials, the method handles these types of properties.
In implementations, determining/selecting the optimal components materials within the provided set of materials may be based on a selection strategy process that takes into account four aspects:
As discussed previously, materials evaluation and, more precisely, the distinction between “Accepted” and “Rejected” materials may be performed by classifying them into two ordered categories Accepted and Rejected using an NCS model. The NCS model parameters may be learned according to the product designers' preferences (learning set) by mobilizing an efficient SAT/MaxSAT-based learning formulation, which is discussed hereinafter, and the learning may evaluate the quality of the models with respect to the compatibility with the learning set.
Since the materials that constitute the product must be evaluated as “Accepted” according to component-oriented properties, they must be selected from the category Accepted after classifying all items by the learned NCS model. In some situations, the learning phase could end up with several compatible NCS models. To obtain the best mix of materials, choosing the best model issued from the learning step may not always be guaranteed. To avoid a complete enumeration of all compatible NCS models, the method formalizes the task of identifying the best mix of materials as an integrated task within the learning problem, i.e., learning a constrained NCS model.
For that, the method integrates in implementations the product-oriented properties and constraints into the sorting model learning formulation. In other terms, the generation of the final solution is coupled with the learning of the compatible NCS model, which ensures that all the materials that constitute the selected solution are considered as “accepted” by the same compatible sorting model. If the method cannot output a compatible model, the method may comprise performing a preliminary phase of reviewing the learning set as discussed hereinafter. At the end, the learning formulation returns a compatible sorting model used to classify the materials into Accepted and Rejected categories, and therefore the preferred solution is represented by Accepted category. A formal representation of these implementations is now detailed.
Implementations of the learning of the NCS model have been discussed. Going back to the method (i.e., not necessarily implementations thereof as just discussed), the learning may comprise encoding learning clauses based on the constraints. The encoding uses a SAT-based encoding (or MaxSAT as discussed hereinafter). SAT-based encoding is known per se. Implementations of the SAT encoding are discussed hereinbelow.
A SAT instance is a Boolean formula that has three components; a set of variables (x1, x2, . . . , xn), a set of literals, each literal is a variable x or its negation −x, and a set of distinct clauses (C1, C2, . . . , Cm). Each clause consists of only literals combined with logical “or” (V) connectives. The goal of the satisfiability problem is to determine whether there exists an assignment of truth values to variables (x1, x2, . . . , xn), that makes the following Conjunctive Normal Form (CNF) formula satisfiable: C1∧C2∧ . . . ∧Cm, where ∧ is a logical “and”. In some situations, satisfying all the constraints is not possible. Thus, it is aimed at finding an assignment to satisfy the most of the constraints (clauses), i.e., resolving the Maximum Satisfiability (MaxSAT) problem. Particularly, people are often interested in solving Partial MaxSAT, where the constraints are divided into hard and soft constraints. The goal is to satisfy all hard constraints and maximize the number (or the total weight) of satisfied soft constraints. To distinguish between hard and soft clauses, let wmax denote the weight associated with each hard clause and therefore, it is considered clauses associated with a weight less than wmax as soft clauses. Many real world problems can be modeled as MaxSAT problems naturally.
The specifications may comprise incompatible specifications. In this case the learning includes finding a compromise between the incompatible specifications. The incompatible specifications may be provided by different users. This may comprise iteratively modifying the learning set until reaching an extent of compatibility between the specifications. This may be performed using an interactive procedure comprising iterations of (until a consensus is reached): asking the users for their preferences (e.g., reference materials assignments to “accepted” or “rejected”), implementing a voting rule to remove incompatible preferences, and ask the users for their agreement on the result of the voting rule. The interactive procedure may be that which is discussed hereinbelow. Still in this case, the learning may comprise (instead of or in addition to the interactive procedure) encoding learning clauses based on the constraints and the encoding may use a MaxSAT-based encoding.
Thus, to handle incompatible specifications, MaxSAT-based encoding and/or said interactive procedure may be performed. In implementations thereof, which may be combined with the above-discussed implementations (N.B. all implementations discussed in the present disclosure are combinable), the learning formulation is expressed in SAT/MaxSAT language, and to study the compatibility of the preferences of product designers with the problem requirements, the method may first perform any suitable satisfiability test (i.e., to check whether the provided specifications are compatible, or to which extent (e.g., percentage) they are). If the result is satisfying (SAT) and therefore a collective NCS model can be deduced, this model is learnt to select for each component the appropriate material. In the case where the result is not satisfying (UNSAT), the interactive procedure is performed to revise the specifications in order to learn collaboratively a collective NCS model.
Implementations of the interactive procedure, which are combinable with all the above-discussed implementations, are now discussed. When multiple product designers are involved in the decision process, a set of reference materials assignments (specifications) Ai* is provided by each one and the objective is to learn a collective constrained NCS model aligned with all the product designers' points of view. Putting all the learning sets together in an aggregated learning set A* may add inconsistency and therefore it may be that no NCS model can be elicited. In this case, the method may perform the interactive procedure to revise the learning set A* by proposing at each iteration to the product designers to agree on the assignment of a reference material among those which generate the inconsistency. The procedure seeks to reach a consensus between decision makers and derive a collective NCS model. The procedure may be as follows (as illustrated on
Step 1: Each product designer provides a learning set. The product designers may have given their opinion on the same material or on different ones. In the first case, the procedure constructs the collective learning set by assigning each material to the winning category according to the chosen voting rule. In the second case, the collective learning set is simply the union of the individual ones. The iterative procedure takes as input the collective learning set constructed as before.
Step 2: The procedure checks whether there is a model that restores all the reference materials of the collective learning set. If this is the case, the procedure directly deduces a compatible NCS model. Otherwise, the procedure goes to the next step.
Step 3: The procedure generates all minimal UNSAT sets (MUS): A MUS is a set of unrestored reference materials of minimal cardinality.
Step 4: The procedure evaluates all the reference materials present in at least one MUS with an appropriate indicator in order to determine the reference material which collective assignment will be modified at this iteration.
Step 5: The procedure generates the set of categories possibly assigned to the reference material chosen in Step 4 so as not to generate any additional conflict.
Step 6: By taking into account all the possible categories, a chosen voting rule, the relative role of the product designers as well as the preferred assignments of the product designers for this reference material, the new assignment is determined and the procedure updates the collective learning set.
Step 7: The procedure checks if the new learning set is satisfiable. If this is the case, the procedure deduces the collective NCS model. If this is not the case, the procedure goes back to Step 2.
Implementations of the SAT/MaxSAT learning formulation which is used to learn a constrained NCS model from the component-oriented and product-oriented requirements and from the specifications are now discussed.
Variables: Learning the collective NCS model involves 3 family of variables:
The implementations consider two types of product-oriented rigid requirements that are assumed to be generic. The first is the budget constraint, which encompasses cardinality constraints and pseudo-Boolean constraints and represented by a tuple <gc,ϕ(gc), σ, , representing the computation rule defined by the product designers to aggregate the individual values of items on the criterion gc, and σ (respectively
ϕ(gc) is the scalarization function defined by the product designers to evaluate the portfolio (i.e., the set of materials for all the components) as a whole. It is assumed here that this function is a linear combination of binary variables xk, ∀k∈A, i.e., the associated portfolio constraint is expressed as follows; σ≤Σk∈Awkxk≤
SAT encoding of both types of constraints results in a set of clauses in the conjunctive normal form, which operates on additional binary variables.
Given a set of items A such that n=|A|, a vector of weights <wi>i∈{1, . . . , n} and an upper bound σ∈, the boolean function ϕLC is defined with variables <xi>i∈{1, . . . ,n} and <Ri,k>i∈{1, . . . ,n-1},k∈{1, . . . ,σ}, as the conjunction of clauses:
All clauses of the formulation ϕLC, associated to the weight wmax.
The second product-oriented rigid requirement considered in this disclosure is the material compatibility matrix. Indeed, during manufacturing of a multi-component product, the materials that make up the components must be compatible with each other. To do this, given for each component i a set of materials candidates MCi, the compatibility matrix CM=(cm1≤a≤Σ
The compatibility matrix is implied in the learning formulation via the following clauses:
Clauses ϕCM are hard, associated to the weight wmax.
An example of application of the above-discussed implementations of the method for material selection of a multi-components knee prothesis is now discussed, with reference to
The application of the method to this knee prothesis example follows the followings steps, with the ultimate purpose of manufacturing the prothesis.
Step a)—Criteria/Constraints Selection
The user(s) (designer(s)) starts by providing the constraints/requirements, by choosing the component-oriented and the product-oriented rigid and soft requirements/constraints. This may be done by the user(s) interacting with the user-interface illustrated on
Step b)—Performance Table
Next, the method comprises providing the set of materials by displaying, for each component, the set of materials with for each material its evaluation/score on each constraint associated for the component. This may be done as illustrated on
Step c)—Construction of an Evaluation Model on the Candidate Materials
At this step, the method aims at building an evaluation model based on the component-oriented and the product-oriented requirements. The construction of this evaluation model involves in the discussed example three designers and the objective is to learn the model from their preferences (provided specifications). The profiles of the three designers may be displayed on the interface, by selecting (e.g., by graphical user interaction) the “consensus building” tab, as illustrated on
After the preference elicitation phase, a collective model is constructed from the preferences provided by the three designers. At this stage, a compatibility test is performed. In the case where the preferences of the three designers are consistent, a learning formulation, which involves the product-oriented requirements, is used to generate a compatible evaluation model. Otherwise, the method proceeds interactively with the three designers to revise their preferences in order to build a consensus.
The interactive procedure used to revise the designers' preferences starts by generating reference materials that represent the inconsistency source. Once these materials are generated, in each iteration, the procedure randomly takes a material and generates the associated compatible evaluations, which removes the associated inconsistency. This is illustrated on
The interactive procedure for preference revision process is repeated until the preferences of the designers become consistent. In this case, a collective evaluation model can be learned from these preferences and this model is used to select for each component of the knee prosthesis the most suitable material. In the last stage, the selected materials are displayed, as illustrated on
As illustrated by the above example, the implementations of the method provide the following advantages, which are now discussed.
Given a manufacturing product consisting of a set of components, several materials can be used to design each component. In the implementations, a two-level selection strategy is proposed; a component level and a product level. In the component level, it is searched for intrinsically evaluating the candidate materials for each component based on a set of component-oriented rigid requirements (e.g., geometry, external forces, properties of stock material before processing), component-oriented soft requirements (e.g., mechanical properties, physical properties, economic criteria), a set of materials properties (constraints), and the preferences of the product designers. The idea is to ensure the feasibility of the decision as well as the gain of the product designers' confidence. Moreover, in the product level, the implementations choose the mix of materials for all the components of the manufacturing product, while optimizing a set of soft and rigid criteria called product-oriented criteria and respecting the product designers' preferences.
In the implementations, a two-level selection strategy is thus proposed to handle the materials selection problem for a multi-component manufacturing product. The selection process is formulated as a multicriteria constrained sorting problem i.e., the objective is to select the mix of materials compatible with each other, which respect component-oriented requirements and product-oriented requirements. It is ensured that the selected materials have the best performance at the component level and at the product level. It is proposed to use a constrained NCS model to assign the materials into two ordered categories Accepted/Rejected and with the consideration of all the dimensions of the problem defined above, the materials will be used to design the manufacturing product will be given by the category Accepted. The used NCS model is learned interactively according to the preferences of a group of product designers. Therefore, it is developed an innovative elicitation procedure for several product designers (a decision committee) allowing each member of the committee to provide individual preference information in order to build a collective preference model accepted by each product designer. This work contributes to bring together recent advances in both multicriteria decision analysis and social choice domains and provide multicriteria multifactor decision support solutions, thereby the hybridization of these disciplines.
The implementations enable the combination of both evaluation systems i.e., component-oriented evaluation and product-oriented evaluation with the fact considering them in the same devise. This allows to select the mix of the best materials intrinsically and as a whole, and respecting all the constraints of the problem. Moreover, considering the product designers preferences when constructing the two evaluation models ensures the feasibility of the decision as well as the gain of the product designers' confidence and avoids the black box problem. Decisions made in this type of applications (material selection) are not generally the responsibility of a unique decision maker but a group of decision makers. The implementations account for that by providing decision support tools that involve a committee of product designers. As regards material qualifications, the implementations provide a solution that handles different types of evaluations such as latent monotone evaluations, single peaked and single valley evaluations. The consideration of these types of criteria in these implementations is coupled with the elicitation of its patterns e.g., it is learned an evaluation system for materials and the parameters of these criteria simultaneously, which may be useful for a future use of the solution. In the implementations, it is adopted the sorting method NCS to evaluate materials according to component-oriented and product-oriented properties. With NCS model, interactions between evaluation properties are considered and one of its specificities is to account for the alternative evaluations in an ordinal perspective avoiding compensation and enabling to deal meaningfully with qualitative data. Finally, the provided SAT/MaxSAT based solution through the implementations is investigated to be computationally efficient and eventually be able to solve problems with large datasets.
The method is computer-implemented. This means that steps (or substantially all the steps) of the method are executed by at least one computer, or any system alike. Thus, steps of the method are performed by the computer, possibly fully automatically, or, semi-automatically. In examples, the triggering of at least some of the steps of the method may be performed through user-computer interaction. The level of user-computer interaction required may depend on the level of automatism foreseen and put in balance with the need to implement user's wishes. In examples, this level may be user-defined and/or pre-defined.
A typical example of computer-implementation of a method is to perform the method with a system adapted for this purpose. The system may comprise a processor coupled to a memory and a graphical user interface (GUI), the memory having recorded thereon a computer program comprising instructions for performing the method. The memory may also store a database. The memory is any hardware adapted for such storage, possibly comprising several physical distinct parts (e.g., one for the program, and possibly one for the database).
The product to be manufactured contemplated in the method may a (e.g., mechanical) part or assembly of parts (or equivalently an assembly of parts, as the assembly of parts may be seen as a part itself from the point of view of the method), or more generally any rigid body assembly (e.g., a mobile mechanism). The product may be a product from various and unlimited industrial fields, including: aerospace, architecture, construction, consumer goods, high-tech devices, industrial equipment, transportation, marine, and/or offshore oil/gas production or transportation. The product may thus be an industrial product which may be any mechanical part, such as a part of a terrestrial vehicle (including e.g., car and light truck equipment, racing cars, motorcycles, truck and motor equipment, trucks and buses, trains), a part of an aerial vehicle (including e.g., airframe equipment, aerospace equipment, propulsion equipment, defense products, airline equipment, space equipment), a part of a naval vehicle (including e.g., navy equipment, commercial ships, offshore equipment, yachts and workboats, marine equipment), a general mechanical part (including e.g., industrial manufacturing machinery, heavy mobile machinery or equipment, installed equipment, industrial equipment product, fabricated metal product, tire manufacturing product), an electro-mechanical or electronic part (including e.g., consumer electronics, security and/or control and/or instrumentation products, computing and communication equipment, semiconductors, medical devices and equipment), a consumer good (including e.g., furniture, home and garden products, leisure goods, fashion products, hard goods retailers' products, soft goods retailers' products), a packaging (including e.g., food and beverage and tobacco, beauty and personal care, household product packaging).
The client computer of the example comprises a central processing unit (CPU) 1010 connected to an internal communication BUS 1000, a random access memory (RAM) 1070 also connected to the BUS. The client computer is further provided with a graphical processing unit (GPU) 1110 which is associated with a video random access memory 1100 connected to the BUS. Video RAM 1100 is also known in the art as frame buffer. A mass storage device controller 1020 manages accesses to a mass memory device, such as hard drive 1030. Mass memory devices suitable for tangibly embodying computer program instructions and data include all forms of nonvolatile memory, including by way of example semiconductor memory devices, such as EPROM, EEPROM, and flash memory devices; magnetic disks such as internal hard disks and removable disks; magneto-optical disks. Any of the foregoing may be supplemented by, or incorporated in, specially designed ASICs (application-specific integrated circuits). A network adapter 1050 manages accesses to a network 1060. The client computer may also include a haptic device 1090 such as cursor control device, a keyboard or the like. A cursor control device is used in the client computer to permit the user to selectively position a cursor at any desired location on display 1080. In addition, the cursor control device allows the user to select various commands, and input control signals. The cursor control device includes a number of signal generation devices for input control signals to system. Typically, a cursor control device may be a mouse, the button of the mouse being used to generate the signals. Alternatively or additionally, the client computer system may comprise a sensitive pad, and/or a sensitive screen.
The computer program may comprise instructions executable by a computer, the instructions comprising means for causing the above system to perform the method. The program may be recordable on any data storage medium, including the memory of the system. The program may for example be implemented in digital electronic circuitry, or in computer hardware, firmware, software, or in combinations of them. The program may be implemented as an apparatus, for example a product tangibly embodied in a machine-readable storage device for execution by a programmable processor. Method steps may be performed by a programmable processor executing a program of instructions to perform functions of the method by operating on input data and generating output. The processor may thus be programmable and coupled to receive data and instructions from, and to transmit data and instructions to, a data storage system, at least one input device, and at least one output device. The application program may be implemented in a high-level procedural or object-oriented programming language, or in assembly or machine language if desired. In any case, the language may be a compiled or interpreted language. The program may be a full installation program or an update program. Application of the program on the system results in any case in instructions for performing the method. The computer program may alternatively be stored and executed on a server of a cloud computing environment, the server being in communication across a network with one or more clients. In such a case a processing unit executes the instructions comprised by the program, thereby causing the method to be performed on the cloud computing environment.
| Number | Date | Country | Kind |
|---|---|---|---|
| 23305499.8 | Apr 2023 | EP | regional |
