Method for generating a database for tracking component compliance indicators and associated devices

Abstract

The invention relates to a method for generating a semantic base (B) containing indicator tracking data determining whether a component is suitable for a specific use, which method comprises: - a step of obtaining an initial database associating an initial identification datum and a plurality of initial features with the component; and - a step of converting the initial database into the semantic base (B), having a modified data structure and a set of rules, each of which is intended to be applied to a feature stored in the semantic base (B) in order to evaluate an indicator, the data structure and the set of rules of the semantic base (B) determining the evaluation of the indicators by a computer (15). The method comprises a step of evaluating the indicator for the component by applying a rule to the associated features.

Description

[0001] DESCRIPTION

[0002] TITLE: Method for generating a database for monitoring compliance indicators of components and associated devices

[0003] FIELD OF THE INVENTION

[0004] The present invention relates to a method for generating a semantic database for tracking component compliance indicators.

[0005] The present invention also relates to a semantic database for tracking component compliance indicators, a computer program product and an information medium.

[0006] Conformity is taken here in the sense that measured or calculated indicators conform to predefined expectations for a particular use.

[0007] TECHNOLOGICAL BACKGROUND OF THE INVENTION

[0008] The invention relates to the field of monitoring conformity indicators of mechanical, electronic or software components, for example for the manufacture, maintenance or reconditioning of mechanical, electronic or software devices.

[0009] In industry, it is necessary to know the characteristics of different components, in order to be able to develop, manufacture, maintain or recondition different devices.

[0010] For this purpose, it is known to use tabloid-type databases containing certain information relating to components.

[0011] However, these databases are generally specific to each party involved in the manufacturing, maintenance or reconditioning chain of the device. Thus, the user does not necessarily know where to look for the information.

[0012] Additionally, with tabloid-style databases, searching for information can be time-consuming and tedious.

[0013] Moreover, given the proliferation of databases, a user does not necessarily know where to look for the information he needs.

[0014] Even assuming that the user is aware of all the databases, he is often faced with a terminology problem since each database tends to use its respective terminology specific to the stakeholders for which the said database was established. Even assuming further that the user is aware of all the terminologies, the same characteristic can have different values ​​or different meanings, also called semantics in different databases depending on the source from which the characteristic originates.

[0015] Thus, even assuming that the user consults each database, he is then faced with a problem of consistency between the characteristics present in the different databases and does not have an objective indicator to determine which characteristic he should trust the most.

[0016] These difficulties then lead to risks of non-conformity of manufactured devices, resulting from maintenance or reconditioning.

[0017] For example, to manufacture a device intended to be immersed in water, a user needs to know the pressure resistance of the various components that will be used to form the device in order to choose those that will actually be used.

[0018] Thus, for some stakeholders, an order of magnitude of the pressure resistance of a component is sufficient, while a precise value is necessary for others.

[0019] In the example presented above, there is therefore a risk that one of the components will not withstand the required pressure.

[0020] SUMMARY OF THE INVENTION

[0021] There is therefore a need to make the provision of information more reliable and to simplify access and use for a user, concerning the conformity of use of components for the manufacture, maintenance or reconditioning of devices, in particular mechanical, electronic or software.

[0022] To this end, the subject of the invention is a method for generating a semantic database for monitoring component conformity indicators, the components being intended to be used for a specific use, the specific use being the manufacture, maintenance or reconditioning of mechanical, electronic or software devices comprising at least one of the components, indicators of conformity to the use determining whether or not the component is suitable for use for the specific use, the method being implemented by a computer system comprising a computer and a memory, the method comprising an initialization phase comprising:

[0023] - a step of obtaining an initial database, the initial database having an initial structure, the initial structure being a tabular structure associating, with each component of a plurality of initial components, at least one initial identification data of the component and a plurality of initial characteristics of said component, at least one of the characteristics resulting from a measurement of a quantity characterizing the component, and

[0024] - a step of converting the initial database into the semantic database for monitoring compliance indicators, called the converted database, having a data structure modified with respect to the initial database and a set of evaluation rules, each evaluation rule being intended to be applied to at least one characteristic stored in the converted database to evaluate a compliance indicator for the use of the component associated with the at least one characteristic, the data structure and the set of evaluation rules of the converted database determining the implementation of the evaluation of the compliance indicators by the calculator, the method further comprising a use phase comprising:

[0025] - an evaluation step, for each component, of the indicator of conformity to the use of the component, by application, to the associated characteristics, of at least one rule of the set of evaluation rules.

[0026] According to particular embodiments, the generation method according to the invention comprises one or more of the following characteristics, taken in isolation or in all technically possible combinations.

[0027] - the use phase also includes:

[0028] - an alert step, during which an alert is issued to a user if at least one compliance indicator violates a predefined compliance criterion established from the evaluation rules;

[0029] - the method further comprises an update phase comprising:

[0030] - a step of receiving at least one new characteristic relating to at least one component of the converted database,

[0031] - a step of updating the converted database, to obtain an updated database, comprising: o a sub-step of re-evaluation, by the calculator, of each conformity indicator associated with said component, by applying the evaluation rules to a set comprising the characteristics associated with said component in the converted database and the new characteristic, and o a sub-step of storing, in the updated database, each re-evaluated conformity indicator, - an alert step, during which an alert is issued to a user if at least one re-evaluated conformity indicator infringes a predefined conformity criterion established from the evaluation rules;

[0032] - an edition date is associated with each characteristic, the edition date being the date on which the characteristic was measured, determined or established, the use phase further comprising:

[0033] - a prediction step, for each component, of a risk of non-compliance over a predefined time horizon, based on the results of the evaluation rules and the date of publication of each characteristic associated with the component, and

[0034] - a warning step, during which a warning is issued to a user if at least one risk of non-compliance is higher than a predefined threshold;

[0035] - the data structure of the converted database is a semantic graph;

[0036] - each evaluation rule is specific to at least one component of the converted database, at least one rule of the set of evaluation rules depending on the specific use of the associated component(s);

[0037] - at least one conformity indicator quantifies the confidence placed in one of the characteristics of the component to describe an aspect of the component;

[0038] - the method further comprises a phase of integrating a new component into the converted database, comprising: a step of receiving identification data of the new component and new specific evaluation rule(s), a step of integrating, into the modified structure of the converted database, the identification data, and the new evaluation rule(s), the integration comprising the establishment of a semantic link between characteristic(s) of the converted database and the identification data of the new component, and a step of evaluating the conformity indicator associated with the new component by applying the new evaluation rule(s) to the characteristic(s) linked to the identification data;

[0039] - at least one of the components of the converted database is a seal whose specific use is the manufacture or maintenance of a waterproof housing intended to be immersed in water to a predetermined depth to prevent water infiltration into the housing despite overpressure induced by the water on an exterior of the housing, a characteristic of said component being a measured resistance to overpressure, one of the evaluation rules being capable of calculating a difference between the measured resistance to overpressure and a predefined value depending on the predetermined depth, the conformity indicator obtained by application of said evaluation rule being a function of the calculated difference;

[0040] - another component of the converted database is the same seal for a separate specific use, the separate specific use being the manufacture or maintenance of an aircraft fuselage to prevent air leaks from an interior of the fuselage to an exterior where the atmospheric pressure is lower, a characteristic of said component being a measured resistance to depression, one of the evaluation rules being suitable for calculating a difference between the measured resistance to depression and a predefined value depending on a predefined maximum flight altitude of the aircraft, and the conformity indicator obtained by application of said evaluation rule being a function of the calculated difference; and

[0041] - at least one of the components of the converted database is a capacitor whose specific use is the reconditioning of a switching power supply, the characteristics of said component comprising: a measurement of capacitance of said component, a thermal dispersion of said component and / or one or more acceptable soldering technologies, one of the evaluation rules being specific to the technical characteristic considered and to the context of use defined by a user, the conformity indicator obtained by application of said rule.

[0042] The present invention also relates to a semantic database for tracking component conformity indicators, the components being intended to be used for a specific use, the specific use being the manufacture, maintenance or reconditioning of mechanical or electronic devices comprising at least one of the components, indicators of conformity to the use determining whether or not the component is suitable for use for the specific use, the database for tracking conformity indicators being capable of being obtained by a generation method as described above.

[0043] The present invention also relates to a computer program product comprising software instructions which, when implemented by a computer, implement a method as described above. The present invention also relates to an information medium comprising program instructions forming a computer program, the computer program being loadable onto a data processing unit and implementing a method as described above when the computer program is implemented on the data processing unit.

[0044] In this description, the expression "suitable for" means indifferently "adapted for", "adapted to" or "configured for".

[0045] BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The invention will be better understood by reading the description which follows, made with reference to the appended drawings in which:

[0047] [Fig 1] Figure 1 is a schematic representation of a computer system for generating a semantic database,

[0048] [Fig 2] Figure 2 is a schematic representation of the semantic database resulting from the computer system generating Figure 1,

[0049] [Fig 3] Figure 3 is a flowchart of a generation method implemented by the generation computer system of Figure 1, and

[0050] [Fig 4] Figure 4 is a detailed view of a self-correction phase of the process, a flowchart of which is shown in Figure 3.

[0051] DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

[0052] System Description

[0053] In Figure 1 is shown a generation 10 computer system, also called computer system 10. The computer system 10 comprises a computer 15 and a memory 20.

[0054] The computer 15 is capable of implementing a method for generating a semantic database which will be described later. The generation method is therefore a generation method implemented by computer.

[0055] The calculator 15 is an information system designed to manipulate and / or transform represented data.

[0056] The method is carried out in the form of one or more software programs, i.e. in the form of a computer program, also called a computer program product, it is furthermore capable of being recorded on a medium, not shown, readable by a computer. The computer-readable medium is, for example, a medium capable of storing electronic instructions and of being coupled to a bus of a computer system. For example, the readable medium is an optical disk, a magneto-optical disk, a ROM memory, a RAM memory, any type of non-volatile memory (for example FLASH or NVRAM) or a magnetic card. A computer program comprising software instructions is then stored on the readable medium.

[0057] To implement the generation method, the computer 15 communicates with the memory 20 suitable for storing data to be processed by the computer 15 and data processed by the computer 15.

[0058] Purpose of the generation process

[0059] Description of the semantic database

[0060] The operation of the system 10 is now described with reference to Figure 2.

[0061] The generation method aims to obtain a semantic database B for monitoring conformity indicators I of components 30.

[0062] The components 30 are here real physical entities forming part of a real physical device, which is here a mechanical, electronic or software device. Such a device is represented schematically by a rectangle D in FIG. 2 and will be referred to as device D in the remainder of the description.

[0063] Specific examples of components 30 will be given later.

[0064] The components 30 are intended to be used for a specific use U.

[0065] This specific use U is a technical use chosen from the development, manufacture, maintenance or reconditioning of the device D.

[0066] The conformity indicators I to use U determine whether or not the component 30 is suitable for use for the specific use U.

[0067] Each conformity indicator I is thus a measure of the suitability of the component 30 for use in use U.

[0068] Reading the semantic database B by the computer system 10 provides access to a set of information.

[0069] The set of information comprises, for each component of a plurality of initial components, at least one initial identification data item for the component and a plurality of initial characteristics C of said component, at least one of the characteristics C resulting from a measurement of a quantity characterizing the component.

[0070] The information set therefore includes measurements, also called observation points measured by means of metrics or indicators.

[0071] Reading the semantic database B by the computer system 10 therefore provides access to measurements carried out for each component 30.

[0072] However, the semantic database B is not a simple database. The reading of the semantic database by the computer system 10 provides specific instructions for the evaluation of the conformity indicator I for each component 30.

[0073] This is because the semantic database B is characterized by a data structure and a set of evaluation rules Rs which determine the implementation of the evaluation of the conformity indicators I by the calculator 15.

[0074] In fact, to calculate the conformity indicators I, the computer system 10 applies to the characteristics C available in the data structure of the semantic database B and relating to the component 30 to be characterized at least one rule R from the set of evaluation rules Rs.

[0075] The content of the semantic database B is thus a command of the actions carried out by the computer system 10.

[0076] By means of the semantic database B, the computer system 10 is thus able to obtain the conformity indicators I based on measurements carried out previously.

[0077] Preferably, each characteristic is associated with several conformity indicators I quantifying or qualifying the characteristic respectively according to one of the aspects described below.

[0078] For example, the I conformity indicators are divided into four types of aspects. The four types are as follows:

[0079] - the intrinsic type which quantifies the confidence placed in the characteristic as such and in the source which produced this characteristic;

[0080] - the contextual type which quantifies the confidence granted in the characteristic for the use of the associated component(s) in the specific use considered;

[0081] - the representational type which quantifies the confidence in the clarity of obtaining information from the characteristic; and the accessibility type which quantifies the confidence in the availability of data for different users involved in the specific use of the component(s) associated with the characteristic.

[0082] These types are defined in a manner specific to the device D considered or to the components 30 constituting it. Alternatively, these types depend on the recipient of the semantic database.

[0083] The intrinsic attribute is independent of the specific use of the component(s) associated with the characteristic. For example, intrinsic type I indicators include: credibility, truthfulness, accuracy, objectivity, and reputation.

[0084] Credibility is the degree to which a characteristic is believable. Credibility is, for example, calculated based on the characteristic under consideration, i.e., its order of magnitude, its unit of measurement, or its freshness, as explained below. Credibility is further calculated based on the source that produced the characteristic, with sources being pre-evaluated by experts, for example.

[0085] Veracity defines whether information is accurate and correct in relation to the actual object it describes. Thus, veracity depends on the source of the characteristic, for example, the entity that produced the characteristic in the case of a certificate and / or the way in which the characteristic was collected. For example, veracity depends on the measuring instrument in the case where the characteristic results from a measurement. The calculation of veracity is based on business definitions and rules. The result is binary: 0 = FALSE, 1 = TRUE.

[0086] Accuracy quantifies whether the characteristic is certified error-free, defect-free, or how easily errors are identifiable, as well as the precision of the characteristic. Accuracy is a score between 0 and 1.

[0087] Accuracy is calculated, for example, using the following equation:

[0088] Or :

[0089] • a) is the characteristic value,

[0090] • is the actual corresponding value, and

[0091] • d(. , . ) is a distance measure such as a Euclidean distance or a Hamming distance.

[0092] For example, if the characteristic to be qualified has a precision of 10 -3 near, then the accuracy is approximately equal to 0.997.

[0093] Objectivity, also called impartiality, quantifies the degree to which a characteristic is impartial or is affected by subjective criteria in its development. For example, objectivity is calculated from belief-theoretic approaches (extension of probability functions in which the measure of the universe is less than or equal to 1, which allows characterizing the subjectivity of a decision-maker) or from fuzzy logic.

[0094] Reputation quantifies an overall subjective assessment of the quality of the characteristic over a predefined period of time by a user. Reputation is, for example, calculated by labeling similar to NutriScore-type labeling, also called energy score (“ABCD or E”) or by a financial rating. For example, contextual type I indicators include: added value, relevance, completeness, and freshness.

[0095] Added value quantifies whether the feature provides a competitive advantage in the specific use. Added value is, for example, calculated using functional analysis and value analysis (FAVA) methods.

[0096] Relevance quantifies the importance of the feature for the specific use considered. In other words, relevance quantifies whether the feature is interesting, applicable, and usable for the specific use. Relevance is, for example, calculated based on the domain of use (ODD: Operational Design Domain, defined during the specifications). It is therefore verified that all the types of data necessary to cover the ODD are present in order to answer the questions asked by users. The relevance calculated is then, for example, a coverage rate.

[0097] Completeness quantifies whether the feature describes the component in depth. Completeness is preferably complementary to relevance. For example, completeness is calculated as the coverage of data to answer the question asked over the entire context studied.

[0098] Freshness quantifies the age of the characteristic. In other words, freshness defines how long ago the characteristic was established.

[0099] For example, freshness is calculated using the following equation: Or :

[0100] • max(. , . ) is the maximum function

[0101] • age of the characteristic is the time difference (in number of days) between the insertion of the characteristic in the converted database and the last freshness assessment associated with this characteristic,

[0102] • retention period is a predefined value respective to the characteristic corresponding to the number of days during which the characteristic is considered up to date, and

[0103] • s is a parameter adjustable by an expert and dependent on the characteristic.

[0104] The representational approach quantifies the ability to present and exploit the information contained in the characteristic, that is, in a file, a document, a message or even a graphical interface used to transmit the information. Thus, we consider the form in which the information is "formatted" or "layout", depending on whether we are interested in an application or a human. For example, the I indicators of the representational type include: interpretability, comprehensibility, coherence of the representation and conciseness of the representation.

[0105] Interpretability quantifies the ability to make the user understand the meaning of what is shown to them. It is calculated through user feedback (surveys, problem reports or questions).

[0106] Understandability quantifies the readability and clarity of information in the feature. Similarly, understandability is calculated through user feedback.

[0107] Representation consistency quantifies a deviation in the presentation of information in the feature from other similar features C or over time. Representation consistency is preferably guaranteed by constructing and maintaining the domain taxonomy and ontology, i.e. the semantic database

[0108] Conciseness of representation quantifies whether the feature does not contain too much unnecessary information and whether the data formats are optimal for performance and usability. Conciseness is calculated as a ratio of useful information (as defined by the ODD and ontology) to file size.

[0109] For example, indicators I of accessibility type include accessibility and access security.

[0110] Accessibility quantifies how identifiable, retrievable, and available the feature is, both at a given moment and in a lasting manner, for different users. Accessibility also quantifies how quickly the feature can be accessed by the user(s). Accessibility is, for example, the ratio between the number of accesses to information over all queries made, weighted by the effort (time and energy) to obtain useful information.

[0111] Access security quantifies how the feature is secured and in particular whether the feature is only accessible to users authorized to view it (need to know and right to know). Access security is defined by the coverage of security rules and implemented in relation to the sensitivity of each piece of information.

[0112] It is clear that not every compliance indicator is necessarily assessed for every characteristic. In such a case, either the indicator(s) take a neutral value, such as a median value, or the corresponding compliance indicator(s) are not calculated.

[0113] Likewise, each indicator is evaluated according to a given context. Each conformity indicator I being, as indicated previously, a measure of the suitability of the component 30 to be used in the use U, the computer system 10 is a measurement system since the computer system 10 takes into account measurements, performs calculations on these measurements and obtains a new measurement.

[0114] The semantic database B is thus capable of controlling a computer system 10 so that it obtains measurements, performs calculations and obtains one or more new measurements.

[0115] Regardless of the value for a user of the technical data contained, the semantic database B therefore has an additional advantage over a simple database which would have the same data since it allows a measuring system to be ordered to carry out the desired measurement.

[0116] It should be emphasized here that without the semantic database B, the computer system 10 would not be able to precisely calculate the conformity indicators I and therefore would not be able to measure whether the component 30 is suitable or not for the specific use.

[0117] In the following section, another advantage of this semantic database B is shown, namely its generic character. For this, examples of the previously used notions will be given, namely specific use II, device D, components 30 and conformity indicator I.

[0118] Examples of elements present in the semantic base or used to obtain the semantic database B

[0119] In a first example, in the mechanical field, the specific use is the manufacture or maintenance of a waterproof housing. In this example, one component is a seal, another component being a cover of the housing.

[0120] In a second example, also in the mechanical field, the specific use is the manufacture or maintenance of an aircraft fuselage. In this second example, the components include: a seal, a fuselage sheet, and a nut.

[0121] In a third example, in the electronics field, the specific use is the reconditioning of a switching power supply. In this third example, the components include: a protective fuse, a diode bridge, a capacitor, a transistor, a transformer, a coil, and a control circuit.

[0122] In a fourth example in the software field, the specific use is the exploitation of open-source software. The components include: terms of use of the software, operating licenses, dependencies on other software or technical infrastructure, documents and source code of the software.

[0123] Conformity indicators I for use determine whether or not the component is suitable for use in the specific use. Conformity indicators I thus quantify or qualify the confidence that can be given to a plurality of characteristics C describing each component. Thus, if the confidence is too low, for reasons that will be detailed below, the component is considered non-compliant for the specific use.

[0124] Preferably, the semantic database B includes components for different specific uses. Preferably, the same semantic database B simultaneously includes components intended for mechanical uses, components intended for electronic uses and components intended for software uses.

[0125] To generate the semantic database B, the memory 20 stores for example one or more initial databases 25. Each initial database 25 has an initial structure. The initial structure is a tabular structure associating, with each component 30 of a plurality of initial components, at least one initial identification data of each component and a plurality of initial characteristics C of said component 30, at least one of the characteristics C resulting from a measurement of a quantity characterizing the corresponding component.

[0126] The identification data of each component 30 is for example a name of the component, a reference of the component 30 intrinsic to the initial database 25, or a standardized reference of the component 30.

[0127] For example, if component 30 is a seal, the identification data is, for example, "seal", "seal", or "seal REF '1'".

[0128] For example, if component 30 is a resistor (in the component sense), the identification data is for example “resistor” or “RT100000025M”.

[0129] Each initial characteristic quantifies or qualifies an aspect of the component.

[0130] For example, for mechanical components 30, one of the initial characteristics C quantifies a geometry of the component 30, a mass of the component 30, a volume of the component 30, a resistance of the component 30, a thermal capacity of the component 30, a drying time or a covering power.

[0131] For example, for electronic components 30, one of the initial characteristics C quantifies a resistance of component 30, an inductance of component 30, a capacity of component 30, a maximum intensity that component 30 can support, a transformation ratio of component 30, or even a memory storage capacity, a reading speed or a calculation capacity of a computer component...

[0132] For example, for software components 30, one of the characteristics C quantifies a memory space occupied by the software component 30, a version of the software, a license contract.

[0133] In all the preceding examples, another initial characteristic is for example a certificate of validity of a component 30 for use in a predefined device D or an authorization to market a device D comprising said component 30 in a specific country.

[0134] It is thus clear that the identification data is preferably a string of characters. Each characteristic is either a document, such as a PDF document or a table, or a value, preferably a binary, a relative integer or a real number.

[0135] Each conformity indicator I is preferably a binary, a relative integer or a real number.

[0136] Each evaluation rule R is, for example, a computer function comprising one or more formulas for calculating the associated conformity indicator(s) I. In the following, it will be explained how such a semantic database B can be obtained.

[0137] It should be understood that the methodology to be outlined is generic and can take many different forms depending on the components 30 and specific uses considered.

[0138] Also, a generic description will first be described with reference to Figure 3 and then some concrete examples will be used that are sufficiently different to clearly understand how the semantic database B can be obtained in each case using the generation method that is the subject of the present invention.

[0139] Description of the generation process according to Figure 3

[0140] With reference to Figure 3, the generation method comprises five phases: an initialization phase 100, a use phase 200, an update phase 300, an integration phase 400 and a correction phase 500 corresponding to a return loop to the initialization phase 100.

[0141] Each of these phases will be described in turn below. Examples of initialization phase 100

[0142] The initialization phase 100 aims to obtain an initial semantic database B.

[0143] The initialization phase 100 comprises an obtaining step 110 and a conversion step 120.

[0144] During the obtaining step 110, the calculator 15 obtains the or each initial database 25, for example from the memory 20.

[0145] Then, during the conversion step 120, the calculator 15 converts the or each initial database into the semantic database B of compliance indicator tracking data I, called converted database B. The converted database B has a modified data structure compared to the or each initial database 25 and a set of evaluation rules. This modification is preferably carried out either automatically or semi-automatically. Alternatively, it can be carried out completely manually.

[0146] The data structure and the set of evaluation rules of the converted database B determine the implementation of the evaluation of the conformity indicators I by the calculator 15.

[0147] For example, the converted database B is a semantic graph as schematically illustrated in Figure 2. That is, the converted database B comprises, for example, nodes and edges. Each node is a component or a feature associated with a component. Each edge is a semantic relationship associating a component and an associated feature.

[0148] According to one embodiment, in the converted database B, at least one characteristic is connected to several components by different arcs.

[0149] This is achieved because the converted database B is a semantic graph. In a tabloid database, each feature is associated with only one component.

[0150] For example, if several components are tested and a single validity certificate is issued for all of these components, the converted database B includes only one copy of this certificate, which forms a node in the semantic graph. An arc then connects this node to each component considered in this certificate.

[0151] Preferably, each component is characterized, in the converted database B, by the following elements: identification, classification, characteristic(s), application, use and possibly associations.

[0152] The identification is for example the identification data of the component present in the or one of the initial databases 25. The classification defines the type of component corresponding to the identification. For example, when the identification of a component is “RT100000025M”, the associated classification is “resistor”.

[0153] The C characteristic(s) are those present in the initial database(s) 25. The C characteristics are for example of two types: documentation and technical characteristics.

[0154] The documentation includes the set of files included in the initial database(s) 25 relating to the components.

[0155] Documentation includes, for example, component data sheets, component descriptions and certificates.

[0156] The technical characteristics are quantified values ​​directly entered into the initial database(s) 25 or automatically extracted from the documentation during the conversion step 120.

[0157] For example, certain technical characteristics result from a measurement carried out prior to the implementation of the method and, either of which measured values ​​are directly integrated into the initial database 25, or of which values ​​are compiled in a document such as a PDF file. In this second case, the values ​​of the technical characteristics are preferably extracted automatically from said documentation.

[0158] For example, applications define the domain or context of use of the component. For example, applications indicate usage recommendations from the component manufacturer. In a particular example where the component is an industrial paint, applications define, for example, recommended temperature or humidity values ​​for applying the paint.

[0159] The uses come from the initial database(s) 25 and define who the users of the components are as well as the use cases. Here the uses are associated with the roles of the users of the device D. These roles can be of different natures: a user of the device such as a buyer or an engineer; a business expert in charge of producing and maintaining the data.

[0160] Associations characterize the relationships between components. For example, associations indicate whether two components are compatible, equivalent, derived from each other, or substitute components for each other.

[0161] Each evaluation rule is intended to be applied to at least one characteristic stored in the converted database B to evaluate a conformity indicator I to the use of the component associated with the at least one characteristic. The evaluation rules are for example predefined and include the equations and the ways in which each indicator I is calculated.

[0162] Each evaluation rule is optionally specific to at least one component of the converted database B. At least one rule in the set of evaluation rules depends on the specific use of the associated component(s).

[0163] Examples of use phase 200

[0164] The use phase 200 comprises an evaluation step 210, an alert step 220 and preferably a prediction step 230 and a warning step 240.

[0165] During the evaluation step 210, the calculator 15 evaluates, for each component, the indicators of conformity I to the use of the component, by application, to the associated characteristics C, of ​​at least one rule of the set of evaluation rules.

[0166] Preferably, each rule also includes a predefined conformity criterion associated with each conformity indicator I. If at least one conformity indicator I does not respect the predefined criterion, then the associated component is indicated as non-compliant in the converted database B.

[0167] The use phase 200 then comprises an alert step 220, during which the calculator 15 issues an alert to a user if at least one conformity indicator I violates the predefined conformity criterion. Indeed, in such a situation, the component is not compliant with the specific use and the alert step allows the user to be informed thereof.

[0168] The alert is, for example, data available and made available to the user through different means depending on the use cases and user experience, such as a visual alert (pop-up message, icon), a summary dashboard, a global or local map, or any type of signal likely to alert the user.

[0169] In the example where the characteristic is a certificate that must be issued annually and certifies that a component withstands a stress such as mechanical stress, if the freshness is dated more than one year ago, then the conformity indicator I associated with said components does not meet the predefined conformity criterion. Thus, during the alert step 220, an alert is issued to the user to warn them that the component is not compliant since the certificate was issued more than one year ago.

[0170] Optionally, an edit date in the source and / or an update date in the converted database B is associated with each characteristic in the converted database B. The edit date is the date on which the characteristic was measured, determined or established. The update date is the date on which the characteristic was recorded in the converted database B on which the indicators are calculated.

[0171] As an optional addition, the use phase 200 further comprises a prediction step 230, for each component, of a risk of non-compliance at a predefined time horizon, based on the results of the evaluation rules and the date of publication of each characteristic associated with the component.

[0172] The predefined time horizon advantageously depends on the characteristic considered and / or the component considered and preferably on the employment context.

[0173] For example, in the case of a certificate that must be issued annually, during the prediction step 230, the calculator 15 predicts a risk of non-compliance at a predefined time horizon of the certificate if the anniversary date of the certificate expires before the predefined time horizon.

[0174] For example, compliance risk has a binary value. Alternatively, compliance risk has an integer or real value. The value of compliance risk depends on the predefined time horizon, the issue date of the characteristic, and the evaluation rules.

[0175] According to this optional addition, the use phase 200 further comprises a warning step 240, during which the calculator 15 issues a warning to a user if at least one risk of non-compliance is greater than a predefined threshold.

[0176] At the end of the usage phase 200, the converted database is functional and usable. For example, it can be queried by the user to check the conformity indicators I associated with each component.

[0177] Update Phase Examples 300

[0178] The update phase 300 comprises a reception step 310, an updating step 320 and an alert step 330.

[0179] During the update phase 300, at least one new characteristic concerning at least one component of the converted database B is added to said converted database B.

[0180] The update phase 300 comprises a step 310 of receiving at least one new characteristic relating to at least one component of the converted database B.

[0181] Each new characteristic received is for example a certificate, a measurement, or an authorization to market a component. The update phase 300 further comprises a step 320 of updating the converted database B, to obtain an updated database.

[0182] The updating step 320 comprises a re-evaluation sub-step 321, during which the calculator 15 re-evaluates each conformity indicator I associated with said component, by applying the evaluation rules to a set comprising the characteristics C associated with said component in the converted database B and the new characteristic C. The updating step 320 is also called the data updating step.

[0183] Preferably, during the re-evaluation sub-step 321, only the conformity indicators I associated with the new characteristic are evaluated.

[0184] Alternatively, the converted database B identifies other characteristics C dependent (directly or indirectly on) the new characteristic and the conformity indicators I associated with these characteristics C are also re-evaluated.

[0185] For example, if the new characteristic is an authorization to market the component only if a measurement of said component is below a threshold, then, during the re-evaluation sub-step 321, the calculator 15 re-evaluates the conformity indicators I of the measurement characteristic.

[0186] The updating step 320 further comprises a sub-step 322 of storing, in the updated database, each re-evaluated conformity indicator I.

[0187] The update phase 300 further comprises an alert step 330, during which the calculator 15 issues an alert to a user if at least one re-evaluated conformity indicator I violates a predefined conformity criterion established from the evaluation rules.

[0188] The alert issued during the alert step 330 of the update phase 300 is preferably similar to the alert issued during the alert step 220 of the use phase 200.

[0189] Examples of integration phase 400

[0190] The integration phase 400 comprises a reception step 410, an integration step 420, an evaluation step 430 and an alert step 440.

[0191] Integration phase 400 concerns the integration of a new component into the converted database B.

[0192] The new component is, for example, a new item for manufacturing, maintenance or reconditioning, a new device D or an alternative to a component already included in the converted database B. For example, the new component is theoretically identical to an existing component but comes from a competing supplier or is intended for a specific use distinct from that intended for the first component.

[0193] The integration phase 400 comprises a step 410 of receiving identification data for the new component and preferably new specific evaluation rule(s).

[0194] The integration phase 400 further comprises a step 420 of integrating, into the modified structure of the converted database B, the identification data and the new evaluation rule(s). The integration step 420 comprises establishing a semantic link between characteristic(s) of the converted database B and the identification data of the new component.

[0195] In particular, if certain characteristics C already present in the converted database B present information associated with the new component, said characteristics C are linked to the identification data of the new component by a respective semantic link.

[0196] For example, if a feature of the converted database B is a marketing authorization for a plurality of components including the new component, said feature is linked to the identification data of the new component.

[0197] The integration phase 400 further comprises an evaluation step 430, during which the calculator 15 evaluates the conformity indicator(s) I associated with the new component, by applying new evaluation rule(s) to the characteristic(s) which is / are linked to the identification data of the new component.

[0198] Preferably, the integration phase 400 further comprises an alert step 440, during which the calculator 15 issues the aforementioned alert to the user if a conformity indicator I does not meet a predefined criterion.

[0199] According to one embodiment, the update 300 and integration 400 phases are implemented simultaneously. Thus, during the reception steps 310, 410, the identification data of a new component and new characteristic(s) relating to this component are received. Thus, the re-evaluation sub-steps 321 and the evaluation step 430 are implemented simultaneously and comprise the evaluation of the conformity indicators I of the characteristic(s) C linked to the new component. Examples of correction phase 500

[0200] During the correction phase 500, the characteristics whose indicators imply non-conformities (for example below a desired threshold) make it possible to identify the corrections to be made to the input data.

[0201] The correction phase 500 comprises a self-correction step 510 and, where appropriate, a correction recommendation step 520.

[0202] Alternatively or as an optional addition, the method further comprises an automatic correction phase 500, during which the converted database B makes corrections or correction recommendations in the respective identification, classification, characteristic(s), application, use and possibly associations of each component of the converted database. The automatic correction phase 500 is preferably implemented in parallel with the evaluation 200, update 300 and integration 400 phases.

[0203] The automatic correction phase 500 comprises a self-correction step 510, during which errors, such as typographical errors or missing information, are automatically detected and corrected, for example from a dictionary and an automatic corrector included in the converted database B.

[0204] With reference to Figure 4, the self-correction step 510 comprises an error detection sub-step 512, during which the automatic corrector detects, in one or more words, an error by comparing the word with the words included in the dictionary.

[0205] The self-correction step 510 then comprises a sub-step 514 of generating candidates for correction, during which the automatic corrector determines, for example among the words in the dictionary, those which are similar to the word containing an error, also called an erroneous word.

[0206] The self-correction step 510 then comprises a sub-step 516 of classifying the generated candidates, during which the automatic corrector calculates for each candidate, a distance with the erroneous word. The distance is for example the Levenshtein distance, also called minimum editing distance. This distance quantifies the minimum number of operations required to transform the erroneous word into the candidate. The operations considered preferably include insertion, deletion, substitution and transposition.

[0207] The self-correction step 510 then comprises a comparison sub-step 518 during which the calculator 15 compares the distance between the candidate having said lowest distance and the erroneous word, with a predefined threshold. If the distance is less than or equal to the predefined threshold, the self-correction step 510 comprises a replacement sub-step 519 during which the calculator 15 automatically replaces the erroneous word with said candidate.

[0208] Otherwise, the self-correction phase 500 includes a correction recommendation step 520. Preferably, the correction recommendation phase 520 is based on word embedding techniques.

[0209] The correction recommendation step 520 comprises an initialization sub-step 522 during which the calculator 15 associates, with each word of the dictionary, a vector of initial numerical values.

[0210] Then, the correction recommendation step 520 comprises a learning sub-step 524 during which the calculator 15 adjusts the numerical values ​​of each vector so that the vectors associated with two words regularly present in the same sentence and / or in the same context have substantially similar values. For this purpose, the CBOW or Skip-Gram process is for example used. This then provides a semantic representation of the words according to the occurrences of appearance in the converted database B.

[0211] The correction recommendation step 520 comprises a selection sub-step 526 during which the calculator 15 selects, as new candidates, the words whose Levenshtein distance with the erroneous word is less than a predefined threshold.

[0212] Then the correction recommendation step 520 includes a proposal sub-step 528 during which the calculator 15 sends, to the user, a list of new candidates.

[0213] Finally, the correction recommendation step 520 comprises an implementation sub-step 529 during which the calculator 15 receives an instruction of a new candidate selected by the user, from the list of new candidates. The implementation sub-step 529 comprises the replacement of the erroneous word by the new candidate selected by the user.

[0214] The correction step 500 preferably also includes the transmission, to the initial databases, of the identified errors and the corrections made or recommendations as well as the validations made by the users.

[0215] According to a first variant, the method only comprises the initialization phase 100 and the use phase 200.

[0216] According to a second variant, the method further comprises only the update phase 300 or the integration phase 400.

[0217] According to a third variant, compatible with the first and second variants, the method further comprises the correction phase 500. According to a fourth variant, compatible with the first, second and third variants, the use phase 200 does not comprise the prediction 230 and warning 240 steps.

[0218] Specific examples of application of the generation process

[0219] Examples of application of the method will now be described for specific uses.

[0220] According to a first example of application of the method, at least one of the components of the converted database B is a sealing gasket whose specific use is the manufacture or maintenance of a waterproof housing. The waterproof housing is intended to be immersed in water to a predetermined depth to prevent water infiltration into the housing despite an overpressure induced by the water on an exterior of the housing.

[0221] In this first example of application of the method, a characteristic of said component is a measured resistance to overpressure.

[0222] One of the evaluation rules is specific to calculating a difference between the measured overpressure resistance and a predefined value depending on the predetermined depth. The conformity indicator I resulting from the application of this rule is a function of said difference.

[0223] Thus, if the difference is less than a predefined safety threshold, the associated conformity indicator I takes a low value and during the alert step 220, 330 or 440, the alert is issued to the user to indicate to him that the use of the seal is not compliant for the manufacture or maintenance of the waterproof housing.

[0224] Still according to this first example of application of the process, the measured resistance was measured by one of the operators in the production room and another characteristic is a certificate of measurement of the pressure resistance, from a certified metrology institute.

[0225] In such a case, the conformity indicators I of credibility and veracity will take higher values ​​for the characteristic from the certified metrology institute than for the resistance measured by the operator.

[0226] Still according to the first example of application of the method, another component of the converted database B is the same seal for a distinct specific use. The distinct specific use is the manufacture or maintenance of an aircraft fuselage to prevent air leaks from an interior of the fuselage to an exterior where the atmospheric pressure is lower. A characteristic of said component is a measured resistance to depression. Similar to previously, one of the evaluation rules is capable of calculating a difference between the measured resistance to depression and a predefined value depending on a predefined maximum flight altitude of the aircraft. The conformity indicator I obtained by applying said evaluation rule is a function of the calculated difference.

[0227] It is clear that the conformity indicators I evaluated, during the evaluation steps 210, 430 or during the re-evaluation sub-step 321, are not necessarily the same for the component intended for use in the sealed housing and for that intended for use in the fuselage of an aircraft, although these components are theoretically the same components.

[0228] According to a second example of application of the method, at least one of the components of the converted database B is a capacitor whose specific use is the reconditioning of a switching power supply. The characteristics C of said component comprise a thermal dispersion of said component and / or one or more acceptable soldering technologies.

[0229] One of the evaluation rules is specific to the requirements (constraints) of the device such as its conditions of use, the device integration environment and / or the operating conditions, and is based on the indicators previously defined for the given context.

[0230] According to a third example of application of the method, at least one of the components of the converted database B is a source code whose specific use is the exploitation of Open Source software. A characteristic of said component is a version of the software. One of the evaluation rules evaluates the correspondence between the source code and the version of the software, for example using an identifier present in the source code capable of being incremented at each version.

[0231] The conformity indicator I obtained by applying said rule is representative of a correspondence between the version and the source code, such as a binary value equal to “1” if the source code corresponds to the version, or equal to “0” otherwise.

[0232] Thus, the converted database B makes it possible to concentrate all the information and characteristics C associated with each component throughout its cycle, from the development of a device D, to the production, maintenance or reconditioning of said device D, including its marketing.

[0233] In addition, the conformity indicator evaluation technique I makes it possible to automatically evaluate the confidence that can be given to each characteristic and therefore to the use of the component.

[0234] The modified data structure of the converted database B allows for greater agility in the connections between the features C and components. Indeed, when the converted database B is a semantic graph, the connections, or arcs, make it possible to give meaning to each link between the components and the features C. In addition, the semantic graph structure makes it possible to infer knowledge from already existing connections between these elements. This makes it possible, in particular, to predict future non-conformities.

[0235] Thus, the method, the semantic database, the computer program product and the information medium according to the invention make it possible to simplify and make more reliable the provision of information to a user concerning the conformity of use of components for the manufacture, maintenance or reconditioning of devices, in particular mechanical, electronic or software devices.

Claims

CLAIMS 1. Method for generating a semantic base (B) of data for tracking conformity indicators (I) of components (30), the components being intended to be used for a specific use, the specific use being the manufacture, maintenance or reconditioning of mechanical, electronic or software devices (D) comprising at least one of the components, conformity indicators (I) for the use determining whether or not the component is suitable for use for the specific use, the method being implemented by a computer system comprising a computer (15) and a memory (20), the method comprising an initialization phase (100) comprising: - a step (1 10) of obtaining an initial database, the initial database having an initial structure, the initial structure being a tabular structure associating, with each component of a plurality of initial components, at least one initial identification data item of the component and a plurality of initial characteristics of said component, at least one of the characteristics (C) resulting from a measurement of a quantity characterizing the component, and - a step (120) of converting the initial database into the semantic database (B) of compliance indicator tracking data (I), called the converted database (B), having a data structure modified with respect to the initial database and a set of evaluation rules, each evaluation rule being intended to be applied to at least one characteristic (C) stored in the converted database (B) to evaluate a compliance indicator (I) with the use of the component associated with the at least one characteristic (C), the data structure and the set of evaluation rules of the converted database (B) determining the implementation of the evaluation of the compliance indicators (I) by the calculator (15), the method further comprising a use phase (200) comprising: - a step (210) of evaluating, for each component, the indicator of conformity (I) to the use of the component, by application, to the associated characteristics (C), of at least one rule from the set of evaluation rules.

2. Method according to claim 1, wherein the use phase (200) further comprises: an alert step (220), during which an alert is issued to a user if at least one compliance indicator (I) violates a predefined compliance criterion established from the evaluation rules.

3. Method according to claim 1 or 2, further comprising an updating phase (300) comprising: a step of receiving (310) at least one new characteristic (C) relating to at least one component of the converted database (B), a step of updating (320) the converted database (B), to obtain an updated database, comprising: o a sub-step of re-evaluation (321), by the calculator (15), of each conformity indicator (I) associated with said component, by applying the evaluation rules to a set comprising the characteristics (C) associated with said component in the converted database (B) and the new characteristic (C), and o a sub-step of storing (322), in the updated database of each re-evaluated conformity indicator (I), an alert step (330),in which an alert is issued to a user if at least one re-evaluated compliance indicator (I) violates a predefined compliance criterion established from the evaluation rules., 4. Method according to any one of claims 1 to 3, in which an edition date is associated with each characteristic (C), the edition date being the date on which the characteristic (C) was measured, determined or established, the use phase (200) further comprising: a prediction step (230), for each component, of a risk of non-compliance at a predefined time horizon, from the results of the evaluation rules and the edition date of each characteristic (C) associated with the component, and a warning step (240), during which a warning is issued to a user if at least one risk of non-compliance is greater than a predefined threshold.

5. Method according to any one of claims 1 to 4, in which the data structure of the converted database (B) is a semantic graph.

6. Method according to any one of claims 1 to 5, in which each evaluation rule is specific to at least one component of the converted database, at least one rule of the set of evaluation rules depending on the specific use of the associated component(s).

7. Method according to any one of claims 1 to 6, in which at least one conformity indicator (I) quantifies the confidence given to one of the characteristics (C) of the component to describe an aspect of the component.

8. Method according to any one of claims 1 to 7, further comprising a phase (400) of integrating a new component into the converted database (B), comprising: a step (410) of receiving identification data of the new component and new specific evaluation rule(s), a step (420) of integrating, into the modified structure of the converted database (B) the identification data, and the new evaluation rule(s), the integration comprising the establishment of a semantic link between characteristic(s) (C) of the converted database and the identification data of the new component, and a step (430) of evaluating the conformity indicator (I) associated with the new component by applying the new evaluation rule(s) to the characteristic(s) (C) linked to the identification data.

9. Method according to any one of the preceding claims, in which at least one of the components of the converted database (B) is a seal whose specific use is the manufacture or maintenance of a waterproof housing intended to be immersed in water to a predetermined depth to prevent water infiltration into the housing despite overpressure induced by the water on an exterior of the housing, a characteristic (C) of said component being a measured resistance to overpressure, one of the evaluation rules being capable of calculating a difference between the measured resistance to overpressure and a predefined value depending on the predetermined depth, the conformity indicator (I) obtained by application of said evaluation rule being a function of the calculated difference.

10. Method according to the preceding claim, in which another component of the converted database (B) is the same seal for a distinct specific use, the distinct specific use being the manufacture or maintenance of an aircraft fuselage to prevent air leaks from an interior of the fuselage to an exterior where the atmospheric pressure is lower, a characteristic (C) of said component being a measured resistance to depression, one of the evaluation rules being capable of calculating a difference between the measured resistance to depression and a predefined value depending on a predefined maximum flight altitude of the aircraft, and the conformity indicator (I) obtained by application of said evaluation rule being a function of the calculated difference.

11. Method according to any one of the preceding claims, in which at least one of the components of the converted database (B) is a capacitor whose specific use is the reconditioning of a switching power supply, the characteristics (C) of said component comprising: a measurement of capacitance of said component, a thermal dispersion of said component and / or one or more acceptable soldering technologies, one of the evaluation rules being specific to the technical characteristic considered and to the context of use defined by a user, the conformity indicator (I) obtained by application of said rule.

12. Method according to any one of the preceding claims, in which at least one of the components of the converted database (B) is a source code whose specific use is the exploitation of Open Source software, a characteristic (C) of said component being a version of the software, one of the evaluation rules evaluating the correspondence between the source code and the version of the software, the conformity indicator (I) obtained by application of said rule being representative of a correspondence between the version and the source code.

13. Semantic database for tracking conformity indicators (I) of components (30), the components being intended to be used for a specific use, the specific use being the manufacture, maintenance or reconditioning of mechanical or electronic devices (D) comprising at least one of the components, conformity indicators (I) for use determining whether or not the component is suitable for use for the specific use, the compliance indicator tracking database (I) being capable of being obtained by a generation method according to any one of the preceding claims.

14. Computer program product comprising software instructions which, when implemented by a computer, implement a method according to any one of claims 1 to 12.

15. Information medium comprising program instructions forming a computer program, the computer program being loadable onto a data processing unit and implementing a method according to any one of claims 1 to 12 when the computer program is implemented on the data processing unit.

Images