Method and apparatus for checking tyres

Abstract

Method for checking tyres, comprising: providing a plurality of first mould drawings (10) for moulding corresponding first tyres; providing a plurality of acquired images (20) of said corresponding first tyres; providing a generative artificial intelligence system (GenAI); executing an operation of training said generative artificial intelligence system (GenAI), wherein said first mould drawings (10) and said acquired images (20) are inputted to said generative artificial intelligence system (GenAI); after said training operation, inputting to said generative artificial intelligence system (GenAI) a plurality of second mould drawings (60) of corresponding second tyres; obtaining from said generative artificial intelligence system (GenAI), as output, a plurality of resulting images (70), generated on the basis of said second mould drawings (60); training an anomaly detection neural network (NN) on the basis of said resulting images (70); moulding and curing a finished tyre (FT), and preparing said finished tyre (FT) for checking; taking one or more images of said finished tyre (FT), thereby obtaining one or more corresponding taken images (50); checking said finished tyre (FT) by Inputting said one or more taken images (50) to said anomaly detection neural network (NN); generating a notification signal (NS) through said anomaly detection neural network (NN) when one or more anomalies are detected.

Description

[0001] " METHOD AND APPARATUS FOR CHECKING TYRES"

[0002] DESCRIPTION

[0003] The present invention relates to a method for checking tyres.

[0004] The present invention further relates to an apparatus for checking tyres.

[0005] A tyre for vehicle wheels generally comprises a carcass structure including at least one carcass ply having respectively opposite end flaps in engagement with respective annular anchoring structures, generally referred to as "bead cores", integrated into the regions usually identified as "beads", the inside diameter of which substantially matches a so-called "fitting diameter" of the tyre for fitting it onto a respective rim. The tyre also comprises a crown structure including at least one belt strip located in a radially external position relative to the carcass ply, and a tread band which is radially external to the belt strip. Between the tread band and the belt strip(s) a so-called "underlayer" of elastomeric material may be interposed, the properties of which are suitable for providing a stable union between the belt strip(s) and the tread band. In addition, respective sidewalls of elastomeric material are applied to the side surfaces of the carcass structure, each extending from one of the side edges of the tread band to the respective annular bead anchoring structure. In tyres of the "tubeless" type, the carcass ply is internally coated with a layer of elastomeric material, preferably a butyl-based one, commonly referred to as "liner", which has air tightness properties and extends from one bead to the other.

[0006] A "mould drawing" is a graphics file representative of the shape and dimensions of the mould. The mould drawing may be a graphics file containing the mould design in its final version, which is used for producing the mould, e.g. a CAD (Computer Aided Design) graphics file.

[0007] The tyre moulding and curing process has a dual purpose of giving the tyre its final mechanical / structural characteristics and forming a series

[0008] . i.of symbols on the tyre sidewalls.

[0009] Such symbols may represent technical and / or commercial information, such as, for example: manufacturer's name / logo, tyre size, tyre type, intended season, possible certifications, etc.

[0010] Typically, the outcome of the process of moulding and curing the sidewalls of a tyre is verified by a human operator who, through a visual inspection, checks whether there are any anomalies / defects.

[0011] Document EP 1901059 describes a method of creating master data, i.e. reference data, used for inspecting a concave-convex figure, like that of a tyre. From the figure obtained from tyre CAD drawings, an image is cut out; the position of the cut-out image is then identified within the tyre. Then, the cut-out image is transformed into a grey-scale image based on the thickness of the various portions; finally, the cut-out image is further processed to be used for inspecting the concave-convex figure of a tyre.

[0012] The Applicant observes that the method described in EP 1901059, while it represents an improvement over traditional visual inspections, still suffers from a number of criticalities.

[0013] According to EP 1901059, in fact, the reference images (master data) are created artificially by applying grey scales to images derived from the mould drawing. As perceived by the Applicant, reference images thus created, although processed by considering also the shape of the tyre, are not sufficiently realistic and cannot be effectively used as terms of comparison for acquired images of a tyre to be inspected.

[0014] Moreover, when the technique described in EP 1901059 is applied, it is only possible to generate one reference image starting from a given CAD drawing, thus setting a limit to the number of master images that can be created and reducing, as a result, the overall accuracy of the inspection system.

[0015] In this frame, the Applicant has realized that the generation ofreference images to be used for tyre inspections can be improved by also taking into account real images acquired from tyres already manufactured.

[0016] The Applicant has also realized that using one or more neural networks for the creation of said reference images can lead to a considerable increase in the efficiency of the generation process.

[0017] Lastly, the Applicant has found that, by training a generative artificial intelligence system on CAD images of tyre moulds and corresponding real images acquired from tyres without defects built by using such moulds, it is possible to generate extremely realistic reference images in an efficient manner. The images thus obtained can then be used for training other neural networks, preferably intended for executing anomaly detection tasks on tyres to be checked.

[0018] In accordance with a first aspect, the invention relates to a method for checking tyres.

[0019] Preferably, a plurality of first mould drawings for moulding corresponding first tyres are provided.

[0020] Preferably, it is envisaged to provide a plurality of acquired images of said corresponding first tyres.

[0021] Preferably, it is envisaged to provide a generative artificial intelligence system.

[0022] Preferably, it is envisaged to execute an operation of training said generative artificial intelligence system.

[0023] Preferably, it is envisaged to input said first mould drawings to said generative artificial intelligence system.

[0024] Preferably, it is envisaged to input said acquired images of said corresponding first tyres to said generative artificial intelligence system.

[0025] Preferably, it is envisaged to input a plurality of second mould drawings of corresponding second tyres to said generative artificial intelligence system.Preferably, said second mould drawings are inputted to said generative artificial intelligence system after said training operation.

[0026] Preferably, it is envisaged to obtain as output, from said generative artificial intelligence system, a plurality of resulting images.

[0027] Preferably, said resulting images are generated on the basis of said second mould drawings.

[0028] Preferably, it is envisaged to train an anomaly detection neural network.

[0029] Preferably, said anomaly detection neural network is trained on the basis of said resulting images.

[0030] Preferably, it is envisaged to mould and cure a finished tyre.

[0031] Preferably, it is envisaged to prepare said finished tyre for checking. Preferably, it is envisaged to take one or more images of said finished tyre, thereby obtaining one or more corresponding taken images.

[0032] Preferably, it is envisaged to check said finished tyre.

[0033] Preferably, in order to carry out said check, it is envisaged to input said one or more taken images to said anomaly detection neural network, Preferably, it is envisaged to generate a notification signal by means of said anomaly detection neural network when one or more anomalies are detected.

[0034] In accordance with a second aspect, the invention relates to an apparatus for checking tyres.

[0035] Preferably, a processing module is provided.

[0036] Preferably, said processing module comprises a generative artificial intelligence system.

[0037] Preferably, said generative artificial intelligence system is trained by executing a training operation.

[0038] Preferably, during said training operation, said generative artificial intelligence system is inputted a plurality of first mould drawings for moulding corresponding first tyres.Preferably, during said training operation, said generative artificial intelligence system is inputted a plurality of acquired images of said corresponding first tyres,

[0039] Preferably, an input module is provided.

[0040] Preferably, said input module is associated with said processing module.

[0041] Preferably, said input module is configured per receiving or acquiring one or more taken images of a finished tyre.

[0042] Preferably, said input module is configured for receiving, as input, a plurality of second mould drawings of corresponding second tyres, Preferably, said processing module is configured for generating, through said generative artificial intelligence system, a plurality of resulting images.

[0043] Preferably, said resulting images are generated on the basis of said second mould drawings.

[0044] Preferably, said processing module comprises an anomaly detection neural network,

[0045] Preferably, said anomaly detection neural network is trained on the basis of said resulting images.

[0046] Preferably, said processing module is configured for checking said finished tyre.

[0047] Preferably, in order to carry out said check, said processing module inputs one or more taken images to said anomaly detection neural network.

[0048] Preferably, said anomaly detection neural network is trained to generate a notification signal when one or more anomalies are detected.

[0049] The Applicant believes that the proposed solution makes it possible to reliably, effectively, and efficiently in terms of execution times, detect any defects / anomalies that may be present on the sidewall of a tyre, in particular because of the way in which the resulting images aregenerated.

[0050] In one or more of the above aspects, the invention may comprise one or more of the following preferred features.

[0051] Preferably, said resulting images are representative of tyres with substantially no defects.

[0052] Preferably, each one of said first mould drawings, inputted to said generative artificial intelligence system during said training operation, comprises a plurality of drawing portions.

[0053] Preferably, said drawing portions provide, combined together, the corresponding first mould drawing.

[0054] Preferably, each one of said acquired images of said corresponding first tyres, inputted to said generative artificial intelligence system during said training operation, comprises a plurality of image portions.

[0055] Preferably, said image portions provide, combined together, the corresponding acquired image.

[0056] Preferably, an indication of correspondence between each portion of a mould drawing and a corresponding portion of an acquired image is inputted to said generative artificial intelligence system.

[0057] Preferably, said indication of correspondence is inputted to said generative artificial intelligence system during said training operation.

[0058] Preferably, said generative artificial intelligence system is trained to output a plurality of image parts.

[0059] Preferably, said image parts form, combined together, said at least one generated image.

[0060] Preferably, it is envisaged to input a further mould drawing for manufacturing a further tyre to said generative artificial intelligence system.

[0061] Preferably, said further mould drawing is inputted to said generative artificial intelligence system after said training operation.

[0062] Preferably, said further tyre is different from said corresponding firsttyres.

[0063] Preferably, said generative artificial intelligence system outputs at least one generated image of said further tyre.

[0064] Preferably, said finished tyre is moulded and cured by using said mould having the drawing of said further tyre.

[0065] Preferably, it is envisaged to make a comparison between said one or more taken images and said at least one generated image.

[0066] Preferably, said comparison is made in order to identify one or more defects of said finished tyre.

[0067] Preferably, said notification signal is generated also as a function of said comparison.

[0068] Preferably, said input module is configured for receiving, as input, a further mould drawing for manufacturing a further tyre, which is different from said corresponding first tyres.

[0069] Preferably, said finished tyre is moulded and cured by using said mould having the drawing of said further tyre.

[0070] Preferably, said processing module is configured for generating, through said generative artificial intelligence system, at least one generated image of said further tyre.

[0071] Preferably, said processing module is configured for making a comparison between said one or more taken images and said at least one generated image.

[0072] The images thus obtained are then used as sample images for direct comparison with images acquired from tyres to be checked.

[0073] Preferably, said comparison is made in order to identify one or more defects of said finished tyre.

[0074] Preferably, said processing module is configured for generating said notification signal also as a function of said comparison.

[0075] Further features and advantages will become more apparent in the light of the following detailed description of a preferred, but non-limiting,embodiment of the invention. Such description is provided herein with reference to the annexed drawings, which are also supplied by way of non-limiting example, wherein:

[0076] - Figure 1 shows a block diagram of an apparatus according to the present invention;

[0077] - Figures 2-4 show images employed in embodiments of the present invention;

[0078] - Figure 5 schematically shows a first training process used in embodiments of the present invention;

[0079] - Figure 6 schematically shows a first inspection operation executed in embodiments of the present invention;

[0080] - Figure 7 schematically shows a second training process used in embodiments of the present invention;

[0081] - Figure 8 schematically shows a second inspection operation executed in embodiments of the present invention.

[0082] With reference to the accompanying figures, numeral 1 designates as a whole an apparatus for checking tyres.

[0083] The apparatus 1 (Figure 1) comprises an input module 2, configured for receiving or acquiring taken images 50 representative of a finished tyre FT. The finished tyre FT is a moulded and cured tyre.

[0084] In particular, the taken images 50 are representative of a sidewall of the finished tyre FT.

[0085] The taken images 50 can be obtained, for example, through an acquisition operation executed by means of one or more suitable acquisition devices 3, such as one or more cameras, included in or associated with said input module 2.

[0086] Each taken image 50 can be obtained from multiple acquisitions that, when combined together, will form the respective taken image 50. Each acquisition may target a respective sub-part of the finished tyre FT.

[0087] One or more lighting devices 4, appropriately positioned andadjusted to maximize the quality of the acquired images, may be associated with the acquisition devices 3.

[0088] In particular, the acquisition devices 3 and the associated lighting devices 4 may be synchronized to progressively acquire different portions of the finished tyre FT (in particular, of at least one sidewall of the finished tyre FT), so as to make up the taken images 50.

[0089] The apparatus 1 further comprises a processing module 5, associated with the input module 2.

[0090] The processing module 5 may be realized as one or more processors associated with one or more respective memory areas and configured to execute the operations described below.

[0091] For example, the processing module 5 may apply one or more compensation algorithms to the taken images 50 so as to compensate for any deformations (whether small or big) of the finished tyre FT. It is in fact necessary to consider that, during the acquisition process, the finished tyre FT is not typically mounted on a rim and, therefore, is not inflated. For example, the finished tyre FT may be lying on a support table (not shown). As a consequence, the finished tyre FT may not have a perfectly regular shape, and this may lead to obtaining taken images 50 not showing its exact appearance (e.g. writings, logos, etc.). By way of example, one may use the compensation technique described in international patent application WO 2015 / 097635 Al in the name of the present Applicant.

[0092] Preferably, the taken images 50 are grey-scale images.

[0093] The processing module 5 comprises a generative artificial intelligence system GenAI.

[0094] In accordance with the invention, a specific training operation is envisaged for training the generative artificial intelligence system GenAI (Figure 5).

[0095] A plurality of first mould drawings 10 are provided, for mouldingcorresponding first tyres; in addition, a plurality of acquired images 20 of said corresponding first tyres are also provided,

[0096] Advantageously, the acquired images 20 are representative of tyres with substantially no defects.

[0097] In practice, acquired images 20 of first tyres already built (i.e. previously moulded and cured) are provided along with the first mould drawings 10 used for moulding and curing such first tyres.

[0098] In one embodiment, each one of the first mould drawings 10 comprises a plurality of drawing portions; such drawing portions provide, combined together, the corresponding first mould drawing,

[0099] Likewise, each one of the acquired images 20 of the corresponding first tyres comprises a plurality of image portions; such image portions provide, combined together, the corresponding acquired image.

[0100] The first mould drawings 10 and the acquired images 20 of the corresponding first tyres are Inputted to the generative artificial intelligence system GenAI; in particular, the generative artificial intelligence system GenAI is trained by indicating the correspondence between each mould drawing and the acquired image(s) of the respective tyre.

[0101] During the operation of training the generative artificial intelligence system GenAI, the latter is inputted the following: (i) the portions of the first mould drawings 10; (ii) the portions of the acquired images 20; (iii) an indication of correspondence between each portion of a first mould drawing and a corresponding portion of an acquired image.

[0102] In this manner, the generative artificial intelligence system GenAI will learn the correspondence between a mould drawing and the image (or the images) of the corresponding tyre; such correspondence is established between the mould drawing portions and the acquired image portions.

[0103] As a generative artificial intelligence system GenAI, one may use a so-called Generative Adversarial Network, GAN, wherein two neural

[0104] - io -networks, i.e. a generator and a discriminator, compete with each other. A known GAN is described in the paper available at the URL:

[0105]

[0106] Networks (last access: December 18, 2024).

[0107] By way of example, Figure 2 shows a schematized representation of a first mould drawing 10. As can be seen, such representation comprises a decorated zone 11, writings of different sizes 12.1, 12.2, logos 13, and a substantially plain background 14.

[0108] Figure 3 shows a schematized representation of an acquired image 20 of a tyre moulded and cured with a mould corresponding to the drawing schematized in Figure 2. In Figure 3, one can see a first group of pixels 21, which represents a decorated zone, a second group of pixels 22.1, which represents a writing having a relatively big size, a third group of pixels 22.2, which represents a writing having a relatively small size, a fourth group of pixels 23, which represents two logos, and a fifth group of pixels 24, which represents a substantially plain background of the tyre sidewall.

[0109] Once the generative artificial intelligence system GenAI has been trained, it can be operatively used to assist the inspection of other tyres - i.e. tyres that are different from those used for the training operation.

[0110] In particular, the generative artificial intelligence system GenAI is used for checking said finished tyre FT.

[0111] More specifically, the generative artificial intelligence system GenAI can be used for training an anomaly detection neural network NN (Figure 7), so that the latter can detect any defects / anomalies of the finished tyre FT.

[0112] To this end, the generative artificial intelligence system GenAI is inputted, through the input module 2, a plurality of second mould drawings 60 of corresponding second tyres.

[0113] The second mould drawings 60 are similar to the one schematically

[0114] - ii -illustrated in Figure 2.

[0115] The corresponding second tyres are tyres that are different from said corresponding first tyres, that is, the tyres whose images and mould drawings 20, 10 have been used for training the generative artificial intelligence system GenAI.

[0116] Preferably, the second tyres are not actually built (i.e. they are not moulded and cured).

[0117] Having received, as input, the second mould drawings 60, the generative artificial intelligence system GenAI will generate, as output, corresponding resulting images 70.

[0118] The resulting images 70 simulate real images (similar to the one schematized in Figure 3) acquired from the second tyres - which, as aforesaid, may even not be actually moulded and cured.

[0119] In one embodiment, the second mould drawings 60 and the resulting images 70 are all divided into sub-parts.

[0120] The anomaly detection neural network NN is trained by inputting thereto the resulting images 70 (possibly divided into sub-parts),

[0121] The resulting images 70 represent tyres that have substantially no defects; therefore, the anomaly detection neural network NN will learn to recognize substantially flawless tyres and, by contrast, to detect any anomalies or defects.

[0122] The anomaly detection neural network NN thus trained can then be used for checking moulded and cured tyres (Figure 8).

[0123] In particular, the anomaly detection neural network NN can be used for checking the finished tyre FT on the basis of the respective taken images 50. When one or more anomalies are detected, the anomaly detection neural network NN will generate a notification signal NS.

[0124] The notification signal NS may be generated in the form of a visual indication displayed on an operator's screen, which shows the detected anomaly, preferably in the position where it was detected.Known anomaly detection neural networks are described in the scientific paper accessible through the following URL:. MyTec.. AP.„. A

[0125]

[0126] ComDrehensiveJ^abWoridjyataseL.fo^ (last access: December 18, 2024).

[0127] In addition or as an alternative, the inspection operation may be conducted on further images 80 (Figure 4) representative of a tyre different from the finished tyre FT and different from said first tyres and second tyres.

[0128] In this case as well, when an anomaly X is detected, the processing module 5 will generate the notification signal NS representative of said anomaly and also, preferably, of its position.

[0129] Preferably, the finished tyre FT is moulded and cured with a respective mould having a respective drawing 30. It can therefore be stated that such mould has the further drawing of a further tyre, i.e. a tyre different from the tyres used during the operation of training the generative artificial intelligence system GenAI.

[0130] The further mould drawing 30 used for the finished tyre FT is of the type schematized in Figure 2.

[0131] For the purpose of conducting a further check (Figure 6), the further mould drawing 30 used for the finished tyre FT is inputted to the generative artificial intelligence system GenAI to obtain a generated image 40 of said further tyre.

[0132] The generated image 40 is similar to the image schematized in figure 3.

[0133] In particular, the generative artificial intelligence system GenAI is trained to output a plurality of image parts that, combined together, form the generated image 40. Such image parts correspond to said portions of first mould drawings 10 and portions of acquired images 20.

[0134] Ideally, the finished tyre FT should coincide with said further tyre, since both correspond to one same mould; the difference lies in the factthat the finished tyre FT is the result of an actual moulding and curing process (thus being subject to potential errors, anomalies, etc.), whereas the further tyre is a theoretical tyre with no defects and perfectly corresponding to the further mould drawing 30.

[0135] The processing module 5 is configured for making a comparison between the taken images 50 (representative of the finished tyre FT) and the at least one generated image 40 (obtained from the generative artificial intelligence system GenAI). It will thus be possible to confirm any defects already detected in the finished tyre FT by the anomaly detection neural network NN, or to identify other defects.

[0136] Figure 4 schematically shows a taken image 50 which also includes, in addition to the decoration, the writings and the logos, an anomaly X.

[0137] In particular, the taken image 50 is divided into a plurality of parts, and each part of the taken image is compared with a corresponding part of the generated image 40.

[0138] When one or more defects are detected, the processing module 5 will generate the notification signal NS also on the basis of said comparison.

Claims

CLAIMS1. Method for checking tyres, comprising:providing a plurality of first mould drawings (10) for moulding corresponding first tyres;providing a plurality of acquired images (20) of said corresponding first tyres;providing a generative artificial intelligence system (GenAI); executing an operation of training said generative artificial intelligence system (GenAI), wherein said first mould drawings (10) and said acquired images (20) are inputted to said generative artificial intelligence system (GenAI);after said training operation, inputting to said generative artificial intelligence system (GenAI) a plurality of second mould drawings (60) of corresponding second tyres;obtaining, as the output of said generative artificial intelligence system (GenAI), a plurality of resulting images (70) generated on the basis of said second mould drawings (60);training an anomaly detection neural network (NN) on the basis of said resulting images (70);moulding and curing a finished tyre (FT), and preparing said finished tyre (FT) for checking;taking one or more images of said finished tyre (FT), thereby obtaining one or more corresponding taken images (50);checking said finished tyre (FT) by inputting said one or more taken images (50) to said anomaly detection neural network (NN);if one or more anomalies are detected, generating a notification signal (NS) by means of said anomaly detection neural network (NN).

2. Method according to claim 1, wherein said resulting images (70) are representative of tyres with substantially no defects.

3. Method according to claim 1 or 2, wherein:each one of said first mould drawings (10), inputted to said generative artificial intelligence system (GenAI) during said training operation, comprises a plurality of drawing portions, wherein said drawing portions provide, combined together, the corresponding first mould drawing (10).

4. Method according to any one of the preceding claims, wherein each one of said acquired images (20) of said corresponding first tyres, inputted to said generative artificial intelligence system (GenAI) during said training operation, comprises a plurality of image portions, wherein said image portions provide, combined together, the corresponding acquired image (20).

5. Method according to claims 3 and 4, wherein, during said training operation, an indication of correspondence between each portion of a first mould drawing (10) and a corresponding portion of an acquired image (20) is inputted to said generative artificial intelligence system (GenAI).

6. Method according to any one of the preceding claims, comprising:after said training operation, inputting to said generative artificial intelligence system (GenAI) a further mould drawing (30) for manufacturing a further tyre different from said corresponding first tyres, so that said generative artificial intelligence system (GenAI) will output at least one generated image (40) of said further tyre;wherein said finished tyre (FT) is moulded and cured by using said mould having the drawing of said further tyre;wherein said method comprises making a comparison between said one or more taken images (50) and said at ieast one generated image, in order to identify one or more defects of said finished tyre (FT);said notification signal (NS) being generated also as a function of said comparison.

7. Method according to claim 6, wherein said generative artificial intelligence system (GenAI) is trained to output a plurality of image parts that, combined together, form said at least one generated image (40).

8. Apparatus for checking tyres, comprising:a processing module (5), comprising a generative artificial intelligence system (GenAI), wherein said generative artificial intelligence system (GenAI) is trained during a training operation, wherein in said training operation a plurality of first mould drawings (10), for moulding corresponding first tyres, and a plurality of acquired images (20) of said corresponding first tyres are inputted to said generative artificial intelligence system (GenAI);an input module (2), associated with said processing module and configured for receiving or acquiring one or more taken images (50) of a finished tyre (FT);wherein:said input module (2) is configured for receiving, as input, a plurality of second mould drawings (60) of corresponding second tyres;said processing module (5) is configured for generating, by means of said generative artificial intelligence system (GenAI), a plurality of resulting images (70) generated on the basis of said second mould drawings (60);said processing module (5) comprises an anomaly detection neural network (NN), wherein said anomaly detection neural network (NN) istrained on the basis of said resulting images (70);said processing module (5) is configured for checking said finished tyre (FT) by inputting said one or more taken images (50) to said anomaly detection neural network (NN);said anomaly detection neural network (NN) is trained to generate a notification signal (NS) when one or more anomalies are detected.

9. Apparatus according to claim 8, wherein said resulting images (70) are representative of tyres with substantially no defects.

10. Apparatus according to claim 8 or 9, wherein:each one of said first mould drawings (10), inputted to said generative artificial intelligence system (GenAI) during said training operation, comprises a plurality of drawing portions, wherein said drawing portions provide, combined together, the corresponding first mould drawing (10).

11. Apparatus according to any one of claims 8-10, wherein each one of said acquired images (20) of said corresponding first tyres, inputted to said generative artificial intelligence system (GenAI) during said training operation, comprises a plurality of image portions, wherein said image portions provide, combined together, the corresponding acquired image (20).

12. Apparatus according to claims 10 and 11, wherein, during said training operation, an indication of correspondence between each portion of a first mould drawing (10) and a corresponding portion of an acquired image (20) is inputted to said generative artificial intelligence system (GenAI).

13. Apparatus according to any one of the preceding claims, wherein:said input module (2) is configured for receiving, as input, a further mould drawing (30) for manufacturing a further tyre different from said corresponding first tyres, wherein said finished tyre (FT) is moulded and cured by using said mould having the further drawing (30) of said further tyre;said processing module (5) is configured for generating, through said generative artificial intelligence system (GenAI), at least one generated image (40) of said further tyre;wherein said processing module (5) is configured for:making a comparison between said one or more taken images (50) and said at least one generated image (40), in order to identify one or more defects of said finished tyre (FT);generating said notification signal (NS) also as a function of said comparison.

14. Apparatus according to claim 13, wherein said generative artificial intelligence system (GenAI) is trained to output a plurality of image parts that, combined together, form said at least one generated image (40).

Images