Method and apparatus for predicting print quality

Abstract

There is provided a method, the method comprising obtaining print data, the print data indicative of a mark to be applied to a substrate by a printing apparatus and determining predicted print quality data, the predicted print quality data indicative of a predicted print quality of one or more prints of the mark when applied to the substrate using the printing apparatus with one or more sets of printer settings, wherein determining predicted print quality data comprises processing the print data using a machine learning model.

Description

[0001] Method and apparatus for predicting print quality

[0002] Technical Field

[0003] The present disclosure relates to predicting, using a machine learning model, a print quality for marks to be applied by a printing apparatus.

[0004] Background

[0005] Industrial printers are those found in an industrial setting, such as those used on production lines, packaging lines, filling lines, etc. For example, industrial printers may be used to mark products being conveyed along a conveyor. Of course, conveyors may not be used, and the products may be manually presented to the industrial printer for in another manner. For example, the product may comprise a continuous film or foil which is advanced along a path adjacent the industrial printer. Examples of industrial printers are thermal transfer printers, laser coders, continuous ink jet printers, drop on demand printers, etc.

[0006] Industrial printers are highly complex machines, and typically have a number of adjustable printer settings. It can be difficult for a user to know which combination of printer settings should be used to achieve a desired print result. User’s may therefore incorrectly set up an industrial printer which can lead to wasted marking medium (e.g. ink), and reduce the operational life of the industrial printer.

[0007] Summary

[0008] In a first aspect, there is provided a method, the method comprising obtaining print data, the print data indicative of a mark to be applied to a substrate by a printing apparatus, and determining predicted print quality data, the predicted print quality data indicative of a predicted print quality of one or more prints of the mark when applied to the substrate using the printing apparatus with one or more sets of printer settings, wherein determining predicted print quality data comprises processing the print data using a machine learning model. Printer settings of a printing apparatus affect the print quality of an applied mark. In some cases, it may be important that the print quality of the mark is high. For example, if the mark is a dosage regime for a medicine, the legibility of the mark must be sufficiently good for a user to read without difficulty. As such, a high print quality will be desired. In other cases, some marks may not require a high print quality in order to be legible. For example, a lot number may be printed with a low print quality and still be legible. However, it can be difficult for an operator of a printing apparatus to select a specific set of printer settings that would result in a desired print quality. This is particularly true of industrial printers, which have a large number of printer settings. Typically, operators of industrial printers will not know which of the printer settings will necessarily affect print quality, nor how changing the printer settings will affect the overall operation of the industrial printer. As such, the operator may be using the industrial printer in a sub-optimal way. The method of the first aspect advantageously provides a way in which a print quality of the applied mark can be predicted using a machine learning model. Predicting print quality removes the need for an operator to experiment with different printer settings, which is wasteful and could damage the printing apparatus, to obtain a desired print quality. For example, based on the predicted print quality data satisfying one or more criteria, printer settings associated with the predicted print quality data may be identified and used.

[0009] The print data may be a digital representation of the mark. The print data may be the data that would be sent to the printing apparatus to print the mark. Alternatively, the print data may comprise image data. For example, the print data may be a digital image of the mark. The print data may be a true representation of what is to be printed. The print data may be capable of being displayed on a display device in order for a user to view.

[0010] The predicted print quality data may comprise image data. The predicted print quality data may comprise one or more images of predicted prints of the mark when printed with the printing apparatus using the one or more printer settings. The image data may be output on a display of a user device for a user to review the predicted print or prints of the mark. The user can then select one of the images that satisfies the user’s quality requirements. The set of printer settings associated with the selected image can then be automatically determined. That is, the set of printer settings on which the prediction is based can be determined. The method may be a computer implemented method. The steps of obtaining print data and determining predicted print quality data, as well as the steps described below, may be carried out using any suitable computer hardware. The computer hardware may comprise one or more processors capable of executing computer readable instructions. The one or more processors may be located remote from the printing apparatus. For example, the one or more processors may be located at a server. Alternatively, the one or more processors may be located on a local computer, the local computer co-located with the printing apparatus. Alternatively or additionally, the one or more processors may be located at a technical support site.

[0011] The printing apparatus may be an industrial printer. An industrial printer is the type typically found in warehouses and production lines, and is typically used to mark products being conveyed along a conveyor. Examples of industrial printers are thermal transfer printers, laser coders, continuous inkjet printers, drop on demand printers, etc.

[0012] The method may further comprise determining a set of printer settings based on the predicted print quality data.

[0013] For example, a user may review the predicted print quality data associated with a set of printer settings and based on the predicted print quality data, select the set of printer settings that meet the user’s requirements. When the predicted print quality data comprises image data, the image data may be output as an image on a display of a user device as described above. If a single image is displayed of a predicted print of the mark, the prediction made with a specific set of printer settings, the user may select the image as having sufficient quality, and thus the specific set of printer settings may be automatically selected. Where multiple images are displayed for the user, each image displaying a predicted print of the mark when using a different specific set of printer settings, the user can then select one of the images that satisfies the user’s print quality requirements. The printer settings associated with the selected image can then be determined.

[0014] Alternatively, determining a set of printer settings based on the predicted print quality data may be carried out automatically. For example, where it is known what the user’s print quality requirements are, a second machine learning model can process the predicted print quality data in order to determine the set of printer settings. For example, when the predicted print quality data comprises image data as described above, the second machine learning model may process an image of the predicted print of the mark, and categorise the image as acceptable or not acceptable.

[0015] Determining the set of printer settings based on the predicted print quality data may comprise generating image data based on the predicted print quality data, the image data comprising a predicted image of the mark when printed with the set of printer settings, and determining the set of printer settings based on the image data.

[0016] For example, the predicted print quality data may be used to generate image data, e.g. an image file for output on a display of a user device. The predicted print quality data may be a vector, where each element of the vector corresponds to a pixel, and where each value in each element corresponds to an intensity value. The predicted print quality data can then be translated into an image file for display on a display of a user device, such as a computer screen, smart phone, tablet, etc.

[0017] The method may further comprise outputting the image data to a display of a user device, receiving an input from a user, determining the set of printer settings based on the input.

[0018] For example, the user may view a predicted image of a mark (based on the predicted print quality data) on the display of the user device, and decide that the predicted print quality is sufficient. The user may then provide an input to the user device to indicate that they are satisfied with the predicted print quality. On receipt of the input from the user, the set of printer settings associated with the predicted image (e,g, associated with the predicted print quality data) can automatically be determined.

[0019] Outputting the image data to the display of the user device may comprise generating, using the predicted print quality data, one or more user selectable options, each of the one or more user selectable options associated with a predicted image of the mark, each of the one or more predicted images associated with a set of printer settings for the printing apparatus, and outputting the one or more user selectable options to the display, and wherein receiving the input from the user comprises, receiving a selection of one of the one or more user selectable options. Selecting the set of printer settings based on the predicted print quality data may comprises comparing the predicted print quality data to a predefined print quality threshold, identifying one or more sets of printer settings that satisfy the print quality threshold, and selecting one of the set of printer settings that satisfy the print quality threshold.

[0020] The predefined print quality threshold may be a threshold set by the user. The predicted print quality data, such as the predicted images described above, may be compared against the predefined print quality threshold in order to identify the predicted print quality data that satisfy the predefined print quality threshold. Once the predicted print quality data (such as the predicted images) that satisfy the predefined print quality threshold hold are determined, the associated set of printer settings may be obtained. Multiple sets of printer settings may be obtained if multiple predicted images satisfy the predefined print quality threshold. Any suitable method may then be used to select a particular set of printer settings from the multiple set of printer settings that satisfy the threshold, such as selecting the set of printer settings that use the least amount of ink. A machine learning model may be used to carry out the comparison and / or selection.

[0021] The method may further comprise configuring the printing apparatus with the set of printer settings.

[0022] Once the printer settings have been determined, such as when the user selects a particular set of printer settings that give an acceptable predicted print quality, the printing apparatus may automatically be configured with the set of printer settings. The set of printer settings may be transmitted to the printing apparatus. For example, a control signal may be transmitted to the printing apparatus that, when on receipt by the printing apparatus, causes the printing apparatus to automatically configure the printing apparatus with the set of printer settings. Alternatively, the set of printer settings may be provided to the user for manual input to the printing apparatus. The user device may transmit the set of printer settings to the printing apparatus. Alternatively, a server may transmit the set of printer settings to the printing apparatus. Processing the print data using the machine learning model comprises providing as input to the machine learning model the print data, obtaining output data from the machine learning model based on the print data, and determining the predicted print quality data based on the output data.

[0023] The print data may be a digital representation of the mark. The print data may be pre- processed prior to being input into the machine learning model. For example, when the print data comprises image data, the image data may be converted into a vector, where each element of the vector indicates an intensity value of a specific pixel of the image. The dimensions of the image may be reduced to match the input size of the machine learning model. A feature map may be generated based on the print data, where the feature map is used as input to the machine learning model.

[0024] The output data may be the predicted print quality data, or may be data that can be used to determine the predicted print quality data. For example, the output data may be a predicted image of how the mark will look when printed with the printing apparatus. The output data may be a vector, where each element of the vector indicates an intensity value of a specific pixel of the predicted image.

[0025] The method may further comprise providing as input to the machine learning model a set of printer settings, and wherein obtaining output data from the machine learning model comprises obtaining output data from the machine learning model based on the print data and the set of printer settings.

[0026] That is, the machine learning model may take as input both the print data and a set of printer settings, and output the output data. The output data output from the machine learning model may be a predicted image of how the mark (represented by the print data) will look when printed with the printing apparatus using the set of printer settings. The set of printer settings may be pre-processed into a vector for processing by the machine learning model. The print data and the set of printer settings may be concatenated into a single input (e.g. vector) to be processed by the machine learning model.

[0027] The method may further comprise providing as input to the machine learning model substrate data, the substrate data indicative of a property of the substrate, and wherein obtaining output data from the machine learning model comprises obtaining output data from the machine learning model based on the print data, the set of printer settings, and the substrate data.

[0028] That is, the machine learning model may take as input the print data, a set of printer settings, and the substrate data, and output the output data. The property may be any one or more of the material of the substrate, physical dimensions of the substrate, color of the substrate, and graphic data indicating any pre-applied graphics to the substrate. The substrate data may be concatenated with the print data and the set of printer settings for input into the machine learning model.

[0029] The machine learning model may comprise a neural network, such as artificial neural network.

[0030] The neural network may comprise an input layer, one or more hidden layers, and an output layer. The neural network may be embodied in software or hardware.

[0031] The neural network may comprise a deep neural network, a convolutional neural network, a generative neural network, etc. Examples of generative neural networks are transformers, conditional generative adversarial networks, and diffusion models. The neural network may be conditioned on the print data and the set of printer settings.

[0032] The method may further comprise training the machine learning model with a training data set.

[0033] The training data set may comprise training print data and associated training print quality data. For example, training print data may comprise a plurality of example digital representations of a mark, where each example has an associated set of printer settings. The associated training print quality data may be actual data recorded when a printing apparatus executed a printing operation to apply the mark represented by the training print data with the associated set of printer settings. For example, the training print quality data may be one or more captured images of the mark after it has been applied to the substrate. The associated training print quality data therefore represents target data, or ground truth data. Training the machine learning model with the training data set may comprise inputting the training print data into the machine learning model to obtain a training output, comparing the training output to the associated training print quality data and updating the machine learning model based on the comparing.

[0034] For example, comparing the training output to the associated training print quality data may comprise calculating a loss function. The loss function can be used to update the machine learning model such as by using backpropagation to minimise the loss function.

[0035] The method may further comprise inputting, with the training print data, a training set of printer settings into the machine learning model to obtain the output. That is, the input to the machine learning model during training may be the training print data and a training set of printer settings.

[0036] The print data may comprise a digital representation of the mark.

[0037] The mark may be any one or more of a barcode, QR code, lot number, date, graphic, or text.

[0038] The mark may comprises any alpha numeric characters.

[0039] The set of printer settings may comprise any one or more of: distance of the print head, angle and rotation of the print head, encoder setup, ink, marking I label position (HW setup), optical configuration / setup, laser wave length, dependency to the marking object, vibrations, ribbon type / colour, raster, bold, Character Gap, inverse, mirror, invert, print mode, print resolution, jump speed, marking speed, jump delay, mark delay, stroke delay, marking intensity, pulse frequency, on delay, off delay, x / y print position, iAssure reference, print speed, print darkness (heat), print force, printhead position (continous mode only), font, dpi horizontal, dpi vertical, print density, barcode fine adjust, label feed speed, pixeltrimming I pixelreduction (1 D / 2D Code fine adj.), Graphics / Logos, 1D / 2D Code - Parameter-Set.

[0040] Configuring the printing apparatus with the set of printer settings may comprise transmitting the set of printer settings to the printing apparatus. For example, the set of printer settings may be transmitted from a computer hardware (such as a server, or a computer (e.g. user device)), to the printing apparatus via a suitable network, such as the internet.

[0041] The method may further comprise executing a printing operation on the printing apparatus to apply a mark on a substrate using the determined set of printer settings.

[0042] In a second aspect there is provided a method, the method comprising, obtaining print data, the print data indicative of a mark to be applied to a substrate by a printing apparatus, obtaining data indicating an acceptable print quality, processing the print data and data indicating the acceptable print quality by a machine learning model to obtain a predicted set of printer settings, the predicted set of printer settings being a prediction of the printer settings required in order to achieve the acceptable print quality when the mark is applied by the printing apparatus to the substrate.

[0043] Optional aspects of the first aspect may be combined with the second aspect where appropriate.

[0044] In a third aspect there is provided a computing apparatus comprising one or more processors and a memory, the memory storing instructions thereon that when executed by the one or more processors, cause the one or more processors to carry out the method of any of the first or second aspects.

[0045] In a fourth aspect there is provided a system comprising an industrial printer, a user device, and a computing apparatus comprising one or more processors, a memory, the memory storing instructions thereon that when executed by the one or more processors, cause the one or more processors to carry out the method of any of the first or second aspects.

[0046] The industrial printer, user device and computing apparatus may be collocated, or may be remote from each other. The user device may be a user device comprising a display (e.g. computer display for displaying image files). The user device may be controlled by the computing apparatus, or may have its own computing hardware. Brief of drawings

[0047] The present disclosure will now be further described by way of example only with reference to the accompanying drawings, in which:

[0048] Figure 1 shows a schematic diagram of a machine learning model;

[0049] Figure 2 shows a schematic diagram of another machine learning model;

[0050] Figure 3 is a flow diagram of a method disclosed herein;

[0051] Figure 4 is a flow diagram of another method disclosed herein; and

[0052] Figure 5 is a schematic diagram of a computing apparatus capable of carrying out the disclosed methods.

[0053] Detailed

[0054] With reference to Figure 1 , a method of predicting print quality will be described. In Figure 1 , print data 101 and a set of printer settings 102 are input into a machine learning model 103. The machine learning model 103 processes the print data 101 and the set of printer settings 102 and outputs a predicted image 104.

[0055] The print data 101 comprises a digital representation of the mark. For example, a user, or computer, may generate a mark in the form of a digital image which is to be applied to a substrate of a product. The print data 1 therefore comprises an ideal representation of the mark, without any errors, artefacts, blurring, etc. created by the industrial printer during the printing process. The mark may be a barcode, QR code, lot number, date, graphic, or text, for example.

[0056] The set of printer settings 102 comprise a set of values for different printer settings. Each individual industrial printer will have its own particular adjustable printer settings. However, merely as an example, such printer settings may be any one or more of: distance of the print head, angle and rotation of the print head, encoder setup, ink, marking / label position (HW setup), optical configuration / setup, laser wave length, dependency to the marking object, vibrations, ribbon type / colour, raster, bold, Character Gap, inverse, mirror, invert, print mode, print resolution, jump speed, marking speed, jump delay, mark delay, stroke delay, marking intensity, pulse frequency, on delay, off delay, x / y print position, iAssure reference, print speed, print darkness (heat), print force, printhead position (continous mode only), font, dpi horizontal, dpi vertical, print density, barcode fine adjust, label feed speed, pixeltrimming I pixelreduction (1 D / 2D Code fine adj.), Graphics / Logos, 1 D / 2D Code - Parameter-Set.

[0057] The predicted image 104 is an example of predicted print quality data. For example, the predicted image 104 is a digital representation of a prediction of what the mark will look like when it has been applied to the substrate of the product using the set of printer settings 102 (e.g. printed onto the product by the industrial printer when the industrial printer is configured based on the set of printer settings). The predicted image 104 may differ from the ideal representation of the mark represented by the print data 1. For example, a particular set of printer settings 102 may cause the mark to be blurry when compared with the ideal representation. In this case, the predicted image 104 output by the machine learning model 103 would comprises a predicted blurriness caused by the use of that particular set of printer settings 102. While it is described that a predicted image 104 is output by the machine learning model 103, it will be understood that the predicted image 4 may be data representing a predicted image 4. That is, data that can be used to generate and display a predicted image 4 on a display of a user device.

[0058] The predicted image 104 can be provided to a user, such as displayed on a display of a user device. This allows a user to make a determination as to whether the printer settings 102 that were used to generate the predicted image 104 are suitable. For example, a user may require a specific level of quality associated with their print. A slight blurring, or streaking, of the mark may not be an issue for the user, particularly if the mark is still legible. That is, the quality of the mark may be relatively low, but may be of sufficient quality that the user deems to be acceptable. For example, the set of printer settings 102 may be more efficient than a set of printer settings that would give a higher quality print. Merely as an example, the set of printer settings 102 may use less ink than a set of printer settings that would give a higher quality print. Alternatively or additionally, the set of printer settings 102 may reduce the risk of mechanical service issues, than a set of printer settings that would give a higher quality print.

[0059] The set of printer settings 102 may be selected by the user, or may be obtained from a database of predetermined printer settings 102. For example, a database may comprise multiple sets of printer settings for an industrial printer. The sets of printer settings may be categorised. For example, the sets of printer settings may be categorised as high quality, medium quality, low quality. The user may decide that a medium quality is required, and may select a set of printer settings associated with a medium quality. The user can then test the selected set of printer settings by processing them with the machine learning model 103 to determine if the predicted image 104 meets the desired standard.

[0060] If the user is satisfied with the predicted image 104, the set of printer settings 102 associated with generating the predicted image 104 can be identified and can be used to configure the industrial printer. For example, a user may select a set of printer settings marked as “low quality”. The predicted image 104 would then give the user an indication of the quality of the predicted mark (e.g. its visual appearance once printed) when using the low quality settings. The user may decide that the quality is acceptable. Alternatively, the user may decide that the quality is not acceptable, and may instead select a different set of printer settings 102 to input into the machine learning model

[0061] 103 with the print data 1. In this case, the user may select a different set of printer settings marked as “medium quality”, and the predicted image 104 would then give the user an indication of the quality of the predicted mark when using the different set of printer settings. Of course, the user may decide the printer settings themselves, rather than using predefined settings that are categorised.

[0062] While it has been described that a user may decide that the print quality is acceptable, this may be done automatically. For example, the predicted image 104 may be compared with a predefined quality threshold. If the predicted image 104 satisfies the quality threshold, the set of printer settings 102 associated with the predicted image

[0063] 104 are selected. Otherwise, the set of printer settings 102 are rejected. Another set of printer settings 102 may be obtained and processed by the machine learning model 103. A separate machine learning model may be used to carry out the comparison and / or selection. As noted above, once the user is satisfied with the predicted image 104, the set of printer settings 102 associated with generating the predicted image 104 are identified and can be used to configure the industrial printer. For example, the set of printer settings 102 may be provided to the user, such as displayed on a display, for the user to manually configure their industrial printer. Alternatively, the set of printer settings 102 may be transmitted to the industrial printer for automatic configuration. For example, a control signal may be transmitted to the industrial printer that, when on receipt by the industrial printer, causes the industrial printer to automatically configure the industrial printer with the set of printer settings 102.

[0064] Data input into the machine learning model (e.g. the print data 101 and the set of printer settings 102) may be pre-processed in any suitable way. For example, when the print data 101 comprises image data (such as an image of the mark to be printed), the image data may be converted into a vector, where each element of the vector indicates an intensity value of a specific pixel of the image. The dimensions of the image may be reduced to match the input size of the machine learning model. The vector may be directly input into the machine learning model. Alternatively, a feature map may be generated based on the print data, where the feature map is used as input to the machine learning model. The feature map may be generated using a separate machine learning model, such as a separate convolution neural network.

[0065] The set of printer settings 102 may be converted into a vector for processing by the machine learning model 103. For example, each element of the vector may correspond with a specific printer setting, and each value of each element may represent a specific value associated with the specific printer setting.

[0066] The print data 101 and set of printer settings 102 may be concatenated for processing by the machine learning model 103. For example, where the print data 101 and set of printer settings 102 are each converted into vectors, the vectors can be concatenated to provide a single vector for processing by the machine learning model 103. The predicted image 104 may comprise a vector, where each element of the vector indicates an intensity value of a specific pixel of the predicted image. However, any suitable form of output may be used. As well as processing the print data 101 and the set of printer settings 102, the machine learning model 103 may also take as part of its input, substrate data, the substrate data indicative of a property of the substrate of the product onto which the mark is to be applied. For example, the property may be any one or more of the material of the substrate, physical dimensions of the substrate, color of the substrate, and graphic data indicating any pre-applied graphics to the substrate. The substrate data may be concatenated with the print data 101 and the set of printer settings 102 for input into the machine learning model 103.

[0067] The machine learning model 103 may take any suitable form. For example, the machine learning model 103 may be a neural network that comprises an input layer, one or more hidden layers, and an output layer. The neural network may comprise a deep neural network, a convolutional neural network, a generative neural network, etc. Examples of generative neural networks are transformers, conditional generative adversarial networks, and diffusion models.

[0068] The machine learning model 103 may be trained using any suitable method. For example, the machine learning model 103 may be trained with a training data set. The training data set may comprises training print data, training sets of printer settings and ground truth data. For example, a specific entry in the training database may comprise a training print data example, a training set of printer settings, and ground truth data which may comprise actual data recorded when a printing apparatus executed a printing operation to apply the mark represented by the training print data with the training set of printer settings. For example, the ground truth data may be one or more captured images of the mark after it has been applied to the substrate.

[0069] Training the machine learning model 103 may comprise selecting individual training print data and associated set of printer settings from the training data set, processing the training print data and associated set of printer settings with the machine learning model to obtain output data from the machine learning model, and comparing this output data to the ground truth data associated with the selected training print data. For example, where the ground truth data comprises an image of the mark when printed, the output from the machine learning model is a predicted image of the mark when it has been printed (or data that represents a predicted image). Comparing the output data to the ground truth data may comprise calculating a loss function. The loss function can be used to update the machine learning model 103 such as during an optimization step, by using backpropagation to minimise the loss function. That is, weights, or parameters, of the machine learning model may be updated so as to minimise the loss function. Any suitable loss function may be used, such as mean square error.

[0070] The machine learning model 103 may be continually trained. For example, print data 101 and a set of printer settings 102 may be input into the machine learning model 103 to obtain a predicted image 104 as described above. The user may be satisfied with the print quality indicated in the predicted image 104 and the industrial printer may be configured with the set of printer settings 102. An image of the actual mark applied to the substrate may be captured and compared with the predicted image 4. The image may be captured by a vision capture system of the industrial printer. This comparison may allow the machine learning model 4 to be further trained, e.g. by minimising a loss function.

[0071] While it has been described that the machine learning model 103 takes as input a single set of printer settings along with print data 101, in some implementations, the machine learning model 103 may take as input multiple sets of printer settings 102 along with the print data 101 , and output multiple predicted images 104, where each predicted image corresponds to one of the sets of printer settings 102. In this way, the predicted effect of multiple printer settings on the print data 101 may be determined during execution of the machine learning model.

[0072] In an alternative implementation, shown in Figure 2, print data 201 (which may correspond to print data 101) and data indicating an acceptable print quality 202 are input into a machine learning model 203. The machine learning model 203 outputs a predicted set of printer settings 204, which if used by the industrial printer, would result in a print corresponding to the data indicating an acceptable print quality 202. The predicted set of printer settings 204 are therefore a prediction of the printer settings required in order to achieve a desired acceptable print quality, represented by the data indicating an acceptable print quality 202. For example, the data indicating the acceptable print quality 202 may comprise an image of the mark having an acceptable print quality. For example, the image may be computer generated and based on the print data 201. Alternatively, the data indicating an acceptable print quality 202 may comprise a particular category, such as low quality, medium quality or high quality.

[0073] With reference to Figure 3, there is shown a method according to the disclosed subject matter.

[0074] At Step S1, print data is obtained. The print data is indicative of a mark to be applied to a substrate by a printing apparatus. An example of print data is described above.

[0075] At step S2, predicted print quality data is determined. The predicted print quality data is indicative of a predicted print quality of one or more prints of the mark when applied to the substrate using the printing apparatus with one or more sets of printer settings. The predicted print quality data may be the predicted image 104 described above. Determining the predicted print quality data comprises processing the print data using a machine learning model. For example, the print data and associated set of printer settings are processed using the machine learning model to obtain the predicted print quality data as described above with respect to Figure 1.

[0076] With reference to Figure 4, there is shown another method according to the disclosed subject matter.

[0077] At Step S3, print data is obtained. The print data is indicative of a mark to be applied to a substrate by a printing apparatus. An example of print data is described above.

[0078] At Step S4, data indicating an acceptable print quality is obtained. An example of data indicating an acceptable print quality is described above.

[0079] At step S5, the print data and data indicating the acceptable print quality are processed by a machine learning model to obtain a predicted set of printer settings, the predicted set of printer settings being a prediction of the printer settings required in order to achieve the acceptable print quality when the mark is applied by the printing apparatus to the substrate. An example is described above with respect to Figure 2.

[0080] Figure 5 shows a computing apparatus 501 configured to carry out the methods disclosed herein. The computing apparatus 501 comprises a processor 502 which is configured to read and execute instructions stored in a volatile memory 503 which takes the form of a random access memory. The volatile memory 503 stores instructions for execution by the processor 502 and data used by those instructions.

[0081] The computing apparatus 500 further comprises non-volatile storage in the form of a hard disc drive 504. The computing apparatus 500 further comprises an I / O interface 505 to which are optionally connected data capture and peripheral devices used in connection with the computing apparatus 505. In the example shown, a display 506 is connected to the I / O interface 505 to display output from the computing apparatus 505. The display 506 may be provided locally to the computing apparatus 500 (e.g. as a screen), or remotely from the computing apparatus 500. For example, a display associated with a separate device (e.g. a mobile computing device) may be used as a display for the computing apparatus 500. Additionally or alternatively, a touchscreen associated with the display 506 may operate as a user input device, so as to allow a user to interact with the computing apparatus 500. Alternatively or additionally, separate input devices may be also connected to the I / O interface 505, such as a mouse and / or keyboard. A network interface 507 allows the computing apparatus 500 to be connected to an appropriate computer network so as to receive and transmit data from and to other computing devices, such as an industrial printer. The processor 502, volatile memory 503, hard disc drive 504, I / O interface 505, and network interface 507, are connected together by a bus 508. The computing apparatus 500 may be connected to an external computer / server via the network interface 507.

[0082] The computing apparatus 500 may be incorporated into any or all of the industrial printer, user device, and or remote server.

[0083] It will be appreciated that embodiments disclosed herein can be implemented in any convenient form. For example, embodiments disclosed herein may be implemented by appropriate computer programs which may be carried on appropriate carrier media which may be tangible carrier media (e.g. disks) or intangible carrier media (e.g. communications signals). Embodiments disclosed herein may also be implemented using suitable apparatus which may take the form of programmable computers running computer programs arranged to implement the embodiments disclosed herein. Embodiments of the subject matter and the operations described in this specification can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware. Embodiments of the subject matter described in this specification can be implemented as one or more computer programs, i.e., one or more modules of computer program instructions, encoded on computer storage medium for execution by, or to control the operation of, data processing apparatus. Alternatively or in addition, the program instructions can be encoded on an artificially-generated propagated signal, e.g., a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to suitable receiver apparatus for execution by a data processing apparatus. A computer storage medium can be, or be included in, a computer-readable storage device, a computer-readable storage substrate, a random or serial access memory array or device, or a combination of one or more of them. Moreover, while a computer storage medium is not a propagated signal, a computer storage medium can be a source or destination of computer program instructions encoded in an artificially-generated propagated signal. The computer storage medium can also be, or be included in, one or more separate physical components or media (e.g., multiple CDs, disks, or other storage devices).

[0084] The operations described in this specification can be implemented as operations performed by a processor on data stored on one or more computer-readable storage devices or received from other sources.

[0085] The term “processor” encompasses all kinds of apparatus, devices, and machines for processing data, including by way of example a programmable processor, a computer, a system on a chip, or multiple ones, or combinations, of the foregoing. The apparatus can include special purpose reprogrammable logic circuitry, e.g., an FPGA (field programmable gate array) or an ASIC (application-specific integrated circuit). The apparatus can also include, in addition to hardware, code that creates an execution environment for the computer program in question, e.g., code that constitutes processor firmware, a protocol stack, a database management system, an operating system, a cross-platform runtime environment, a virtual machine, or a combination of one or more of them. The apparatus and execution environment can realize various different computing model infrastructures, such as web services, distributed computing and grid computing infrastructures. Processors suitable for the execution of a computer program include, by way of example, both general and special purpose microprocessors, and any one or more processors of any kind of digital computer. Generally, a processor will receive instructions and data from a read-only memory or a random access memory or both. Devices suitable for storing computer program instructions and data include all forms of computer-readable media and memory devices, including by way of example semiconductor memory devices, e.g., EPROM, EEPROM, and flash memory devices; magnetic disks, e.g., internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks. The processor and the memory can be supplemented by, or incorporated in, special purpose logic circuitry and fiber-optic platform for faster data transfer remotely.

[0086] To provide for interaction with a user, embodiments of the subject matter described in this specification can be implemented on a computer having a display device, e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor including audio, for displaying information (e.g. an indication and / or alert) to the user. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback, e.g., visual feedback, auditory feedback, or tactile feedback.

[0087] Although the disclosure has been described in terms of preferred embodiments as set forth above, it should be understood that these embodiments are illustrative only and that the claims are not limited to those embodiments. The skilled person will be able to make modifications and alternatives in view of the disclosure which are contemplated as falling within the scope of the appended claims. Each feature disclosed or illustrated in the present specification may be incorporated in the disclosure, whether alone or in any appropriate combination with any other feature disclosed or illustrated herein.

Claims

CLAIMS:

1. A method, the method comprising: obtaining print data, the print data indicative of a mark to be applied to a substrate by a printing apparatus; determining predicted print quality data, the predicted print quality data indicative of a predicted print quality of one or more prints of the mark when applied to the substrate using the printing apparatus with one or more sets of printer settings, wherein determining predicted print quality data comprises processing the print data using a machine learning model.

2. The method according to claim 1 , the method further comprising determining a set of printer settings based on the predicted print quality data.

3. The method according to claim 2, wherein determining the set of printer settings based on the predicted print quality data comprises: generating image data based on the predicted print quality data, the image data comprising a predicted image of the mark when printed with the set of printer settings; determining the set of printer settings based on the image data.

4. The method according to claim 3, further comprising outputting the image data to a display of a user device; receiving an input from a user; determining the set of printer settings based on the input.

5. The method according to claim 4, wherein outputting the image data to the display of the user device comprises: generating, using the predicted print quality data, one or more user selectable options, each of the one or more user selectable options associated with a predicted image of the mark, each of the one or more predicted images associated with a set of printer settings for the printing apparatus; and outputting the one or more user selectable options to the display; wherein receiving the input from the user comprises: receiving a selection of one of the one or more user selectable options.

6. The method according to claims 2 and 3, wherein selecting the set of printer settings based on the predicted print quality data comprises: comparing the predicted print quality data to a predefined print quality threshold; identifying one or more sets of printer settings that satisfy the print quality threshold; and selecting one of the set of printer settings that satisfy the print quality threshold.

7. The method according to any one of claims 2 to 6, the method further comprising configuring the printing apparatus with the set of printer settings.

8. The method according to any preceding claim wherein processing the print data using the machine learning model comprises: providing as input to the machine learning model the print data; obtaining output data from the machine learning model based on the print data; and determining the predicted print quality data based on the output data.

9. The method according to claim 8, further comprising: providing as input to the machine learning model a set of printer settings; and wherein obtaining output data from the machine learning model comprises obtaining output data from the machine learning model based on the print data and the set of printer settings.

10. The method according to claim 8 or 9, further comprising providing as input to the machine learning model substrate data, the substrate data indicative of a property of the substrate, and wherein obtaining output data from the machine learning model comprises obtaining output data from the machine learning model based on the print data, the set of printer settings, and the substrate data.

11. The method according to any preceding claim wherein the machine learning model comprises a neural network.

12. The method according to any preceding claim, further comprising training the machine learning model with a training data set.

13. The method according to claim 12, wherein the training data set comprises training print data and associated training print quality data.

14. The method according to claim 13, wherein training the machine learning model with the training data set comprises: inputting the training print data into the machine learning model to obtain a training output; comparing the training output to the associated training print quality data updating the machine learning model based on the comparing.

15. The method according to any preceding claim, wherein the print data comprises a digital representation of the mark.

16. The method according to any preceding claim, wherein the mark is any one or more of a barcode, QR code, lot number, date, graphic, or text.

17. The method according to any preceding claim, wherein the set of printer settings comprise any one or more of: distance of the print head, angle and rotation of the print head, encoder setup, ink, marking I label position (HW setup), optical configuration / setup, laser wave length, dependency to the marking object, vibrations, ribbon type / colour, raster, bold, Character Gap, inverse, mirror, invert, print mode, print resolution, jump speed, marking speed, jump delay, mark delay, stroke delay, marking intensity, pulse frequency, on delay, off delay, x / y print position, iAssure reference, print speed, print darkness (heat), print force, printhead position (continous mode only), font, dpi horizontal, dpi vertical, print density, barcode fine adjust, label feed speed, pixeltrimming I pixelreduction (1 D / 2D Code fine adj.), Graphics / Logos, 1 D / 2D Code - Parameter-Set.

18. The method according to any of claims 7 to 17, wherein configuring the printing apparatus with the set of printer settings comprises transmitting the set of printer settings to the printing apparatus.

19. The method according to any preceding claim, the method comprising: executing a printing operation on the printing apparatus to apply a mark on a substrate using the determined set of printer settings.

20. A method, the method comprising: obtaining print data, the print data indicative of a mark to be applied to a substrate by a printing apparatus; obtaining data indicating an acceptable print quality; processing the print data and data indicating the acceptable print quality by a machine learning model to obtain a predicted set of printer settings, the predicted set of printer settings being a prediction of the printer settings required in order to achieve the acceptable print quality when the mark is applied by the printing apparatus to the substrate.

21. A computing apparatus comprising: one or more processors, a memory, the memory storing instructions thereon that when executed by the one or more processors, cause the one or more processors to carry out the method of any preceding claim.

22. A system comprising: an industrial printer; a user device; a computing apparatus comprising: one or more processors, a memory, the memory storing instructions thereon that when executed by the one or more processors, cause the one or more processors to carry out the method of any of claims 1 to 20.

Images