Determination of BALE laydowns based on feedback loop

Abstract

In a computer-implemented method for determining bale laydowns in a textile production process a first bale laydown (310) is provided A computer system (390) receives textile-quality data determined at a location (330) in the textile production process downstream of the first bale laydown (310), the textile-quality data relating to fiber material from the first bale laydown (310). The received textile-quality data are taken into account when determining, by the computer system (390), a second bale laydown (320). Thus, the predictability and consistency of quality characteristics of the textile intermediate products and / or products produced from the bale laydowns (310, 320) is improved.

Description

[0001] DETERMINATION OF BALE LAYDOWNS BASED ON FEEDBACK LOOP

[0002] FIELD OF THE INVENTION

[0003] The present invention lies in the field of quality management in the textile industry. It relates to a computer-implemented method, a computer system, and a computer program for determining bale laydowns in a textile production process, according to the preambles of the independent patent claims.

[0004] DESCRIPTION OF THE PRIOR ART

[0005] Textile raw material such as cotton is usually delivered to a spinning mill in the form of fiber bales. The fiber bales are stored in a warehouse of the spinning mill. From there, they are transported to an opening department of the spinning mill and placed in continuous rows to form a so-called bale laydown. Typically, a bale laydown belongs to a so-called laydown mixing, which is a group of bale laydowns with similar characteristics form which a certain textile product or intermediate product is produced. Fiber bales of various varieties and / or qualities are often collocated to form one bale laydown and / or one laydown mixing in order to obtain a desired fiber blend. They are opened, and fiber tufts are taken off in layers by means of an automatic bale opener.

[0006] DE-40’38'685 Al and US-5,489,028 A each disclose a bale opener provided with a detection device for detecting the presence of foreign material at the surface of the fiber bales.

[0007] The software USTER® FIBERQ optimizes the sourcing and utilization of textile raw material, taking into account inventory, quality, and cotton parameters. It automatically calculates and proposes bale laydowns and laydown mixings and thus offers higher inventory utilization, quality consistency and yield. The raw-material parameters needed as an input are either provided by a cotton-classing authority or measured in the spinning

[0008] - 1 -

[0009] P0640NE-WO mill, e.g., by means of the fiber classification and analysis system USTER® HVI 1000. The parameters can be provided per lot or per individual fiber bale.

[0010] US-5,210,909 A discloses a method for optimizing and maintaining constant fiber quality in automated mixing of bales having fibers of differing qualities. A fiber sample is automatically withdrawn in each case from the bale processed by the bale opener and is analyzed in a measuring and control instrument. The analytical values so measured are continuously compared by a program with nominal values and the comparison measurements so obtained are used to maintain the mixing values by instructing the baleremoving device to remove corresponding quantities from the particular bales. Another sampling device for checking the measured values can be present at the output of a mixer connected to the bale opener.

[0011] A fiber processing method is disclosed in US-6,087,608 A. It includes the steps of removing fiber tufts from a series of fiber bales; pneumatically conveying the fiber tufts in a duct; forming images of the conveyed material as it passes through a first location of the duct; determining the degree of lightness or the color of the fiber tufts by an image processing device; and comparing the measuring values obtained by the image processing device with a settable limit value. If lightness differences in the bale mixture are detected, individual bales, e.g., excessively light or excessively dark bales, can be removed from the laydown.

[0012] According to US-2001 / 049860 Al, textile fibers are supplied in a treatment installation by way of a transport system after one another to various treatment stations. With a sensor system at least two different physical variables are continuously detected at the fibers. From the detected variables certain fiber properties are deduced and actual values are formed which are compared to a nominal value for each fiber property. With the presence of deviations from the nominal value the operating condition at least of one treatment station and / or of the transport system is changed, or there may be set jobs in order to remove or exchange certain bales in the laydown.

[0013] - 2 -

[0014] P0640NE-WO SUMMARY OF THE INVENTION

[0015] It is an object of the present invention to provide a computer-implemented method, a computer system, and a computer program for determining bale laydowns in a textile production process that improve with regard to the state of the art the predictability and consistency of quality characteristics of the textile intermediate products and / or products produced from the bale laydowns. It is a further object to optimize the sourcing and utilization of textile raw material, as well as the management of a bale warehouse.

[0016] These and other objects are solved by the computer-implemented method, the computer system, and the computer program as defined in the independent claims. Advantageous embodiments are specified in the dependent claims.

[0017] The invention is based on the idea of providing a feedback loop from a location in the textile production process downstream of the bale laydown to a computer system determining the bale laydown. Textile-quality data are determined at said location, fed back to the computer system, and taken into account when determining a future bale laydown.

[0018] The computer-implemented method serves for determining bale laydowns in a textile production process. It comprises the steps of: providing at least one first bale laydown; determining, by a computer system, at least one second bale laydown; receiving, by the computer system, textile-quality data determined in the textile production process downstream of the at least one first bale laydown, the textile-quality data relating to fiber material from the at least one first bale laydown; and taking into account the received textile-quality data when determining, by the computer system, the at least one second bale laydown.

[0019] One embodiment further comprises the step of calculating by the computer system a rating metrics from the received textile-quality data and taking into account the rating metrics when determining, by the computer system, the at least one second bale laydown.

[0020] - 3 -

[0021] P0640NE-WO One embodiment further comprises the steps of: storing, in a memory of the computer system, fiber-quality data of the at least one first bale laydown; assigning, by the computer system, the stored fiber-quality data of the at least one first bale laydown to the received textile-quality data of the at least one first bale laydown; and further taking into account the fiber-quality data of the at least one first bale laydown when determining, by the computer system, the at least one second bale laydown.

[0022] In the two embodiments described above, the stored fiber-quality data of the at least one first bale laydown can further be taken into account in the calculation of the rating metrics.

[0023] According to one embodiment, a machine-learning algorithm is used by the computer system for determining the at least one second bale laydown. The machine-learning algorithm can be based on reinforcement learning and the rating metrics can be used as a reward to be maximized.

[0024] According to one embodiment, exactly one second bale laydown is determined by the computer system, and a location parameter and / or a dispersion parameter of textile-quality data determined for fiber material from exactly one first bale laydown is taken into account when determining the second bale laydown.

[0025] According to one embodiment, a second laydown mixing consisting of a plurality of second bale laydowns is determined by the computer system essentially simultaneously, and a location parameter and / or a dispersion parameter of textile-quality data determined for fiber material from a first laydown mixing consisting of a plurality of first bale laydowns is taken into account when determining the second laydown mixing.

[0026] One embodiment further comprises the steps of: storing, in a memory of the computer system, fiber-quality data of at least part of textile-fiber bales from an inventory of textilefiber bales; receiving, by the computer system, at least one specification regarding the fiber-quality data; receiving, by the computer system, a size information about the first bale laydown and the second bale laydown; and further taking into account the stored fiber-quality data, the at least one received specification, and the received size information when determining, by the computer system, the at least one second bale laydown. The

[0027] - 4 -

[0028] P0640NE-WO stored fiber-quality data can comprise values of at least one parameter from the following set: reflectance characteristics, color characteristics, fiber-length characteristics, fiberstrength characteristics, fiber tensile properties, fiber-fineness characteristics, fibermaturity characteristics, micronaire, short-fiber characteristics, trash characteristics, dust characteristics, nep characteristics, foreign-matter characteristics. Preferred parameters are micronaire, yellowness, and reflectance. The at least one received specification regarding the fiber-quality data can comprise a value range for the at least one parameter.

[0029] One embodiment further comprises the steps of storing, in a memory of the computer system, further information for at least part of the textile-fiber bales, the further information being from the following set: bale identifier, plant variety, geographic origin, year of harvest, growth conditions, ginning mill, price, storage location; and further taking into account the stored further information when determining, by the computer system, the at least one second bale laydown. The further information can comprise a bale identifier, and the at least one second bale laydown can be determined by creating a map of the bale laydown indicating positions of single fiber bales and their bale identifiers.

[0030] According to one embodiment, the textile-quality data are determined for fiber material having the form of a single fiber, a fiber tuft, a fiber mat, a sliver, a roving, a yarn, or a fabric.

[0031] According to one embodiment, the determined textile-quality data comprise values of at least one parameter from the following set: reflectance characteristics, color characteristics, fiber-length characteristics, strength characteristics, tensile properties, fiber-fineness characteristics, fiber-maturity characteristics, micronaire, short-fiber characteristics, trash characteristics, dust characteristics, nep characteristics, foreign-matter characteristics, density, mass per unit length, thickness, diameter, roundness characteristics, hairiness characteristics. Preferred parameters are micronaire, yellowness, and reflectance.

[0032] According to one embodiment, the textile-quality data are determined in-line by a production-monitoring instrument or off-line by a laboratory instrument.

[0033] - 5 -

[0034] P0640NE-WO The invention also relates to a computer-implemented method for providing at least one bale laydown in an opening department of a spinning mill from an inventory of textilefiber bales. The method comprises the steps of: determining at least one second bale laydown according to the method described above; and providing the at least one second bale laydown from the inventory of textile-fiber bales according to the determined at least one second bale laydown.

[0035] The invention further relates to a computer system comprising means for carrying out the method as described above.

[0036] The computer system according to the invention serves for determining bale laydowns in a textile production process. It comprises a processor configured to determine at least one bale laydown. The computer system further comprises a receiver for receiving textilequality data determined in the textile production process downstream of at least one first bale laydown, the textile-quality data relating to fiber material from the at least one first bale laydown. The processor is configured to determine at least one second bale laydown, wherein the received textile-quality data are taken into account in the determination of the at least one second bale laydown.

[0037] The invention further relates to a computer program having instructions which when executed by a computer system cause the computer system to perform the method as described above.

[0038] By means of the feedback loop provided according to the invention, the determination of the bale laydowns is continuously improved. The determination of the bale laydowns is based on measured textile-quality data resulting from the bale laydowns, rather than merely on theoretical considerations such as variations of fiber-quality data within a bale laydown and between two bale laydowns. Thus, the predictability and consistency of quality characteristics of the textile intermediate products and / or products produced from the bale laydowns is improved with regard to the state of the art.

[0039] - 6 -

[0040] P0640NE-WO The invention also optimizes the sourcing and utilization of textile raw material by spinning mills, as well as the management of the bale warehouse. Precisely those bales that are needed for an optimum quality and consistency are retrieved from the warehouse.

[0041] BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In the following, the invention is explained in detail based on the drawings.

[0043] Figure 1 schematically shows a part of a textile production process and a computer system according to the invention.

[0044] Figure 2 shows a flowchart for embodiments of the method according to the invention.

[0045] Figures 3 and 4 show block diagrams of embodiments of the method according to the invention.

[0046] Figure 5 shows fictitious time courses of a textile-quality parameter and rating metrics calculated therefrom according to the invention.

[0047] IMPLEMENTATION OF THE INVENTION

[0048] Figure 1 schematically shows a part of a textile production process 100 that takes place in a spinning mill. The material flow is depicted with thick arrows, whereas the data or information flow is depicted with thin arrows. In the textile production process 100, raw fiber material, e.g., raw cotton, is spun to yam, which can further be woven or knitted to a fabric.

[0049] Figure 1 also schematically shows a computer system 190 according to the invention. The computer system 190 can be realized as a single device or as network of several computer devices. It can be located inside and / or outside of the spinning mill. It can be realized by means of cloud computing, i.e., employ remote shared computer resources. The computer system 190 comprises means for carrying out the method according to the invention.

[0050] - 7 -

[0051] P0640NE-WO Textile raw fiber material is delivered to a warehouse in the form of fiber bales and stored there as an inventory 110.

[0052] Fiber-quality data of at least part of the fiber bales, and preferably of all fiber bales form the inventory 110, are stored in a memory of the computer system 190. The fiber-quality data can comprise values of at least one parameter from the following set: reflectance characteristics, color characteristics, fiber-length characteristics, fiber-strength characteristics, fiber tensile properties, fiber-fineness characteristics, fiber-maturity characteristics, micronaire, short-fiber characteristics, trash characteristics, dust characteristics, nep characteristics, foreign-matter characteristics. Preferred parameters are micronaire, yellowness, and reflectance.

[0053] In the embodiment shown in Figure 1, the fiber-quality data are determined by measuring parameters of fiber samples.

[0054] For determining the fiber-quality data, the fiber samples can be taken from the inventory 110. The measurements can be taken in a textile laboratory by means of a fiber-testing instrument 112 such as USTER® HVI 1000 or USTER® AFIS PRO 2. Fiber-quality data of each fiber bale or only of part of the fiber bales can be measured. The fiber bales are usually delivered to the spinning mill in lots, a lot consisting of about 100 fiber bales of the same fiber variety and relatively homogeneous quality. To reduce costs and efforts, it might be sufficient to determine the fiber-quality data of only part of a lot and assign average values of the measured data to all fiber bales of the lot. The fiber-quality data are transmitted to the computer system 190, which is depicted in Figure 1 by an arrow 113.

[0055] Alternatively or additionally, the fiber-quality data can be determined directly from the bale laydown 120. Examples of in-line detection devices mounted on the bale opener processing the bale laydown 120 are given in DE-40’38'685 Al and US-5,489,028 A. The in-line fiber-quality data are transmitted to the computer system 190, which is depicted in Figure 1 by an arrow 121.

[0056] Alternatively or additionally, the fiber-quality data can be determined from fiber samples taken from the bale laydown 120. The measurements can be taken in a textile laboratory by

[0057] - 8 -

[0058] P0640NE-WO means of a fiber-testing instrument 122 such as USTER® HVI 1000 or USTER® AFIS PRO 2. The off-line fiber-quality data are transmitted to the computer system 190, which is depicted in Figure 1 by an arrow 123.

[0059] In an alternative embodiment not shown in Figure 1, the fiber-quality data are determined already prior to the delivery of the fiber bales to the spinning mill, e.g., by the U.S. Department of Agriculture Cotton Classing Services. They are also stored in the computer system 190.

[0060] The computer system 190 determines a bale laydown 120, which is indicated in Figure 1 by arrows 115, 125. Methods for conventionally determining bale laydowns 120 are known from the state of the art, and software performing such methods is available in the market, e.g., the software USTER® FIBERQ.

[0061] The fiber bales needed for the bale laydown 120 are selected from the inventory 110, transported to an opening department of the spinning mill and placed in continuous rows to form the bale laydown 120 as determined by the computer system 190. They are opened 120, and fiber tufts are taken off in layers by means of an automatic bale opener. The selection of the fiber bales forming a bale laydown 120 is a subject-matter of this invention and will be explained in detail below.

[0062] The textile production process 100 may further include, for example, the following process steps: coarse cleaning, blending, fine cleaning, carding, doubling, combing, drafting, spinning, rewinding, weaving, or knitting. Not all the mentioned process steps need to be passed through, and further process steps may be added. For the sake of simplicity, only a few process steps such as the fine cleaning 130, the carding 140, and the rewinding 150 are schematically drawn in Figure 1, while others are indicated by dotted arrows or omitted. The fiber material traversing the textile production process takes various forms, depending on the respective process step, e.g., the form of a fiber bale, a fiber tuft, a fiber mat, a sliver, a roving, a yarn, or a fabric.

[0063] Textile-quality data are determined for the fiber material in the textile production process 100 downstream of the bale laydown 120. The textile-quality data can comprise values of

[0064] - 9 -

[0065] P0640NE-WO at least one parameter from the following set: reflectance characteristics, color characteristics, fiber-length characteristics, strength characteristics, tensile properties, fiber-fineness characteristics, fiber-maturity characteristics, micronaire, short-fiber characteristics, trash characteristics, dust characteristics, nep characteristics, foreign-matter characteristics, density, mass per unit length, thickness, diameter, roundness characteristics, hairiness characteristics. Preferred parameters are micronaire, yellowness, and reflectance.

[0066] In one embodiment, the textile-quality data are determined by an in-line productionmonitoring instrument at some process steps 130, 140, 150 downstream of the bale laydown or between them. Examples of such in-line production-monitoring instruments include a fiber cleaner such as USTER® JOSSI VISION SHIELD, a sliver-monitoring system such as USTER® SLIVERGUARD 3, and a yam clearer such as USTER® QUANTUM 4.0. The in-line textile-quality data are transmitted to the computer system 190, which is depicted in Figure 1 by arrows 131, 141, 151.

[0067] Alternatively or additionally, samples of the fiber material can be segregated from the textile production process 100 at some process steps 130, 140, 150 downstream of the bale laydown or between them. Textile-quality data can be determined for the segregated samples by an off-line laboratory instrument 132, 142, 152. Examples of off-line laboratory instruments include a fiber-testing instrument such as USTER® HVI 1000 or USTER® AFIS PRO 2, and a yam- or sliver-testing instrument such as USTER® TESTER 6. The off-line textile-quality data are transmitted to the computer system 190, which is depicted in Figure 1 by arrows 133, 143, 153.

[0068] An overview of in-line and off-line textile-quality measurement products suitable for application in the invention is given in the brochure “Think Quality™ Product Portfolio”, Uster Technologies AG, 2023. The products are available in the market.

[0069] When the computer system 190 determines a further bale laydown 120, it takes into account the received textile-quality data. The determination of the further bale laydown 120 is depicted in Figure 1 again by arrows 115, 125.

[0070] - 10 -

[0071] P0640NE-WO Embodiments of the method are further described in the following with reference to the flowchart of Figure 2.

[0072] Fiber-quality data of at least part of the fiber bales, and preferably of all fiber bales form the inventory, are stored 201 in a memory of the computer system 190, together with corresponding bale identifiers.

[0073] In the example discussed in the following, it is assumed that a “first” bale laydown and a “second” bale laydown belong to the same laydown mixing.

[0074] The computer system 190 receives 202 at least one specification regarding the fiber-quality data. The specification can comprise, e.g., a value range for the at least one fiber-quality- data parameter each. For instance, the specification can predetermine that the micronaire of all fiber bales within the laydown mixing should be between 3.5 and 3.8.

[0075] The computer system 190 further receives 203 a size information about the first and the second bale laydown. Examples of the size information are a number of fiber bales within the laydown, or a laydown width and a laydown length. A typical bale laydown consists of up to 140 fiber bales, laydown widths typically ranging from three to six fiber bales, and laydown lengths typically ranging from 20 to 40 fiber bales.

[0076] Based on the stored fiber-quality data, the at least one received specification, the received size information, and possibly further information, the computer system 190 determines 204 a laydown mixing and determines 205 a first bale laydown belonging to the laydown mixing.

[0077] The determined first bale laydown is provided 206 in the opening department of the spinning mill and opened by the bale opener.

[0078] The computer system 190 keeps track of the inventory and is updated 207 upon every change of the inventory. The fiber-quality data of the fiber bales used in a bale laydown are deleted from the memory of the computer system 190, since the corresponding fiber bales are no longer available for further bale laydowns.

[0079] - 11 -

[0080] P0640NE-WO The fiber material from the first laydown is further processed 208 as described with reference to Figure 1.

[0081] During the further processing 208, textile-quality data are determined 209 for fiber material from the first bale laydown as described with reference to Figure 1. As shown in Figure 1, the textile-quality data can be determined 209 in-line of off-line at various stages 130, 140, 150 of the textile production process 100 by one or several measuring instruments.

[0082] A laydown rating metrics is calculated 210 from the determined textile-quality data. The laydown rating metrics is explained below with reference to Figure 3.

[0083] According to the invention, the textile-quality data or, in the embodiment discussed here, the laydown rating metrics calculated therefrom is taken into account 212 when determining a second bale laydown. This is realized by an “inner” feedback loop 213 from the textile production process downstream of the first bale laydown to the computer system 190; see also Figures 1, 3 and 4. In the inner feedback loop 213, the textile-quality data or the laydown rating metrics calculated therefrom are fed back to the computer system 190 and used by it for an optimized or improved determination of the second bale laydown. In the flowchart of Figure 2, the determination of the second bale laydown is designated again with the reference numeral 205.

[0084] The described steps 205-213 are repeated 211 for the second and further bale laydowns as long as the same laydown mixing is being processed.

[0085] When the processing of a first laydown mixing is finished 214, a mixing rating metrics is calculated 215 from the textile-quality data or from the laydown rating metrics of the bale laydowns belonging to the first laydown mixing. The mixing rating metrics is explained below with reference to Figure 4.

[0086] In an “outer” feedback loop 218, the textile-quality data or the mixing rating metrics calculated therefrom are fed back to the computer system 190 and taken into account 216

[0087] - 12 -

[0088] P0640NE-WO when determining a second laydown mixing. In the flowchart of Figure 2, the determination of the second laydown mixing is designated again with the reference numeral 204.

[0089] When a second laydown mixing shall be processed 214, the at least one specification regarding the fiber-quality data and the size information about the bale laydowns for the second laydown mixing are preferably received by the computer system 190. In the flowchart of Figure 2, this is designated again with the reference numerals 202 and 203, respectively.

[0090] If new fiber bales are entered 217 into the inventory and at least some of their fiber-quality data are available, these fiber-quality data are stored in the memory of the computer system 190, which is designated again with the reference numeral 201 in the flowchart of Figure 2.

[0091] Modifications of the example discussed above are possible. For instance, the inner feedback loop 213 can be omitted. In this case, all bale laydowns of a laydown mixing are determined at once, rather than sequentially. The mixing rating metrics is calculated and taken into account when determining the second laydown mixing. The first bale laydown is from a first laydown mixing, and the second bale laydown is from the second laydown mixing.

[0092] In another modification of the example of Figure 2, only individual bale laydowns are considered, whereas the concept of laydown mixings is irrelevant. In this case, the outer feedback loop 218 is omitted.

[0093] The computer system 190 can use a machine-learning algorithm for determining 205 the bale laydowns and / or for determining 204 the laydown mixings. The machine-learning algorithm is preferably based on reinforcement learning. The laydown rating metrics and / or the mixing rating metrics can be used as a reward or an element of a reward. The reward can comprise further elements such as further laydown rating metrics or further mixing rating metrics calculated from fiber-quality data of the bale laydowns. The various elements of the reward can be weighted when calculating the reward. The computer system

[0094] - 13 -

[0095] P0640NE-WO 190 automatically adapts its determination 205 of bale laydowns and / or its determination 204 of laydown mixings by maximizing the reward.

[0096] Figure 3 illustrates an embodiment of the method according to the invention that takes into account one first bale laydown 310 only, which consists of fiber bales 311. The first bale laydown 310 is provided 206, and its fiber material is subsequently processed 208 in a textile production process 100, as described above with reference to Figures 1 and 2. In at least one step 330 of the textile production process, textile-quality data are determined for fiber material from the first bale laydown 310 and received by the computer system 390 via a feedback loop 340. The computer system 390 takes into account the received textilequality data when determining 350 a second bale laydown 320. In the schematic drawing of Figure 3, the bale laydowns 310, 320 are represented as consisting of 2^5 fiber bales 311 and 321, respectively, which shall not limit the generality of the invention.

[0097] The process can be repeated by determining textile-quality data for fiber material from the second bale laydown 320 and receiving them by the computer system 390 via the feedback loop 340, etc.

[0098] Figure 4 illustrates an embodiment of the method according to the invention that takes into account a plurality of first bale laydowns 411 grouped in a first laydown mixing 410. The first bale laydowns 411 are provided and processed one after the other in the textile production process. In at least one step 430 of the textile production process, textile-quality data are determined for fiber material from the first bale laydowns 411 and received by the computer system 490 via a feedback loop 440. The computer system 490 takes into account the received textile-quality data when determining 450 a second laydown mixing 420 with a plurality of second bale laydowns 421. In the schematic drawing of Figure 4, the laydown mixings 410, 420 are represented as each consisting of three bale laydowns 411 and 421, respectively, and the bale laydowns 411, 421 are represented as each consisting of 2x5 fiber bales, which shall not limit the generality of the invention.

[0099] Figure 5 schematically illustrates examples of calculating rating metrics from the textilequality data in embodiments of the method according to the invention; see the steps 210 or 213 of the flowchart of Figure 2. It shows a Cartesian coordinate system 500 spanned by a

[0100] - 14 -

[0101] P0640NE-WO first axis 501 on which a time t is plotted and by a second axis 502 on which a parameter from the textile-quality data, e.g., a reflectance R(t) of the fiber material from the first bale laydowns, is plotted. In the coordinate system 500, fictitious time courses 510, 520, 530 of the parameter R(t) are drawn for three subsequent bale laydowns from a laydown mixing. The parameter values R(t) can be measured continuously or at discrete points in time t.

[0102] For each bale laydown, rating metrics can be calculated by the computer system 190 (see Figure 1) from the textile-quality data, e.g., from the reflectance values R(t). A first example of a rating-metrics parameter is a location parameter such as a mean value of values of a parameter from the textile-quality data, as represented in Figure 5 by a solid line 511, 521, 531 each. A second example of a rating-metrics parameter is a dispersion parameter such as a standard deviation of values of a parameter from the textile-quality data. A parameter range spanning the mean value 511, 521, 531 plus / minus the standard deviation is indicated in Figure 5 by two dotted lines 512, 513; 522, 523; 532, 533 each.

[0103] Rating metrics can also be calculated for the laydown mixing. A first example of a rating metrics for the laydown mixing is a location parameter such as a mean value of the mean values 511, 521, 531 of the individual bale laydowns of the laydown mixing, as represented by a solid line 541 in Figure 5. A second example of a rating-metrics parameter is a dispersion parameter such as a standard deviation of the mean values 511, 521, 531 of the individual bale laydowns of the laydown mixing. A parameter range spanning the mean value 541 plus / minus the standard deviation is indicated in Figure 5 by two dotted lines 542, 543.

[0104] More than one parameter from the textile-quality data can be taken into account in the calculation of the rating metrics.

[0105] One or several parameters from the fiber-quality data can further be further taken into account in the calculation of the rating metrics.

[0106] The computer system 190 (see Figure 1) can store the rating metrics in a memory and assign it to the fiber-quality data of the corresponding individual bale laydowns and / or of the corresponding laydown mixings. The fiber-quality data of the at least one first bale

[0107] - 15 -

[0108] P0640NE-WO laydown and / or the rating metrics assigned thereto can be further taken into account when determining by the computer system 190 the second bale laydown or the second laydown mixing. It is understood that the present invention is not limited to the embodiments discussed above. With knowledge of the invention, the person skilled in the art will be able to derive further variants which are also part of the subject matter of the present invention.

[0109] - 16 -

[0110] P0640NE-WO LIST OF REFERENCE NUMERALS

[0111] 100 Textile production process

[0112] 110 Inventory of fiber bales

[0113] 112, 122 Fiber-testing instruments

[0114] 113, 123 Transmittal of off-line fiber-quality data to computer system

[0115] 115, 125 Determination of bale laydowns

[0116] 120 Bale laydown

[0117] 121 Transmittal of in-line fiber-quality data to computer system

[0118] 130 Fine cleaning

[0119] 131, 141, 151 Transmittal of in-line textile-quality data to computer system

[0120] 132, 142, 152 Off-line laboratory instruments

[0121] 133, 143, 153 Transmittal of off-line textile-quality data to computer system

[0122] 140 Carding

[0123] 150 Rewinding

[0124] 190 Computer system

[0125] 201 Storing of fiber-quality data 202 Receipt of specification regarding fiber-quality data 203 Receipt of size information about first and second bale laydown 204 Determination of laydown mixing 205 Determination of bale laydown 206 Provision of bale laydown

[0126] 207 Update of inventory 208 Processing of fiber material from bale laydown

[0127] 209 Determination of textile-quality data 210 Calculation of laydown rating metrics 211 Decision: shall a further bale laydown be processed? 212 Taking laydown rating metrics into account when determining a further bale laydown

[0128] 213 Inner feedback loop 214 Decision: shall a further laydown mixing be processed? 215 Calculation of mixing rating metrics 216 Taking mixing rating metrics into account when determining a further laydown mixing

[0129] 217 Decision: have new fiber bales been entered into the inventory? 218 Outer feedback loop

[0130] 310, 320 First and second bale laydown 3 H, 321 Fiber bales 330 Step of the textile production process 340 Feedback loop 350 Determination of the second bale laydown 390 Computer system

[0131] 410, 420 First and second laydown mixing 4H, 421 Bale laydowns 430 Step of the textile production process 440 Feedback loop

[0132] - 17 -

[0133] P0640NE-WO 450 Determination of the second laydown mixing

[0134] 490 Computer system

[0135] 500 Cartesian coordinate system 501, 502 First and second axis

[0136] 510, 520, 530 Fictitious time courses of a parameter from the textile-quality data

[0137] 511, 521, 531 Mean value of values of a parameter from the textile-quality data

[0138] 512, 513 Standard deviation of values of a parameter from the textile-quality data 522, 523 Standard deviation of values of a parameter from the textile-quality data

[0139] 532, 533 Standard deviation of values of a parameter from the textile-quality data

[0140] 541 Mean value of mean values of individual bale laydowns of a laydown mixing

[0141] 542, 543 Standard deviation of mean values of individual bale laydowns of a laydown mixing

[0142] P0640NE-WO

Claims

CLAIMS1. A computer-implemented method for determining bale laydowns (120) in a textile production process (100), comprising the steps of: providing (206) at least one first bale laydown; and determining (204, 205), by a computer system (190), at least one second bale laydown (120); characterized by the steps of: receiving, by the computer system (190), textile-quality data determined (209) in the textile production process (100) downstream of the at least one first bale laydown (120), the textile-quality data relating to fiber material from the at least one first bale laydown (120); and taking into account (212) the received textile-quality data when determining, (204, 205) by the computer system (190), the at least one second bale laydown (120).

2. The computer-implemented method according to claim 1, further comprising the step of calculating (210, 213), by the computer system (190), a rating metrics from the received textile-quality data and taking into account (212, 216) the rating metrics when determining (204, 205), by the computer system (190), the at least one second bale laydown (120).

3. The computer-implemented method according to claim 1 or 2, further comprising the steps of: storing (201), in a memory of the computer system (190), fiber-quality data of the at least one first bale laydown (120); assigning, by the computer system (190), the stored fiber-quality data of the at least one first bale laydown (120) to the received textile-quality data of the at least one first bale laydown (120); and further taking into account the fiber-quality data of the at least one first bale laydown when determining (204, 205), by the computer system (190), the at least one second bale laydown (120).- 19 -P0640NE-WO4. The computer-implemented method according to claims 2 and 3, wherein the stored fiber-quality data of the at least one first bale laydown (120) are further taken into account in the calculation (210, 213) of the rating metrics.

5. The computer-implemented method according to any one of the preceding claims, wherein a machine-learning algorithm is used by the computer system (190) for determining (204, 205) the at least one second bale laydown (120).

6. The computer-implemented method according to claim 5 on the one hand and any one of the claims 2 or 4 on the other hand, wherein the machine-learning algorithm is based on reinforcement learning and the rating metrics is used as a reward to be maximized.

7. The computer-implemented method according to any one of the preceding claims, wherein exactly one second bale laydown (120) is determined (205) by the computer system (190), and a location parameter and / or a dispersion parameter of textilequality data determined (209) for fiber material from exactly one first bale laydown (120) is taken into account (212) when determining (205) the second bale laydown (120).

8. The computer-implemented method according to any one of the preceding claims, wherein a second laydown mixing consisting of a plurality of second bale laydowns (120) is determined (204) by the computer system (190) essentially simultaneously, and a location parameter and / or a dispersion parameter of textile-quality data determined (209) for fiber material from a first laydown mixing consisting of a plurality of first bale laydowns (120) is taken into account (216) when determining the second laydown mixing.

9. The computer-implemented method according to any one of the preceding claims, further comprising the steps of storing, in a memory of the computer system (190), fiber-quality data of at least part of textile-fiber bales from an inventory (110) of textile-fiber bales;- 20 -P0640NE-WOreceiving (202), by the computer system (190), at least one specification regarding the fiber-quality data; receiving (203), by the computer system (190), a size information about the first bale laydown (120) and the second bale laydown (120); and further taking into account the stored fiber-quality data, the at least one received specification, and the received size information when determining (204, 205), by the computer system (190), the at least one second bale laydown (120).

10. The computer-implemented method according to claim 9, wherein the stored fiberquality data comprise values of at least one parameter from the following set: reflectance characteristics, color characteristics, fiber-length characteristics, fiberstrength characteristics, fiber tensile properties, fiber-fineness characteristics, fibermaturity characteristics, micronaire, short-fiber characteristics, trash characteristics, dust characteristics, nep characteristics, foreign-matter characteristics.

11. The computer-implemented method according to claim 10, wherein each data field of a data set represents a fiber-quality parameter from the following set: micronaire, yellowness, and reflectance.

12. The computer-implemented method according to any one of the claims 9-11, wherein the at least one received specification regarding the fiber-quality data comprises a value range for the at least one parameter.

13. The computer-implemented method according to any one of the preceding claims, further comprising the steps of: storing, in a memory of the computer system (190), further information for at least part of the textile-fiber bales, the further information being from the following set: bale identifier, plant variety, geographic origin, year of harvest, growth conditions, ginning mill, price, storage location; and further taking into account the stored further information when determining (204, 205), by the computer system (190), the at least one second bale laydown (120).- 21 -P0640NE-WO14. The computer-implemented method according to claim 13, wherein the further information comprises a bale identifier, and the at least one second bale laydown (120) is determined by creating a map of the bale laydown indicating positions of single fiber bales and their bale identifiers.

15. The computer-implemented method according to any one of the preceding claims, wherein the textile-quality data are determined for fiber material having the form of a single fiber, a fiber tuft, a fiber mat, a sliver, a roving, a yarn, or a fabric.

16. The computer-implemented method according to any one of the preceding claims, wherein the determined textile-quality data comprise values of at least one parameter from the following set: reflectance characteristics, color characteristics, fiber-length characteristics, strength characteristics, tensile properties, fiber-fineness characteristics, fiber-maturity characteristics, micronaire, short-fiber characteristics, trash characteristics, dust characteristics, nep characteristics, foreign-matter characteristics, density, mass per unit length, thickness, diameter, roundness characteristics, hairiness characteristics.

17. The computer-implemented method according to claim 1016, wherein each data field of a data set represents a fiber-quality parameter from the following set: micronaire, yellowness, and reflectance.

18. The computer-implemented method according to any one of the preceding claims, wherein the textile-quality data are determined in-line by a production-monitoring instrument (130, 140, 150) or off-line by a laboratory instrument (131, 141, 151).

19. A computer-implemented method for providing at least one bale laydown in an opening department of a spinning mill from an inventory (110) of textile-fiber bales, comprising the steps of: determining (204, 205) at least one second bale laydown (120) according to any one of the preceding claims; and- 22 -P0640NE-WOproviding (206) the at least one second bale laydown (120) from the inventory (110) of textile-fiber bales according to the determined at least one second bale laydown (120).

20. A computer system (190) comprising means for carrying out the method according to any one of the preceding claims.

21. A computer system (190) for determining bale laydowns (120) in a textile production process (100), comprising a processor configured to determine at least one bale laydown, characterized in that the computer system (190) comprises a receiver for receiving textile-quality data determined in the textile production process (100) downstream of at least one first bale laydown (120), the textile-quality data relating to fiber material from the at least one first bale laydown (120), and the processor is configured to determine at least one second bale laydown (120), wherein the received textile-quality data are taken into account in the determination of the at least one second bale laydown (120).

22. A computer program having instructions which when executed by a computer system (190) cause the computer system (190) to perform the method according to any one of the claims 1-19.- 23 -P0640NE-WO

Images