Method for tracking containers formed with moulds by means of a search by alignment of mould identifier sequences

Abstract

The invention relates to a method for tracking containers transported on a conveyor (100) of containers (300), the containers having been formed by means of a set of moulds, the method comprising: - obtaining a first sequence of inspection results associated with information relating to the mould identifiers, - obtaining a second sequence of inspection results associated with information relating to the mould identifiers, - obtaining mould identifiers for the first sequence of results and for the second sequence of results, - searching by sequence alignment for the first sequence of results (SEQ1) in the second sequence of results (SEQ2) using the obtained mould identifiers, to identify containers that are the same.

Description

Description Title of the invention: Method for tracking containers formed with molds using a mold identifier sequence alignment search Technical Field

[0001] The present invention relates to the field of manufacturing containers formed using molds, for example bottles, jars, or flasks, and more specifically the tracking of these containers on a container conveyor track. Prior art

[0002] The manufacture of containers using molds, for example glass containers, involves the implementation of control steps to detect the presence of defects in the containers.

[0003] Typically, after the containers have been formed, they are transported along one or more conveyor lines so that various checks can be carried out. For example, inspection devices equipped with cameras and illumination means can be used for this purpose to detect defects in the containers, or to check dimensions (e.g. wall thicknesses, etc.). Detection devices using wavelengths other than those in the visible range can also be used to detect defects, depending on the type of material of the containers and the defects to be looked for.

[0004] In the field of glass containers, the inspection of molded containers is particularly critical. As is understood, for glass, in addition to the aesthetic aspect, the presence of defects can be dangerous for users of the containers or also for operators who handle these containers. Defects in glass containers can be inclusions (stones or foreign bodies in English), bubbles (blisters or bubbles in English), cracks such as glazes (checks or cracks in English), etc. Aesthetic defects in glass can be creases, brushmarks. Particularly critical defects are defects at the neck of bottles such as external burrs called knockouts or internal burrs called wire-edges, burrs at the mold joint (called wings), trapezoids (birdswing), glued glass, stretched glass, holes. Other defects affect the geometry of objects such as dip and saddle surface flatness defects, ovalizations, blocked necks, or problems related to dimensions (thickness, external diameters of the body, of the finish, internal diameters of the neck, height, inclination of the body or neck (lean), etc.).

[0005] The detection of these defects in glass containers can use different inspection devices such as templates, pneumatic devices, and also devices equipped with cameras (machine vision) which can use transmission lighting by light panels producing uniform diffuse or structured lighting, or polarized light, directional lighting, laser light, etc.

[0006] Typically, several inspection devices are used in a glass container manufacturing facility.

[0007] This is also the case for containers that are formed using molds from other materials.

[0008] As can be understood, when a defect is detected, it may be relevant to act on a parameter or a manufacturing tool used upstream which could have caused the appearance of this defect. For information, it is known that certain defects can result from the presence of a defect on a mold, and that it is then necessary to replace the mold with another. Other defects can be caused by parameters of the manufacturing equipment, typically parameters which affect the manufacture of the containers, from the distribution of the glass drops (called "gob") in the molds to the passage in an annealing arch.

[0009] A glass container manufacturing facility will now be described for purely illustrative purposes, as different process variants exist. First, glass drops or gobs are loaded into a series of forming sections of an IS machine (Individual Sections) each comprising 1 to 4 blank molds and 1 to 4 finishing molds. The blanks from the blank molds are transferred to the finishing molds for blowing and for the blanks to take the shape of the containers. Within the facility, the sections are generally identified, but also the molds. Typically, for a section identified by the number 1 and capable of accommodating four molds, the four molding cavities, each accommodating one mold in the section, can be designated IA, IB, IC, and 1D (these are the forming cavity identifiers).Within the same manufacturing facility another section identified by the number n can accommodate four other molds, and the molding cavities can be designated nA, nB, nC, and nD.

[0010] Finishing molds (they are the ones that transfer their number by molding to the containers), taken as such, also have their own identifier. Typically, for a glass container manufacturing facility, a set of several dozen finishing molds may be used, some of which occupy all the molding cavities of the facility and another part may be either in repair or in reserve for a later mold replacement in a forming cavity. Molds are identified with mold identifiers, typically identifiers inscribed by means of a coding that is marked in the form of a relief in the molds, and which leave an imprint on all the containers formed by means of the molds, this imprint can be decoded to obtain the mold identifier. In general, this imprint is located at the base, heel (or jable) or bottom of the containers.

[0011] In this description, a "mold identifier" is a code, for example numeric or alphanumeric, which makes it possible to identify each mold uniquely, and for example to find information about the mold. For information purposes, this is a mold number.

[0012] It appears that by reading / decoding a mold identifier on a bottle with a defect resulting from a mold defect, a corrective action can then be triggered in the process by modifying a machine parameter for a given cavity (typically a parameter controlling the start or end times of operations (timing), ventilation or blowing instructions, etc.), or even a replacement by an operator of the mold with a defect. In glass container manufacturing facilities, tables are generally used in which the identifiers of the molds used at each time in the molding cavities are noted.

[0013] In the field of glass containers, it is known that a time ranging from forty-five minutes to two hours can elapse between the moment the containers pass through an annealing arch and the moment the containers are inspected by devices. If a defect is detected by a device, reading the mold number allows action to be taken on the process, on the mold or on the adjustment of the section or cavity but only after the expiry of this time (and, as explained below, with uncertainty on the mold used). The use of said table of mold identifiers makes it possible to know the position of the mold in its forming cavity in order to replace it. The table can also contain information on the sections, and for the molds, indicate which finishing molds are in use at each time.

[0014] Containers have mold identifier markings, but also possibly other markings that may indicate, for example, a forming cavity identifier or even a section number. These numbers can also be obtained by reading by inspection devices.

[0015] From the prior art, the following documents are known which describe in particular the reading of a mold identifier (by decoding) of inspected containers: - US 3923158, in which mold identifiers are read on inspected containers and the mold identifiers are associated with detected defects; - EP3206805, in which containers are illuminated in a particular manner to enable the decoding of a code formed in relief which is visible on an image to be processed, typically to obtain a mold number; - US 4644151, in which the container base has reliefs encoding a mold identifier, the reliefs being arranged in pairs with two possible positions for encoding bits; - US 2010 / 0080442, in which containers having reliefs are also illuminated in a particular manner to obtain images to be processed and decoded mold identifiers. [0016JII It is known to implement statistical analyses to determine, on the basis of defect detections and using mold identifiers, whether it is necessary to replace a mold that frequently produces defective containers or to correct an operating parameter of a defective section or cavity identified by the mold with which it is equipped. Also, by using the table mentioned above which associates, at each moment, the molding cavities / sections with the molds used (finishers and also roughers), it is possible to determine whether an action by an operator is required for the molding sections / cavities. These analyses are nevertheless limited, since a duration for example of between forty-five minutes and two hours can elapse between the moment when the containers pass through an annealing arch and the moment when the containers are inspected by devices, for glass containers.Furthermore, upon exiting an annealing arch, the glass containers are mixed, so that the order in which the containers are inspected is not the order in which the containers exited the forming sections using molds. Therefore, it is not possible to know precisely the time of manufacture of a container based on a reading of a mold identifier read by an inspection device.

[0017] Thus, in the event of a mold replacement, or when an operating parameter is changed, it is not possible in a period following these changes to know whether a defective container was manufactured before or after these changes. This can prevent or delay the intervention of an operator, which only has information on the quality of the containers produced for time periods of several minutes.

[0018] It can be noted that the mixing of containers occurs due to the configuration of the annealing arches and the use of several conveyor lines before transporting the containers on a single conveyor line (among the several which can operate in parallel) on which the inspection devices are arranged.

[0019] Containers made of materials other than glass, typically plastic, may also be mixed before arrival on a conveyor line for inspection. [0020)11 There are also markings engraved (for example by laser) on the containers. These markings contain more information than a mold identifier marking as presented above, and the use of certain marking methods such as laser makes it possible to produce markings specific to each container (the marking is done by scanning). As can be seen, in the methods of the prior art, when a mold is used to mark a mold identifier, reading this identifier only provides information limited to this mold identifier and no information specific to the container: all the containers formed using the same mold are in theory identical and with the same mold identifier inscribed.

[0021] The use of engraved marking allows the inscription of a manufacturing moment directly on the containers, or the recording on a remote computer medium of this manufacturing moment which is found by means of the unique identifier of the containers marked by laser.

[0022] From the prior art, document EP 2368861 is known which describes the time stamping (date, hour, minute, second) of containers using a laser beam to engrave transparent or translucent objects. Unique information for each object is engraved, and it is proposed to use, for example, a Datamatrix coding expressing the time stamp combined with the cavity number. This solution makes it possible to associate the inspection results of a container with its manufacturing time and therefore with the parameters of its manufacturing (identifier of forming cavity for example). It nevertheless requires the use of an inspection machine equipped with a code reader and more precisely a Datamatrix code reader.

[0023] As explained in document EP2368861, the laser marking is installed at the output of the manufacturing machine, to mark, as soon as possible after manufacturing, the containers leaving the sections in a constant and known order (i.e. before mixing the containers). The marker being synchronized with this output order, it is thus possible to mark in a code both the cavity number of each container and the manufacturing time, or to mark an arbitrary but unique serial number which is associated with the manufacturing data.

[0024] EP 2114834 also describes the use of unique codes, which are printed on each container as it leaves the mold, but which are in the form of serial numbers, associated at the time of marking, in a given database, with process data. It is also described that the unique serial numbers are readable at the time of cold inspection.

[0025] In a manner known per se, several inspection devices are used along a conveyor track partly dedicated to inspection. These devices are spaced apart, and reading a unique code (of the Datamatrix type) at one location does not, in prior art installations, make it possible to overcome the problems mentioned above regarding readings of mold identifiers alone. Indeed, it is not easy to precisely track the same container with only the distance traveled on the conveyor track as information, in particular because of possible changes: breakage, replacement of containers, addition of containers, etc.

[0026] In an inspection device, a unit called a supervision unit can carry out tracking of containers, to know the presence and position of each container within an inspection device, using for example optical barriers, displacement encoders, etc. This tracking makes it possible to locally identify containers and to associate them their inspection results, at least while they pass through the inspection device, for example in view of possible ejection of a container.

[0027] Some inspection devices described in US 2009 / 0158865 and EP 2344404 use a star-shaped support to hold containers and transport them to successive inspection stations. These devices are called carousels. Containers held in slots can be easily tracked by means of electronic cams that can be used to track containers and properly associate containers with inspection results (with possible ejections).

[0028] For devices equipped with belts that move containers while holding them laterally, successive inspection stations can also be used. In this context, it is appropriate to use one or more cameras to track the containers within the inspection machine.

[0029] Finally, we know of inspection devices which transport containers on a conveyor, to make them pass through different inspection stations, and which use an encoder to determine the distance traveled on the conveyor by the containers in order to carry out their monitoring.

[0030] These tracking solutions are limited to the inspection devices that implement them.

[0031] As explained above, not all inspection devices have a marking reader including, for example, a unique identifier (of the Datamatrix type). For reasons of space, it is not conceivable to add readers to inspection devices that are not equipped with them and that are already installed within a container manufacturing facility.

[0032] There is therefore a need for a solution to track containers on a conveyor line equipped with different inspection devices.

[0033] In particular, there is a need to overcome the problems associated with the use of mold ID readers alone, which do not allow precise knowledge of the manufacturing time and parameters used for a container, and also do not allow the inspection results of the same container checked in successive spaced devices to be linked. Statement of the invention

[0034] The present invention aims to remedy all or part of the drawbacks of the prior art, by proposing a solution for tracking containers on a conveyor line equipped with inspection devices.

[0035] For this purpose, a method is provided for tracking containers transported on a container conveyor track, the containers having been formed by means of a set of molds (i.e. a finished set of molds) installed in forming cavities (the number of forming cavities may be less than the number of molds in the set of molds; several forming cavities are used), each mold in the set of molds being associated with a mold identifier and / or each forming cavity being associated with a forming cavity identifier, and each container comprising one or more markings of information relating to a mold identifier and / or a forming cavity identifier used for the container, the method comprising: - obtaining a first sequence of inspection results for a first set of containers transported consecutively on the conveyor track (for each container of the first set of containers, there is therefore an inspection result specific to the container which is in the first sequence of inspection results), using a first container inspection apparatus arranged at a first location on the conveyor track, each inspection result of the first sequence of results being associated with information relating to the container mold or forming cavity identifier read by the first apparatus on a marking of the container among said one or more markings, - obtaining a second sequence of inspection results for a second set of containers transported consecutively on the conveyor track (for each container of the second set of containers, we have therefore an inspection result specific to the container which is in the second sequence of inspection results), the second set comprising more containers than the first set (therefore, the second sequence is longer than the first sequence), using a second container inspection apparatus arranged at a second location of the conveyor track, each inspection result of the second sequence of results being associated with information relating to the mold or forming cavity identifier of the container read by the second apparatus on a marking of the container among said one or more markings, - obtaining, for the first sequence of results and for the second sequence of results, mold identifiers from the information relating to the read mold identifiers or forming cavity identifiers from the information relating to the read forming cavity identifiers (it should be noted that information relating to a mold identifier can also be information relating to a forming cavity identifier, insofar as there is a table linking the mold identifiers to the identifiers of forming cavities receiving these molds, conversely, information relating to a forming cavity identifier can also be information relating to a mold identifier, insofar as the same table exists; it can also be noted that from this step, either mold identifiers are obtained for all the containers of the sequences or forming cavity identifiers for all the containers of the sequences), - a sequence alignment search of the first sequence of results in the second sequence of results using the obtained mold identifiers or forming cavity identifiers (in fact, this amounts to aligning sequences of mold numbers or sequences of forming cavity identifiers, and for each mold or cavity identifier, there is an inspection result), to identify the same containers in the first set and in the second set of containers.

[0036] The molds in the mold set here are finishing molds.

[0037] It may be noted that if the method is implemented on the basis of mold identifiers, they are obtained in the obtaining step and the search step will be done by a sequence alignment based only on the mold identifiers. If the method is implemented on the basis of forming cavity identifiers, they are obtained in the obtaining step and the search step will be done by a sequence alignment based only on the forming cavity identifiers.

[0038] This process can be implemented by a computer system (one or more computers).

[0039] It has been observed by the inventors of the present invention that within a given conveyor line, it happens that containers are broken, replaced, inserted, but that this does not prevent the implementation of a sequence search based on information that can be acquired by two different inspection devices. This is notably made possible by the random nature of the order of the containers arriving on the conveyor line, this random nature resulting notably from the formation by means of a set of molds of the containers (typically because the containers are mixed after forming, for example at the exit of an annealing arch in the field of glass containers). This is furthermore made possible by the absence of mixing of the containers as soon as they are transported on the conveyor line.

[0040] It should be noted that by conveyance track, we mean a track which does not have forks between two conveyance tracks, or a track which only has a single branch.

[0041] If containers are determined to be the same containers, this makes it possible to associate with a container for which only a mold identifier is known, or for which only a forming cavity identifier is known, and an inspection result from one inspection device, all other information from the other inspection device. Knowledge of the presence of the same containers in both inspection devices (of course, at different times) is considered as tracking in the present invention (i.e. knowledge of the position at different times).

[0042] By combining the inspection results, problems in the prior art with inspection apparatus that read only mold identifiers or forming cavity identifiers (and, for example, no other information uniquely identifying each container) are overcome.

[0043] Information relating to a mold identifier can be a mold identifier, or information that allows a mold identifier to be retrieved. Also, information relating to a forming cavity identifier can be a forming cavity identifier, or information that allows a forming cavity identifier to be retrieved. Thus, if a unique identifier is marked, this unique identifier can be associated in a remote database with a mold identifier or a forming cavity (from which mold numbers can be obtained, and vice versa). The method described above therefore works with two devices that may not both be equipped with complex code readers in which unique identifiers, etc., are coded.

[0044] The person skilled in the art will be able to choose the lengths of the sequences according to the number of molds in the set of molds or the number of forming cavities which accommodate the molds so that the search by sequence alignment can be implemented.

[0045] It should be noted that the first inspection device and the second inspection device can be arranged in any possible order. Thus, the first inspection device can be upstream or downstream of the second inspection device.

[0046] Also, by, identifying the same containers, it is meant that there is a good chance that they are the same containers. For example, this may be the case if the sequence alignment search shows a zero distance between the first sequence and a portion of the second sequence (all mold identifiers are identical). The distance between two sequences may be a representation of their similarity based in part on order, the distance between two identical sequences being zero. Also, the person skilled in the art will know determine a distance threshold between two sequences to determine that they are the same containers.

[0047] For example, the sequence alignment search may involve scanning the second sequence with the first sequence, determining a distance for each position. The minimum distance may thus indicate the position of the first sequence in the second sequence. Furthermore, if the minimum distance is greater than or equal to a given distance threshold, it may be assumed that the first sequence was not found or aligned in the second sequence. If the minimum distance is less than this threshold, the position corresponding to this minimum distance is that of the first sequence in the second sequence. When it is determined that the mold or forming cavity identifiers are identical between the first sequence and this position in the second sequence, then this may imply that the containers are the same containers.

[0048] The invention, which implements a simple sequence alignment search, is well suited for manufacturing facilities where hundreds of containers are produced per minute. For example, for glass containers, between 150 and 700 containers can be produced per minute. Since the number of molds in use is typically between 10 and 40 (the number of molds in use corresponds to the number of forming cavities), there are between 10 and 40 different mold numbers mixed together.

[0049] Also, the invention uses information that is the same between different containers (mold or forming cavity identifier), but which at first glance does not allow a mold to be identified. Due to the random nature of the order of the containers, searches by sequence alignment allow the same containers to be identified on the basis of simple information, which alone does not identify containers (which allows the use of simple readers not capable of reading unique identifiers that could be written on the containers).

[0050] According to a particular implementation mode, for each same container, the first inspection result of the container is associated with the second inspection result of the container.

[0051] Associate means associating a storage of inspection results together in a single memory or data structure. Associate also means that pointers can be used to link inspection results. Also, a table can be used that has information from one of the inspection results as input and outputs the second inspection result. Other association methods can be used.

[0052] According to a particular implementation mode, the second location is upstream of the first location along the conveyor track.

[0053] This particular mode of implementation is advantageous since the second sequence is longer than the first, and also because the inspections implemented after passing to the second location can lead to ejections of containers: the first sequence is thus sought in a second, more complete sequence.

[0054] According to a particular embodiment, the containers comprise at least one mold identifier marking and the information relating to the mold identifiers read on the mold identifier marking by a device among the first inspection device and the second inspection device are the mold identifiers.

[0055] In this particular implementation, one of the two inspection devices is equipped with a mold identifier reader, typically a mold identifier reader marked by a raised pattern encoding the mold identifier. This reading is direct, and does not use correspondence tables to obtain a mold identifier from other information, except possibly to perform decoding.

[0056] According to a particular implementation mode, the information relating to the mold identifiers read on the mold identifier marking by the first inspection device and by the second inspection device are the mold identifiers.

[0057] In this particular implementation, there are two inspection devices, both equipped with mold identifier readers. In other words, these devices cannot read a unique container identifier, for example. It is understood that the implementation of the method is very advantageous in this case since the search by sequence alignment makes it possible to identify the same containers without using a unique container identifier. This implementation mode is well suited for existing installations that cannot accommodate more complex code readers.

[0058] According to a particular embodiment, the containers comprise at least one marking of a forming cavity identifier and the information relating to the forming cavities read on the marking of the forming cavity identifier by a device among the first inspection device and the second inspection device are the forming cavity identifiers.

[0059] In this particular embodiment, there is a forming cavity identifier marking on the containers. For example, these markings can be inscribed after the forming step, for example by applying a laser beam. This embodiment is suitable in a context where it is not desired for the molds to leave an embossed mold identifier by imprint, which is the case for bottles intended to receive perfume, where a discreet marking can be noted. The forming cavity identifier can be coded according to a chosen coding.

[0060] According to a particular embodiment, the information relating to the forming cavities read on the forming cavity marking by the first inspection device and by the second inspection device are the forming cavity identifiers.

[0061] In this particular implementation, there are two inspection devices, both equipped with forming cavity identifier readers. In other words, these devices cannot read a unique container identifier, for example. It is understood that the implementation of the method is very advantageous. in this case since the search by sequence alignment (here based on the forming cavity identifiers) makes it possible to identify the same containers without using a unique container identifier. This implementation mode is well suited for existing installations which cannot accommodate more complex code readers.

[0062] According to a particular embodiment, the containers comprise the mold identifier marking, and another marking comprising information relating to the mold identifier and / or the forming cavity identifier, in which the information relating to the mold identifiers read on the other marking using the other apparatus among the first inspection apparatus and the second inspection apparatus are forming cavity identifiers or unique identifiers of each container, the method comprising obtaining the mold identifiers from the forming cavity identifiers or the unique identifiers by means of a look-up table.

[0063] For example, the other marking contains more information than the mold ID marking (typically, it is encoded on a higher number of bits).

[0064] In this particular implementation, the sequence alignment search is based on the forming cavity identifiers.

[0065] Also, in this particular embodiment, the containers have two distinct markings. The other device among the first inspection device and the second inspection device is therefore the device which is equipped with a reader capable of reading this other code (at least equipped with an appropriate sensor), while the device which is not the other is the one which is equipped with a mold identifier reader. One can designate both by designating them for example by the expressions mold identifier reader device and other marking reader device, it being understood that this designation can apply as much to the first inspection device as to the second inspection device. The first inspection device can be the mold identifier reader device and the second inspection device can be the device for reading other markings, and vice versa.

[0066] This implementation method is advantageous in that it allows the determination of the same containers inspected by devices, one of which can only read mold identifiers and no unique container identifier, which the other marking can understand. It avoids the space problems associated with adding a marking reader that is more complex than a mold identifier reader.

[0067] According to a particular mode of implementation, the other markings or the forming cavity identifier markings (which are distinguished here from the other markings) are applied to each container after the container has been formed (this forming being that carried out by means of a mold of the mold assembly).

[0068] Thus, the other markings or forming cavity identifier markings are not markings resulting from impressions of the molds of the mold assembly since they are applied after the containers have been formed. For example, the other markings may be formed by applying a laser beam, by printing, by labeling, etc.

[0069] According to a particular embodiment, the other markings are codes in which the forming cavity identifiers and / or the unique container identifiers and / or the mold identifiers are coded.

[0070] In one embodiment, to obtain that the other marking is unique (for each container), this other marking can be applied to the containers when all the containers pass through the same place. Preferably, this other marking is applied when all the containers pass through the same place in an established order and as close as possible to their manufacture, so that the order allows the serial number to be linked with production data such as the timestamp and / or the forming cavity, or so that the marked identifier contains unique information combining the timestamp and / or the forming cavity. The marking of the containers is therefore preferably at the exit of the IS machine and before their mixing at the exit of the arch.

[0071] In fact, in the present application, an “other marking” can be considered as a unique marking for each container, which makes it possible to trace (if they are not directly written therein) information such as the mold identifier, the forming cavity identifier, or even time stamps (unique for each container) relating to a specific step.

[0072] Typically, the other markings may be one-dimensional (barcode type) or two-dimensional (Datamatrix or QR Code type) codes, configured to encode at least any combination of one or more information elements among forming cavity identifiers (the one that received the mold in which the container on which this number is encoded in the other marking), unique container identifiers (each identifier is specific to the container on which it is encoded in the other marking), and mold identifiers (the one used to form the container on which this identifier is encoded in the other marking).

[0073] According to a particular embodiment, the other device among the first inspection device and the second inspection device is a code reader (according to the coding used for the other marking), the inspection results obtained by this other device being code readings.

[0074] Here, the inspection results themselves are code readings, either a decoding or data acquired by a sensor to be decoded later. In other words, the other device is not strictly speaking a fault detection device and its inspection may be limited to reading codes marked in the other markings.

[0075] According to a particular implementation mode, a timestamp of a manufacturing step of each container is coded in the code.

[0076] The step of manufacturing a container may be a forming step using a mold, a step of marking the other marking (considered here as a manufacturing step), an annealing step, a chisel cutting step to obtain the parison which will be transformed into a container, the step of loading this parison into the cavity used for the container, the step of extracting the molded container from its mold (finisher), etc.

[0077] Thus, the method allows this timestamp to be used to assign it to an inspection result that is not associated with a timestamp. This overcomes the problems associated with the use of inspection devices that can only read mold identifiers and, above all, the problems of imprecision on the forming times with a mold.

[0078] According to a particular embodiment, the method comprises determining a manufacturing time of each container from a time of inspection of the container by the first or by the second inspection device.

[0079] In this particular implementation mode, a timestamp is not obtained by reading but is determined.

[0080] According to a particular embodiment, the method comprises a step of associating, for each same container, the unique identifier and / or the timestamp when the latter is available, with the inspection result of said device among the first inspection device and the second inspection device, and / or, the method comprises a step of associating, for each same container, the determined manufacturing time with the first inspection result and with the second inspection result, if this determined manufacturing time is available.

[0081] According to a particular mode of implementation, the method comprises a step of associating, for each same container, manufacturing data with the first and second inspection results.

[0082] Manufacturing data can be mold ID, forming cavity ID, gob weight, glass temperature, mold temperature, IS machine parameters, various measurements, hot inspection results (before arch), cold inspection results (after arch).

[0083] According to a particular mode of implementation, the search by sequence alignment is implemented by means of an algorithm calculating a distance (or metric) measuring the differences between the first sequence and subsequences of the second sequence all associated with different positions alignment, and seeking an alignment position that minimizes the distance. Thus, we search among the possible positions for the position of the first sequence in the second sequence for which the distance is minimal.

[0084] For example, for distance, we can use the Levenshtein distance, calculated via the Levenshtein algorithm.

[0085] Different sequence alignment search algorithms can be implemented. Sequence alignment search algorithms use algorithms for calculating the distance between two sequences, this distance being minimal when an alignment is found. It has been observed by the inventors that in the context of manufacturing glass containers, it is advantageous to use the Levenshtein algorithm given the possible changes that occur within a glass container conveyor line (breakage, ejection, insertion by an operator, etc.). This algorithm makes it possible to obtain good alignments, and a fortiori good identification of the same containers.Other algorithms can be used to implement the invention, such as algorithms that are more complex to implement and more expensive than the Levenshtein algorithm, such as those of the Damerau-Levenshtein or Wagner-Fischer type, which are advantageous in that they can process more disturbances.

[0086] As a guide, the Levenshtein Distance provides a distance between strings of characters, its application in the process may include consideration of mold or cavity identifier sequences as strings of characters. The Levenshtein distance, which takes into account substitutions, insertions (or addition) and deletions (or erasure) between two strings or sequences, is well suited to the possible changes that occur within a glass container conveyor line (breakage, ejection, insertion by an operator, etc.).

[0087] According to a particular method of implementation, prior to transport on the conveyor line, the containers are mixed.

[0088] The process therefore involves mixing the containers, so that the containers are in random order within the conveyor line, identical sequences that can hardly be reproduced in a given period of time, which, in a context where the number of molds in the set of molds is limited, makes it possible to clearly identify sequences and therefore determine the same containers.

[0089] According to a particular implementation method, the containers are mixed at the outlet of an annealing arch opening onto the conveyor track.

[0090] According to a particular implementation mode, the method is implemented by the first or by the second inspection device.

[0091] For example, the first or second inspection device may have a computer system structure and be able to implement the method. In this case, the two inspection devices are in direct or indirect communication (for example, if they are in communication with the same server).

[0092] According to a particular mode of implementation, the method is implemented by a computer system separate from the first and second inspection devices.

[0093] This computer system is in direct or indirect communication with the two inspection devices to obtain at least the information relating to the mold identifiers and the inspection results.

[0094] The invention also provides a system for tracking containers transported on a conveyor track of containers formed by means of a set of molds installed in forming cavities, each mold of the set of molds being associated with a mold identifier and / or each forming cavity being associated with a forming cavity identifier, and each container comprising one or more markings of information relating to a mold identifier and / or a forming cavity identifier used for the container, the system comprising: - a module for obtaining a first sequence of inspection results for a first set of containers transported consecutively on the conveyor track, using a first container inspection apparatus arranged in a first location of the conveyor track, each inspection result of the first sequence of results being associated with information relating to the mold or forming cavity identifier of the container read by the first device on a marking of the container among said one or more markings, - a module for obtaining a second sequence of inspection results for a second set of containers transported consecutively on the conveyor track, the second set comprising more containers than the first set, using a second container inspection apparatus arranged at a second location on the conveyor track, each inspection result of the second sequence of results being associated with information relating to the container mold or forming cavity identifier read by the second apparatus on a marking of the container among said one or more markings, - a module for obtaining, for the first sequence of results and for the second sequence of results, mold identifiers from the information relating to the mold identifiers read or forming cavity identifiers from the information relating to the forming cavity identifiers read, - a search module by sequence alignment of the first sequence of results in the second sequence of results using the obtained mold identifiers or the obtained forming cavity identifiers, to identify the same containers in the first set and in the second set of containers.

[0095] This system can be configured to implement any of the methods of implementing the method defined above.

[0096] The invention also provides a computer program comprising instructions for executing the steps of a method as defined above when said program is executed by a computer.

[0097] Note that the computer programs mentioned in this presentation may use any programming language, and be in the form of source code, object code, or intermediate code between source code and object code. object, such as in a partially compiled form, or in any other desirable form.

[0098] The invention also provides a computer-readable recording medium on which is recorded a computer program comprising instructions for executing the steps of a method as defined above.

[0099] The recording (or information) media referred to in this disclosure may be any entity or device capable of storing the program. For example, the medium may comprise a storage medium, such as a ROM, for example a CD ROM or a microelectronic circuit ROM (EPROM EEPROM), or a magnetic recording medium, for example a floppy disk or a hard disk.

[0100] On the other hand, the recording media may correspond to a transmissible medium such as an electrical or optical signal, which can be conveyed via an electrical or optical cable, by radio or by other means. The program according to the invention can in particular be downloaded from a network such as the Internet.

[0101] Alternatively, the recording media may correspond to an integrated circuit in which the program is incorporated, the circuit being adapted to carry out or to be used in carrying out the method in question. Brief description of the drawings

[0102] [Fig. 1] Figure 1 schematically shows the steps of a process according to an example.

[0103] [Fig. 2] Figure 2 is a schematic representation of a container conveyor track and shows the sequence alignment.

[0104] [Fig. 3] Figure 3 is a table showing the sequence alignment search.

[0105] [Fig. 4] Figure 4 shows an embodiment in which a reader of other marking is used as an inspection device.

[0106] [Fig. 5] Figure 5 shows another embodiment in which a reader of other marking is used as an inspection apparatus.

[0107] [Fig. 6] Figure 6 schematically represents the association of inspection results.

[0108] [Fig. 7] Figure 7 schematically shows an example in which a remote computer system implements the method.

[0109] [Fig. 8] Figure 8 shows yet another embodiment in which forming cavity identifiers are marked. Description of the embodiments [OHO]We will now describe a method and a system for tracking containers transported by a conveyor track equipped with inspection devices. This method and this system find application in the tracking of containers formed by means of molds installed in forming cavities, such as for example glass containers obtained by blow molding.

[0111] The method and system can be applied whenever the containers are formed by means of a set of molds used in forming cavities and whenever the containers include a marking of information relating to a mold or forming cavity identifier. In fact, in the mold set, each mold has an identifier and during the formation or afterward, each container is marked with this information relating to a mold identifier used for the formation of the container.

[0112] This description, which is in no way limiting, first presents examples in which mold identifiers are used (with reference to figures 1 to 7). The use of cavity identifiers will be described with reference to figure 8.

[0113] In a context of manufacturing containers using a set of molds, where several molds can be used in parallel, and where annealing can be carried out through an annealing arch, the containers generally open onto at least one conveyor track in a random order, that is to say without any link to the order in which the containers are formed, so that it is not possible to predict after forming in what order the containers will be on the conveyor track. The conveyor tracks referred to here have only one branch and no forks, to prevent mixing of the containers. That being said, disturbances are possible, as is the case conventionally on a conveyor track dedicated to the inspection of containers (breakage of a container, removal, manual addition or replacement of a container, ejection, etc.). The invention can nevertheless be implemented for several conveyor tracks of the same installation, for example tracks in parallel.

[0114] Figure 1 is a flowchart of an exemplary container tracking method. This method may be implemented by a computer system, typically a computer, and possibly by an inspection device used to inspect the containers.

[0115] In fact, to implement the method, two inspection devices are used, capable of obtaining inspection data and said information relating to the mold identifiers marked on each container.

[0116] Examples of arrangement of these inspection devices will be described with reference to Figures 3 to 5.

[0117] The process of Figure 1 first involves obtaining inspection results.

[0118] More specifically, during a step OBT_1, an inspection apparatus called a first inspection apparatus obtains a first sequence of inspection results for a first set of containers transported consecutively on the conveyor track, using the first container inspection apparatus which is arranged at a first location of the conveyor track, each inspection result of the first sequence of results being associated with information relating to the mold identifier of the container read by the first apparatus on a marking of the container among said one or more markings. For example, one (or all) inspection results of the first sequence of results contains information relating to the mold identifier of the container read by the first inspection device (for information purposes, this is what is meant by the association mentioned above).

[0119] In a step OBT_2, a step analogous to step OBT_1 is implemented but using a second inspection apparatus and with a larger set of containers. In step OBT_2, a second sequence of inspection results is obtained for a second set of containers transported consecutively on the conveyor track, the second set comprising more containers than the first set, using the second container inspection apparatus which is arranged at a second location on the conveyor track, each inspection result of the second sequence of results being associated with information relating to the mold identifier of the container read by the second apparatus on a marking of the container among said one or more markings.Similarly, one (or all) inspection results of the second sequence of results contain information relating to the container mold identifier read by the second inspection device (for information purposes, this is what is meant by the association mentioned above).

[0120] Steps OBT_1 and OBT_2 are implemented in an order that depends on the position of the first location and the position of the second location within the conveyor lane. For example, if the first location is upstream in the DIR direction of the flow of movement of the containers of the second location, step OBT_1 will be implemented before step OBT_2, and conversely if the first location is downstream of the second location, step OBT_1 will be implemented after step OBT_2.

[0121] Also, the first set is preferably at least partially included in the second set, to facilitate the search by sequence alignment (if it is not, the search fails and the same containers are not found). Depending on the size of the second set, and depending on the size of the first set, we can fix the start times of steps OBT_1 and OBT_2. It can be noted that taking into account the distance traveled by the route of conveying, it can be determined when a container inspected in a first device will be inspected by a second device.

[0122] After implementing steps OBT_1 and OBT_2, there are two sequences of inspection results with, for each inspection result, information relating to an associated mold identifier. It may be noted that in the present application, by sequence is meant an ordered list of elements, this order corresponding here to that in which the containers of each set are transported consecutively, this order being maintained.

[0123] We can then implement the OBT_IM step which involves obtaining, for the first sequence of results and for the second sequence of results, the mold identifiers from the information relating to the mold identifiers read.

[0124] In fact, the inspection devices used for mold-formed containers covered by the present application may be equipped with readers of different types, the containers comprising markings readable by these readers.

[0125] A first type of reader is a mold identifier reader. For example, mold identifiers may be marked on the containers by reliefs (according to a given coding), typically arranged at the base of the containers. In this case, step OBT_IM for the result sequences is implemented jointly with step OBT_i if the inspection device i can read these mold identifiers.

[0126] Another type of reader is a coded marking reader, typically coded in the form of a two-dimensional code (Datamatrix, QR code, etc.) or one-dimensional code (barcode, etc.). These markings are called “other marking” in this description, and they may possibly contain more information than the mold identifier markings coded by reliefs. In particular, they may contain mold identifiers (in this case the step OBT_IM for the result sequences is implemented jointly with the step OBT_i for the device that read this marking), and / or forming cavity identifiers, and / or unique container identifiers and also a time stamp.

[0127] If the other marking contains in its code a forming cavity identifier, then the method may use a table indicating the identifier of the mold used within that forming cavity (possibly using either an estimated elapsed time since forming or a timestamp entered in the code), and the OBT_IM step then includes using the table to obtain the mold identifier.

[0128] If the other marking contains in its code a unique container identifier, then the method can, during step OBT_IM, obtain the mold identifier by a table linking the unique container identifiers to the mold identifiers used or to the cavity numbers.

[0129] After implementing the OBT_IM step, we have two sequences of inspection results and each inspection result is associated with a mold identifier.

[0130] It is then possible to implement, during the REC_AL step, a search by sequence alignment of the first sequence of results in the second sequence of results using the mold identifiers obtained, to identify the same containers in the first set and in the second set of containers.

[0131] This step can use any sequence alignment algorithm. However, at least in the context of glass container manufacturing, it is appropriate to use the Levenshtein algorithm, the implementation of which will be detailed with reference to Figure 2.

[0132] At this stage, if necessary, the inspection results for at least one common container from both sequences, or even all the same containers from the first sequence from the first device and the second device, can be combined. For example, this can be implemented if the inspection results from one of the two devices indicate that the container has a defect.

[0133] Alternatively, this can be implemented for all the same containers, i.e. those whose inspection results are in both sequences according to the result of the sequence alignment search. The ASS step shown further below with reference to Figure 6 is optional and aims at associating the results for the same containers.

[0134] The steps described with reference to Figure 1 can be implemented once for two sets of selected containers. They can also be implemented repeatedly for first sets that follow one after the other. The method can be implemented for several consecutive first sets and the same second set, or consecutive second sets. For example, the procedure is implemented for each container inspected by the first device and / or for each container inspected by the second device, or per batch each time a given number of containers is inspected by the first device and / or by the second device.

[0135] Figure 2 shows both a conveyor track and the implementation of the method as described with reference to Figure 1. More specifically, this figure shows an INS installation comprising a conveyor track 100 on which circulate containers not shown in the figure for the sake of simplicity. In fact, the containers circulate from left to right in the figure in the direction DIR. As can be noted, the conveyor track 100 does not have a fork and only has one branch.

[0136] The containers arriving on the conveyor line are mixed. By mixed is meant mixed as containers are at the outlet of an annealing arch. The method may also include the mixing step, for example at the outlet of an annealing arch.

[0137] Two inspection devices 110 and 120 are arranged respectively at locations E1 and E2 of the conveyor track. Location E1 is downstream of location E2, and E1 is called a first location and E2 a second location. In fact, the naming convention first and second also applies to the inspection devices in the present application, first is used to designate the inspection device which obtains a sequence of shorter length than the other inspection device (the second). Inspection device 110 is therefore the first inspection device and inspection device 120 is therefore the second inspection device, in view of their obtained sequences described below.

[0138] In the illustrated example, the first inspection device 110 is equipped with a reader 115. This reader 115 is configured to read the markings referred to above as “other markings”, i.e. markings in which more information is coded than for mold identifier markings marked by a relief, or an imprint left by the mold used. In other words, the information of the markings read by the reader 115 can be coded on a higher number of bits than the information read by a reader which reads markings formed by reliefs on the containers.

[0139] It may be noted that in the present description, the first inspection device may be referred to as another device in that it reads other markings.

[0140] For example, the reader 115 may be a two-dimensional code reader of the Datamatrix type formed by applying a laser beam to each container. This reader is preferably an optical and contactless code reader. These two-dimensional codes may include in their code forming cavity identifiers or even unique container identifiers or even mold identifiers.

[0141] For information, Datamatrix type codes may be in accordance with the “CETIE - International Technical Center for Bottling and related Packaging” standard with reference DT40.00, for example in its September 2020 version.

[0142] For a first set of containers comprising a chosen number of containers (the number is chosen) which pass through the first inspection device 110, a first sequence of inspection results and information read by the reader 115 is obtained by implementing step OBT_1. The first inspection device 110 therefore comprises at least one reader 115, and possibly various other controls or means of controlling containers. The implementation of step OBT_IM makes it possible to obtain a sequence in in which all the results are associated with a mold identifier, and in the figure, a sequence SEQ1 is represented where each number inscribed in a circle is a mold identifier.

[0143] The inspection apparatus 120 is equipped with a reader 125 capable only of reading mold identifiers directly on the containers (the OBT_IM step is implemented upon reading for the containers which pass in front of the reader 125). The reader 125 is for example configured to read mold identifiers marked by means of an imprint left, by the mold used, on each container formed by the mold.

[0144] In the example of the INS installation, the second device 120 is preceded by spacers 130 guaranteeing sufficient spacing between the containers which will be inspected by the second device 120. The addition of a reader such as the reader 115 can be complex to implement at the location E1, due to the fact that certain models of readers 115 require the presence of the spacer 130 and that space problems arise.

[0145] The figure shows the sequence SEQ2, which is longer than the sequence SEQ1.

[0146] The REC_AL sequence alignment search step can be implemented to search for the first sequence SEQ1 in the second sequence SEQ2. As can be seen, here we have a sequence of containers formed by the molds 12, 15, 28, 28, 32, 15, 7, 12, and 12, which forms the sequence SEQ1 and which is found identically in the sequence SEQ2, where the mold identifiers are shown in thick line circles.

[0147] The REC_AL step of sequence alignment search consists of determining, by testing several possible positions (one can also speak of offsets), a position of the sequence SEQ2 in relation to SEQ1 with the maximum number of identical molds opposite. Therefore, when the sequence alignment is found, we can consider that these are the same containers. The container formed with mold 12, the first element of the sequence SEQ1, is also the one formed with mold 12, at the third position of the sequence SEQ2.

[0148] If the reader 115 of the first device has read unique container identifiers, then it may be very advantageous to associate with this unique identifier the inspection results from the second device for this container, which is present in both sequences. In general, the inspection results, for the same containers, which come from the two inspection devices can be associated.

[0149] It can be noted that in the example above, we have a length for the sequence SEQ1 which is 9 containers, and a number of molds in the set of molds of several tens of molds. We found in the sequence SEQ2 a sequence perfectly identical to the sequence SEQ1 (zero Levenshtein distance between SEQ1 and the sequence which starts at the third position in SEQ2 and which is of length 9). There is therefore a high probability that these are the same containers.

[0150] The person skilled in the art will know how to choose the length of the first sequence based on the number of possible molds to make sequence searching possible and have good identification of the same containers. As can be seen, for a limited number of molds, it is preferable to use a longer first sequence than for a high number of possible molds.

[0151] Also, the second sequence is longer than the first sequence, and it can be for example at least twice as long.

[0152] Furthermore, the lengths of the mold sequences will be chosen in the case of a periodicity appearing in the mold identifiers: the length of the sequence SEQ2 is then chosen to be smaller than the period.

[0153] As represented in Figure 2 by a symbol surmounted by a cross, there is no mixing between the first inspection device and the second inspection device, this makes it possible to find the same containers.

[0154] For information, the implementation of the search by sequence alignment using the Levenshtein distance has been included in the table in Figure 3.

[0155] In the table in Figure 3, we see that for different consecutive positions (or offsets) POS1, ..., POS8 of the sequence SEQ1 opposite the sequence SEQ2, we determine the distance between sequence SEQ2 at this position and a sequence SEQ1 whose content appears in each line at the associated position POSi. In other words, we scan sequence SEQ2 with sequence SEQ1 and calculate the distance at each position. In the last line of the table, we have indicated the Levenshtein distances calculated for the different positions, and we observe that for position P0S4, we have a distance equal to 2, which is the smallest distance here. This distance of 2 results from only two erroneous mold identifiers.

[0156] The sequence alignment search can stop when a distance equal to zero is reached between the sequence SEQ 1 and a portion of the sequence SEQ2, or when the position in the second sequence at a lowest distance has been determined.

[0157] Preferably, in the case where the lowest distance is not zero, we can check whether this distance is less than a given distance threshold. If it is greater, then we can consider that we have not found the sequence SEQ1 in the sequence SEQ2, and there are not the same containers in the two sequences.

[0158] A non-zero distance can result from: - one or more missing containers (typically a container ejected, broken, dropped, removed); and / or - one or more false readings; and / or - one or more non-readings (reading was impossible); and / or - inserting a container between two other containers.

[0159] It can be noted that if one of the two sequences contains mold identifiers obtained in a more reliable manner than those of the other sequence, it is possible, when a reading error appears, to consider only the mold identifiers considered reliable. This will be the case in particular for mold identifiers obtained from a table addressed by another code, considered more reliable than a reading of a container relief.

[0160] Also, even when finding the position of a sequence with a distance less than the distance threshold, it is possible that the same containers are not not found. This can happen if there is no equality between mold IDs, for example.

[0161] In Figure 4, another installation is shown in which a first inspection device 110' is placed upstream on a container conveyor track 100' at a first location E1'.

[0162] The first inspection device is equipped with a reader 115' similar to the reader 125 described with reference to figure 2.

[0163] A second 120' inspection device is also installed on the conveyor track at a second location E2'. This second inspection device is here a reader of other marking, and its inspection results will be its reading (the content of the other marking).

[0164] This mode of implementation is advantageous in that it allows the information contained in the other markings to be found to associate them with the inspection results from the device 110', usually only linked to the mold identifiers.

[0165] The second inspection device 120' is arranged on the conveyor track 100'. Downstream of this second inspection device, there is a conveyor structure 200' which is not a conveyor track within the meaning of the present application in that this structure allows the mixing of the containers that it transports, as illustrated in the figure. A third inspection device 140' is arranged at a third location E3' located at the level of the conveyor structure 200'.

[0166] The method described with reference to Figures 1 and 2 cannot be implemented using the third inspection device 140' and one of the inspection devices among the device 110' and the device 120', given the mixture which is implemented within the structure 200'. In the case illustrated in Figure 4, for example if the other marking is unique, and if the third inspection device 140' is equipped with a reader 145' of the other unique marking, as is the second inspection device 120', then the association between the second and the third reader is possible without the alignment search REC_AL, despite the mixing of the containers along the conveyor 200', and since the invention allows the association by the REC_AL alignment search between the information of the second device 120' with that of the first device 110' since there is no mixing of the containers along the conveyor 100', all the information of the devices 110', 120' and 140' can be associated for each container. It is therefore understood that it is advantageous to place a reader / inspection machine 120' before any mixing.

[0167] Figure 5 shows yet another example of an installation INS" corresponding substantially to the installation INS' described with reference to Figure 4, except that the types of inspection devices are reversed. At the first location E1", there is a first inspection device 110" which is a reader of another unique marking, and at the second location E2", there is a second inspection device 120" similar to the device 110' of Figure 4. The third inspection device 140" is provided with a reader 145" of the other unique marking.

[0168] The examples in Figures 4 and 5 are advantageous in that they show how the invention can be implemented within installations equipped with inspection devices, some of which are equipped with mold identifier readers but not with readers of other markings. A reader of other markings can be implemented at any location on the same conveyor track.

[0169] In Figure 6, for the conveyor line 100, the processing of the same container 300 in the inspection device 110 and in the inspection device 120 is shown. Of course, this same container 300 is inspected at different times.

[0170] The first inspection device 110 comprises the reader 115 already described with reference to FIG. 2, a sensor 116 measuring a diameter of the containers, and a sensor 117 verifying the presence of a trapezoid-type defect.

[0171] The second inspection device 120 comprises the reader 125 already described with reference to FIG. 2, a sensor 126 checking for the presence of a glaze-type defect, and a sensor 127 checking whether the neck of the container is blocked.

[0172] After implementing steps OBT_1 and OBT_IM, there is, for the container 300 and by the first inspection device 110, an inspection result RI comprising the mold identifier (result R11), an inspection result R12 indicating the measured diameter, an inspection result R13 indicating the presence of a trapezoid-type defect, and an inspection result R14 indicating a read timestamp.

[0173] There is also, for the container 300 and by the second inspection apparatus 120, an inspection result R2 comprising the mold identifier (result R21), an inspection result R22 indicating the presence of a glaze-type defect, and an inspection result R23 indicating whether the container is blocked.

[0174] During the ASS step described with reference to Figure 1, the RI and R2 results are combined to obtain an overall RG result.

[0175] It may be noted that other information may be associated during the ASS step.

[0176] For example, if manufacturing data for container 300 is available, it can be associated during this step.

[0177] If a timestamp is included in the other marking for example, this read timestamp can be associated during this step.

[0178] If no timestamp is available, a container manufacturing time can be determined from a container inspection time by the first or second device, and then associated with that determined manufacturing time.

[0179] It can be noted that from a timestamp or a specific manufacturing moment, we can obtain manufacturing data to associate.

[0180] In Figure 7, a system 1000 is shown configured for implementing the method described above. This system 1000 is in communication with the inspection devices 110 and 120.

[0181] The system 1000 has a computer system structure and comprises a processor 1001 and a non-volatile memory 1002 in which are stored computer program instructions 1003, which, when executed by the processor 1001, lead to the execution of the method as described above.

[0182] More specifically, the computer program instructions 1003 include: - instructions for obtaining (forming an obtaining module when the program is executed) a first sequence of inspection results for a first set of containers transported consecutively on the conveyor track, using a first container inspection apparatus arranged at a first location on the conveyor track, each inspection result of the first sequence of results being associated with information relating to the mold identifier of the container read by the first apparatus on a marking of the container among said one or more markings (these instructions lead to the implementation of step OBT_1), - instructions for obtaining (forming an obtaining module when the program is executed) a second sequence of inspection results for a second set of containers transported consecutively on the conveyor track, the second set comprising more containers than the first set, using a second container inspection apparatus arranged at a second location on the conveyor track, each inspection result of the second sequence of results being associated with information relating to the mold identifier of the container read by the second apparatus on a marking of the container among said one or more markings (these instructions lead to the implementation of step OBT_2), - obtaining instructions (forming an obtaining module when the program is executed), for the first sequence of results and for the second sequence of results, mold identifiers from the information relating to the mold identifiers read (these instructions lead to the implementation of the OBT_IM step), - search instructions (forming a search module when the program is executed) by sequence alignment of the first result sequence into the second result sequence using the identifiers of molds obtained, to identify the same containers in the first set and in the second set of containers (these instructions lead to the implementation of step REC_AL).

[0183] The 1000 system is here a computer system separate from the two inspection devices. Alternatively, it can be integrated into one or the other of the inspection devices.

[0184] The method of Figure 1 has been described above for use of mold identifiers. The invention is not limited to the use of mold identifiers and can also be implemented using cavity number identifiers. For example, the forming cavity identifiers can be marked after the forming of the containers, on containers that do not have mold identifiers marked in relief. Also, the containers may not have other markings as described above that are of a unique type for each container. Containers with only marked forming cavity identifiers can thus have a marking that occupies little surface area since this cavity identifier marking is not configured to uniquely identify a container.

[0185] In Figure 8, another installation is shown in which a first inspection device 110'" is placed upstream on a container conveyor track 100'" at a first location E1"'. The containers which circulate on the conveyor track each have only a single forming cavity identifier marking.

[0186] The first inspection device is equipped with a 115'' reader capable of reading forming cavity identifiers. For example, if the forming cavity identifiers are coded and obtained by applying a laser beam, by printing, by labeling, the 115'' reader is capable of reading them to obtain the forming cavity identifier.

[0187] A second 120'" inspection device is also installed on the conveyor track at a second location E2'". This second inspection device comprises a 125'" reader which is here analogous to the 115'" reader.

[0188] When implementing the OBT_IM step adapted for use of only the forming cavity identifiers, the identifiers read by the readers 115'" and 125'" are directly obtained.

[0189] When implementing the REC_AL step adapted for use of only forming cavity identifiers, the sequence alignment search based on the forming cavity identifiers is implemented.

[0190] All the examples described above based on mold identifiers can be adapted to forming cavity identifiers. Furthermore, if an inspection device can only read one type of information between the mold identifier and the forming cavity identifier, a correspondence table can be used to implement the method based on either the mold identifiers or the forming cavities.

[0191] The invention is not limited to the examples described and shown because various modifications can be made thereto without departing from its scope.

Claims

Claims

1. A method of tracking containers transported on a container conveyor track (100, 100', 100", 100'") (300), the containers having been formed by means of a set of molds installed in forming cavities, each mold of the set of molds being associated with a mold identifier and / or each forming cavity being associated with a forming cavity identifier, and each container having one or more markings of information relating to a mold identifier and / or a forming cavity identifier used for the container, the method comprising: - obtaining (OBT_1) a first sequence of inspection results for a first set of containers transported consecutively on the conveyor track, using a first container inspection apparatus (110, 110', 110", 110'") arranged at a first location (El, El', El", El"') of the conveyor track, each inspection result of the first sequence of results being associated with information relating to the mold or forming cavity identifier of the container read by the first apparatus on a marking of the container among said one or more markings, - obtaining (OBT_2) a second sequence of inspection results for a second set of containers transported consecutively on the conveyor track, the second set comprising more containers than the first set, using a second container inspection apparatus (120, 120', 120", 120'") arranged at a second location (E2, E2', E2", E2'") of the conveyor track, each inspection result of the second sequence of results being associated with information relating to the mold or forming cavity identifier of the container read by the second apparatus on a marking of the container among said one or more markings, - obtaining (OBT_IM), for the first sequence of results and for the second sequence of results, the mold identifiers from the information relating to the mold identifiers read or the identifiers of forming cavity from the information relating to the read forming cavity identifiers, - a search (REC_AL) by sequence alignment of the first sequence of results (SEQ1) in the second sequence of results (SEQ2) using the obtained mold identifiers or the obtained forming cavity identifiers, to identify the same containers in the first set and in the second set of containers.

2. The method of claim 1, wherein the second location is upstream of the first location along the conveyor path.

3. Method according to claim 1 or 2, in which the first inspection result (RI) of the container is associated (ASS) for each same container with the second inspection result of the container (R2).

4. A method according to any one of claims 1 to 2, wherein the containers have at least one mold identifier marking and the mold identifier information read from the mold identifier marking by one apparatus (120) of the first inspection apparatus and the second inspection apparatus is the mold identifiers.

5. The method of claim 4, wherein the mold ID information read from the mold ID marking by the first inspection apparatus and the second inspection apparatus is the mold ID.

6. The method of claim 1 or 2, wherein the containers have at least one forming cavity identifier marking and the forming cavity information read from the forming cavity identifier marking by one of the first inspection apparatus and the second inspection apparatus is the forming cavity identifiers.

7. The method of claim 6, wherein the information relating to the forming cavities read from the cavity marking forming by the first inspection device and by the second inspection device are the forming cavity identifiers.

8. The method of claim 4, wherein the containers have the mold identifier marking, and another marking comprising information relating to the mold identifier and / or the forming cavity identifier, wherein the information relating to the mold identifiers read from the other marking using the other apparatus (110) of the first inspection apparatus and the second inspection apparatus are forming cavity identifiers or unique identifiers of each container, the method comprising obtaining the mold identifiers from the forming cavity identifiers or unique identifiers by means of a lookup table.

9. A method according to one of claims 6 to 8, wherein the further markings or the forming cavity identifier markings are applied to each container after the container is formed.

10. A method according to claim 8 or claim 9 taken in its connection with claim 8, wherein the other markings are codes in which the forming cavity identifiers and / or the unique container identifiers and / or the mold identifiers are encoded.

11. The method of claim 8, wherein the other apparatus of the first inspection apparatus and the second inspection apparatus is a code reader, the inspection results obtained by this other apparatus being code readings.

12. A method according to any one of claims 10 or 11, wherein a timestamp of a manufacturing step of each container is encoded in the code.

13. A method according to any one of claims 8 to 11, wherein the method comprises: 12, comprising a step of associating, for each same container, the unique identifier and / or the timestamp when the process is taken in its connection with claim 12, to the inspection result of said apparatus among the first inspection apparatus and the second inspection apparatus.

14. A method according to any one of claims 1 to 14. 13, comprising a step of associating, for each same container, manufacturing data with the first and second inspection results.

15. A method according to any one of claims 1 to 15. 14, wherein the sequence alignment search is implemented using an algorithm calculating a distance measuring the differences between the first sequence and subsequences of the second sequence all associated with different alignment positions, and searching for an alignment position that minimizes the distance.

16. A method according to any one of claims 1 to 16. 15, in which prior to transport on the conveyor track, the containers are mixed.

17. Method according to claim 16, in which the containers are mixed at the outlet of an annealing arch opening onto the conveyor track.

18. A method according to any one of claims 1 to 17, carried out by the first or second inspection apparatus.

19. A method according to any one of claims 1 to 19. 16, implemented by a computer system separate from the first and second inspection devices.

20. System for tracking containers transported on a conveyor track (100, 100', 100", 100'") of containers (300) formed by means of a set of molds installed in forming cavities, each mold of the set of molds being associated with a mold identifier and / or each forming cavity being associated with a forming cavity identifier, and each container having one or more markings of information relating to a mold identifier and / or a forming cavity identifier used for the container, the system comprising: - a module for obtaining (1001, 1003) a first sequence of results inspection for a first set of containers transported consecutively on the conveyor track, using a first container inspection apparatus (110, 110', 110", 110'") arranged at a first location (El, El', El", El'") of the conveyor track, each inspection result of the first sequence of results being associated with information relating to the mold or forming cavity identifier of the container read by the first apparatus on a marking of the container among said one or more markings, - a module for obtaining (1001, 1003) a second sequence of inspection results for a second set of containers transported consecutively on the conveyor track, the second set comprising more containers than the first set, using a second container inspection apparatus (120, 120', 120", 120'") arranged at a second location (E2, E2', E2", E2'") of the conveyor track, each inspection result of the second sequence of results being associated with information relating to the mold or forming cavity identifier of the container read by the second apparatus on a marking of the container among said one or more markings, - a module (1001, 1003) for obtaining, for the first sequence of results and for the second sequence of results, mold identifiers from the information relating to the read mold identifiers or forming cavity identifiers from the information relating to the read forming cavity identifiers, - a search module (1001, 1003) by sequence alignment of the first sequence of results (SEQ1) in the second sequence of results (SEQ2) using the obtained mold identifiers or the obtained forming cavity identifiers, to identify the same containers in the first set and in the second set of containers.

21. A computer program comprising instructions for carrying out the steps of a method according to any one of claims 1 to 19, when said program is executed by a computer.

22. A computer-readable recording medium having recorded thereon a computer program comprising instructions for carrying out the steps of a method according to any one of claims 1 to 19.

Images