Method for monitoring containers with communication between a hot sector and a cold sector

Abstract

The invention relates to a method for monitoring glass containers circulating from a hot sector of a facility for manufacturing glass containers to a cold sector of the facility, wherein the following is implemented, for one of the containers: - an inspection (DET_D1) implemented by a first detection device (7H) configured for the inspection of containers circulating within the hot sector and configured for detecting the presence of at least one defect of a given type, - a transmission (TX_M) by the first detection device of a message produced according to an inspection result by the first detection device, - a reception (RX_M), by a second detection device (7C) configured for the inspection of containers circulating within the cold sector, of an alert message, - a modification (MOD_P), by the second detection device, of at least one parameter for detecting defects of the given type.

Description

Description Title of the invention: Method for controlling containers with communication between a hot sector and a cold sector Technical Field The present invention relates to the field of manufacturing containers formed using molds, for example bottles, jars, or vials, and more specifically to the detection of defects affecting these containers. The invention also relates to the control of manufacturing parameters to avoid these defects. Previous technique

[0002] Glass container manufacturing plants include at least two sections: the cold section and the hot section. These two sections are separated by an annealing arch, with containers passing through this arch as they move from the hot section to the cold section. In the cold section, the containers cool down after passing through the annealing arch. [0003JII There are inspection machines for the cold sector and inspection machines for the hot sector. Inspection machines for the cold sector are designed to reject all defective containers, while machines for the hot sector can only detect some critical defects. In fact, inspections are more difficult in the hot sector, and some defects are currently undetectable in this environment. Inspection machines for the hot sector are generally used to detect critical defects as early as possible and limit the risk of sending them to the customer.]

[0004] The vessels are generally inspected in the cold section approximately 35 minutes to 2.5 hours after they are formed (in particular, because it is easier to handle cooled vessels). The passage through the annealing arch can take an average of one hour.

[0005] Furthermore, in container manufacturing facilities, defect detection can sometimes be detrimental to production: for example, it may be advantageous to use particularly sensitive devices to prevent a defective container from being considered acceptable. This can lead to false positives, negatively impacting container production. [0006JII it is therefore necessary to find a good balance between the quality of defect detection and the production of containers. Description of the invention

[0007] This presentation concerns a method for inspecting glass containers circulating from a hot section of a glass container manufacturing installation to a cold section of the installation (the glass containers manufactured within the installation are those circulating from the hot section to the cold section), in which the following is implemented for one of said containers: - an inspection carried out by a first detection device configured for the inspection of circulating vessels within the hot sector and configured for the detection of the presence of at least one defect of a given type, - the transmission by the first detection device of a message generated by the first detection device, based on an inspection result from the first detection device. - the reception by a second detection device configured for the inspection of circulating containers within the cold sector, of an alert message, the reception by the second detection device being implemented after said emission by the first detection device, - a modification, by the second detection device, of at least one fault detection parameter of said given type (thus, the second detection device is also configured for the detection of the presence of at least one fault of the given type), the modification being triggered by said reception.

[0008] This process can be implemented at least by the first and second inspection devices, each having a computer system structure.

[0009] An inspection result delivered by a detection device can be information indicating the presence or absence of a defect, for example, a flag indicator. An inspection result can also be information indicating the presence of a defect combined with location information, for example, location information indicating where the defect is located within an image acquired by the detection device (for example, by means of a bounding box) or location information indicating where the defect is located within the container (for example, at the container's ring). An inspection result can also be an image acquired by the detection device (this image may or may not contain a defect; typically, it may be an image that has not yet been analyzed by the inspection device).

[0010] An inspection result may also include an instruction targeting the type of defect, for example an instruction to modify the inspection parameter (here, the instruction is an inspection result because its elaboration is triggered by an inspection).

[0011] The message sent may include this inspection result.

[0012] The method described above proposes using the inspection results of a device in the hot sector to modify how defects are detected in the cold sector. This approach improves defect detection by inspection machines because it is implemented automatically. For example, rules can be defined to determine how the detection parameter is modified based on the detection results of the first inspection machine (for example, the rule can be implemented using a lookup table or a mathematical function).

[0013] In fact, the inventors of the present invention observed that images transmitted from inspection devices to other inspection systems within existing installations generally result in a simple display of the images on a human-machine interface (on devices in cold or hot sectors). No automatic decision-making process is taken to impact detection, particularly in cold sectors.

[0014] By modifying a detection parameter, one can adjust, for example, the detection sensitivity, and thus reduce the number of false negatives in the cold sector for only a specific period, so that this has minimal impact on the installation's production. Indeed, increasing sensitivity helps avoid false negatives but increases the number of false positives detected.

[0015] Here, by using information from the hot sector, we can modify the downstream detection for containers that will later arrive in the cold sector.

[0016] It should be noted that detection devices for hot sectors are generally less accurate than those for cold sectors. Therefore, information such as the detection of a fault in a hot sector corresponds to a situation where the fault is evident, which should prompt an increase in sensitivity for cold sector detection.

[0017] According to a particular implementation method, the message emitted includes a result indicating the presence of at least one defect of the given type within at least one container inspected by the first detection device.

[0018] In this particular mode of implementation, the message indicates the presence of a fault, but also the type of this fault, which is the given type.

[0019] According to a particular implementation method, the message emitted includes at least one image of a container acquired by the first detection device.

[0020] This method of implementation is particularly advantageous for enabling the implementation, in the second device, of image detection processing based on parameters other than those of the first detection device.

[0021] In one particular implementation, the detection parameter is a detection sensitivity, and the modification is an increase in said detection sensitivity. It should be noted that an increase in sensitivity results in an increase in the number of rejections.

[0022] According to a particular implementation method, the process further includes an inspection configured for the detection of the defect of said given type and implemented by the second detection device using said modified parameter.

[0023] Thus, in this particular implementation mode, after the parameter is modified, a detection is implemented for a container which may provide a different result from the result that would have been obtained before the parameter was modified.

[0024] According to a particular implementation method, said at least one detection parameter is modified for a given period, the given period having a start time separated from said reception by a given delay.

[0025] This particular implementation method allows, by choosing an appropriate delay, the selection of the point at which the parameter will be modified. In fact, it is known that a container, after its inspection in the hot zone, will only return to the cold zone approximately one hour later (due to its passage through the annealing chamber). It therefore appears advantageous to choose the given delay based on the duration of the passage through the annealing chamber.

[0026] By not immediately changing the detection parameter, we avoid over-detection of defects over a period of time, for example, during a period unrelated to what was observed in the hot sector. Over-detection refers to a high number of false positives, which is problematic in terms of production because some defects, if detected, lead to the disposal of containers.

[0027] We can note that the period can be fixed or dynamic, for example it can be defined according to an event such as an intervention within the container forming system (such as a mold change).

[0028] According to a specific implementation method, the period is determined as follows. Each container bears a unique identifier of two possible types, allowing its exact manufacturing time to be determined. In the first case, the unique identifier includes a timestamp indicating the date or time of the container's manufacture, and it may also contain the mold number used to form the container, or the section number and / or cavity number where the container was formed. This information can be encoded, for example, in a Datamatrix code. In the second case, the identifier is a serial number that allows the manufacturing time to be read from a manufacturing memory or database, as well as other manufacturing data for each container, such as the mold, section, or cavity number.In this implementation, the second device includes, or is connected to, an identifier reader, which allows for the acquisition of the manufacturing timestamp—that is, the exact manufacturing time of each inspected container, down to the minute or even the second. This optimizes the parameter modification time of the second device by applying the change only between the passage of the first container, whose manufacturing timestamp is the initial time of the period, and the passage of a second container, whose manufacturing timestamp is also the initial time of the period. This improves the plant's production capacity.

[0029] In an alternative implementation, the period is not defined by a start and end time, but by a manufacturing condition of the inspected containers. For example, a potentially defective manufacturing period is defined. Since the second device is capable of knowing the manufacturing timestamp of the containers, the modification only applies to any container whose timestamp falls within a given time interval, regardless of when it passes through the second device. Alternatively, the modified setting can be applied to the inspection of any container whose timestamp falls within a given manufacturing period, with a mold, section, or cavity number determined by the received message.According to this implementation, if containers manufactured during a production period without a defect detected by the first device circulate in the second device at the same time as containers manufactured during the period with the aforementioned type of defect detected, the modified setting is applied only to the containers manufactured during the defective production period. According to this embodiment, the modified and unmodified settings are used alternately depending on the container inspected, i.e., according to the timestamp.

[0030] According to a particular implementation method, the process includes obtaining, by the first detection device, an identifier of a mold used or a forming cavity used for the manufacture of the container (here, it may be a defective container if a defect was detected during said inspection), said message emitted further including the identifier of the mold or cavity, - the message received by the second detection device, which also includes the mold or cavity identifier, and - the process further comprising obtaining, by the second detection device, an identifier of a mold used or a forming cavity used for the manufacture of a container controlled by the second detection device, said modification being implemented if the identifier of the mold or forming cavity of the container controlled (by the second inspection device) corresponds to the identifier of the mold or forming cavity included in said received message.

[0031] Detection devices can be equipped with readers configured to read mold or cavity identifiers. For example, these identifiers are marked by a relief specific to the mold, or can be marked by a two-dimensional code, for example applied by laser marking.

[0032] Some defects can result from using a faulty mold or an incorrect forming parameter for a given forming cavity. Therefore, this particular implementation allows the detection parameter to be modified only for containers formed with the same mold or in the same forming cavity. Thus, for the second device, the modified and unmodified settings are used alternately depending on the container being inspected, i.e., according to mold or cavity identifiers. These two settings can be stored within the second device.

[0033] Thanks to the reading of the identifiers for the second device, the modified setting and the unmodified setting are used alternately depending on each container inspected, i.e. according to the original mold or cavity identifiers of each container and according to the timestamp of its production date.

[0034] According to a particular implementation method, the message emitted by the first detection device is said message received by the second detection device.

[0035] In this particular implementation, the first detection device generates and transmits a message, which is then received by the second detection device. This implementation is advantageous in terms of simplicity. The message can be communicated via a wired or wireless interface, or through a communication network (for example, an internal network within the installation).

[0036] According to a particular implementation method, the message emitted by the first detection device is received by an installation controller, the installation controller developing and emitting the message received by the second control device.

[0037] In this particular implementation, there is no direct transmission between the first and second detection devices. This implementation is suitable for a system controller (which may have a computer system structure) capable of processing the information needed to determine the message to be generated for the second detection device.

[0038] According to a particular implementation method, the first detection device is configured to acquire one or more images of a controlled container using one or more cameras, in the visible range, in particular by using a light source, or in the infrared (without a light source but using infrared sensors sensitive to the radiation from hot containers).

[0039] According to a particular implementation method, the second detection device is configured to acquire one or more images of a controlled container using one or more cameras, by illuminating the container with a light source, particularly in the visible range.

[0040] According to a particular implementation method, the first detection device and the second detection device are each configured to acquire an image of a container along the same observation direction at plus or minus 45°.

[0041] As an example, this can be achieved by using a plurality of cameras, for example a plurality of cameras in the cold sector so that at least one camera can acquire an image at plus or minus 45° of the observation direction used in the hot sector.

[0042] According to a particular implementation method, the first detection device and the second detection device are each configured to acquire an image of the same portion of the containers.

[0043] The portion could be, for example, the ring of the container, the bottom of the container, or even the body of the container.

[0044] According to a particular implementation method, the first detection device and the second detection device implement the same classifier or a different classifier but configured for classification according to said given type.

[0045] The classifier is used to classify the defects present or absent within a container.

[0046] The first detection device implements a first classifier, and the second detection device implements a second classifier. Both the first and second classifiers are configured to classify, if a container has a defect of a given type, the defect as being of that given type.

[0047] The two classifiers may nevertheless be different, for example have a different complexity (such as a different number of parameters), or be able to handle different classes (although the given type is common to both classifiers).

[0048] As a guide, classifiers include deep learning models. In this case, they may differ in their parameters.

[0049] According to a particular implementation method, the message is issued if a given number of detections implemented by the first detection device for several containers indicate the presence of at least one defect of the given type, during a given time window.

[0050] For example, in this particular implementation mode, an isolated detection of the presence of a defect of a given type does not lead to the development and issuance of the message, but if several containers present this defect in the same time window, this can be considered critical and leads here to the issuance of the alert message.

[0051] The invention also proposes a system for monitoring circulating glass containers from a hot section of a glass container manufacturing installation to a cold section of the installation, the system comprising a first detection device configured for inspecting circulating containers within the hot section and a second detection device configured for inspecting circulating containers within the cold section, wherein the first detection device is configured to implement: - an inspection, - the transmission by the first detection device of a message generated based on an inspection result by the first detection device, and in which the second detection device is configured to implement: - the reception of an alert message, the reception by the second detection device being implemented after said transmission by the first detection device, - a modification of at least one fault detection parameter of said given type, the modification being triggered by said reception.

[0052] This system can be configured to implement the process as defined above according to any of the implementation modes.

[0053] The aforementioned features and advantages, as well as others, will become apparent upon reading the detailed description that follows, along with examples of implementations of the process and system defined above. This detailed description refers to the attached drawings. Brief description of the drawings

[0054] The attached drawings are schematic and are primarily intended to illustrate the principles of the presentation.

[0055] In these drawings, from one figure to another, identical elements (or parts of elements) are identified by the same reference symbols.

[0056] [Fig. IA] Figure IA shows an installation comprising a manufacturing system and detection devices.

[0057] [Fig. IB] Figure IB is a top view of the installation of Figure IA.

[0058] [Fig. 2] Figure 2 is a schematic representation of the steps of a process according to an example.

[0059] [Fig. 3] Figure 3 is a schematic representation of the steps of a process according to another example.

[0060] [Fig. 4] Figure 4 is a schematic representation of the steps of a process according to yet another example.

[0061] [Fig. 5] Figure 5 is a schematic representation of a hot sector inspection device.

[0062] [Fig. 6] Figure 6 is a schematic representation of a cold sector inspection device. Description of the implementation methods

[0063] We will now describe the control of a glass container manufacturing plant. In particular, we will describe the detection of defects in the hot sector and the control systems that can be used to detect these defects. We will also describe the determination of manufacturing system parameters related to these defects, and finally, the issuing and execution of commands affecting these parameters.

[0064] Figures IA and IB schematically represent an INS installation for manufacturing transparent or translucent glass containers. Figure IA shows the installation from the side, and Figure IB from the top. The installation is shown so that both the hot and cold sections are at least partially visible.

[0065] A SF manufacturing system is shown in the figure. This system manufactures generally transparent glass containers of all known types. At the output of the SF manufacturing system, the containers, such as glass bottles or flasks, exhibit a high temperature typically between 300°C and 600°C.

[0066] As is known, the containers 2, which have just been formed by the SF manufacturing system, are successively placed on an output conveyor 5 to form a line of containers. The containers 2 are transported in a line by the conveyor 5 in a direction of travel F in order to convey them successively to different processing stations and in particular an annealing arch 6, upstream of which is placed a first detection device 7H configured to inspect the containers in the hot sector and in particular configured to inspect for defects of a given type.

[0067] The 7H detection device can perform inspections of the body of the containers, or inspections of the rings of the containers, both of which are carried out during the translation of the containers in motion.

[0068] These inspections can be carried out using cameras configured to acquire the infrared radiation emitted by the hot containers, and they can also be carried out using cameras configured to acquire visible radiation, with a light source illuminating the containers (images can be acquired, relative to this light source, in reflection, plunging, transmission, etc.).

[0069] Defects that can be detected in the hot sector can include ring burr defects, non-returned ring defects, trapezoids ("bird swing" in English), fins, inclusions, burst bubbles, deformations or thins.

[0070] Preferably, the 7H detection device operates in transmission mode with a visible or near-infrared light source and / or without a light source but uses cameras to acquire the infrared radiation emitted by hot containers. If both solutions are implemented, the 7H detection device incorporates multiple cameras. It can also monitor the ring by analyzing the light emitted by a light source and reflected off the ring's surface.

[0071] It can be noted that hot inspection is advantageous in that it allows for rapid reaction to modify parameters of container forming in the SF manufacturing system.

[0072] It can be noted that inspection using infrared radiation makes it possible to detect certain defects such as wings, open blisters, uneven glass distribution.

[0073] The 7H inspection device can be of the type described in document FR3131634.

[0074] The SF manufacturing system comprises several distinct forming sections 12, each containing at least one roughing mold 13 and at least one finishing mold 14. The SF system includes a source 16 of malleable glass, i.e., hot glass, and a glass droplet dispenser 17. (Gobs) which distributes, by gravity, drops of malleable glass 18 to each roughing mold 13. As is known, the source 16 of malleable glass is a reservoir supplied with molten glass, at the bottom of which is a basin having one to four circular opening(s). A rotating tube, the height of which is regulated, controls the flow of glass above the basin, and a system of one to four plungers, moving back and forth, extrudes the glass through the one to four opening(s) of the basin in order to deliver, by gravity, the malleable glass in the form of one to four parallel strings.The malleable glass strings are definitively separated into independent drops by a scissor system 19 arranged at the outlet of the hot glass source 16 and which is operated at regular intervals to cut the malleable glass from the source 16 into sections (this scissor system is automatically controllable, for example to change the weight or shape of the section).

[0075] For systems with multiple (up to four) molding cavities per section, several segments are delivered in parallel and simultaneously. In this description, a parison 18 is an extruded droplet or segment of malleable glass as cut by the scissor system 19. In English, at this stage of a forming process, the parison is called a "gob." The malleable glass, at the point of cutting by the scissor system 19, generally has a temperature above 900°C, for example, between 1100 and 1300°C. This parison is essentially a solid cylinder of malleable glass having a volume and length defined by the setting of the source 16 cooperating with the cutting action of the scissor system 19. Indeed, the diameter of the parisons is defined by that of the openings in the gob.The flow rate is controlled both by the height of the tube, which affects the overall flow rate, and by the movements of one to four plungers, allowing the flow rate to be varied separately for each opening of the trough. The time interval between two actuations of the scissor mechanism 19 determines the length of the parison. In summary, the length, weight, and volume of each parison are determined by the parameters of the source 16 (the tube and the plungers) and the scissor mechanism 19. The malleable glass source 16 is positioned above the blank molds 13 to allow the gravity-fed distribution of the parisons, which are loaded through openings 22 in the upper faces of the blank molds 13.

[0076] The distributor 17 extends via several branches between the hot glass source 16 and the roughing molds 13 of each forming section. Generally, the hot glass source 16, through the scissor system 19, simultaneously delivers as many parisons as there are roughing molds (or finishing molds) in a forming section. It is therefore understood that the forming sections are supplied with parisons successively, one after the other.

[0077] The distributor 17 therefore collects the parisons cut by the scissor system 19 and conveys them to each of the roughing molds 13 of each of the forming sections 12 according to a corresponding loading path. The loading paths for the different roughing molds 13 include common portions and specific portions. A specific portion is a part of the loading path corresponding to a roughing mold 13 that is followed only by the parisons directed by the distributor towards that roughing mold.

[0078] The distributor 17 therefore includes a diverting mechanism, which is a type of pivoting chute or group of chutes, and a parison guidance system with chutes and deflectors at the end of its travel, above the roughing molds. Specifically, the position of the deflectors relative to the associated roughing molds partially determines the position and orientation of the loading of each parison into said roughing molds. Within the distributor, the chutes, deflectors, and diverters determine the loading trajectory of the parisons.

[0079] Glass container manufacturing systems employ various processes combining successive filling, pressing, and / or blowing steps. For clarity, the example is taken from container forming using the well-known processes known as BB (Blow-Blow), PB (Press and Blow), or NNPB (Narrow Neck Press and Blow, adapted for containers with a narrow opening).

[0080] In container manufacturing systems, each forming section 12 can include several molds, for example, two molds, one a roughing mold 13 and the other a finishing mold 14. Each section 12 can include a set of roughing molds and a set of associated finishing molds. In this case, a given parison is guided by the distributor 17 towards a roughing mold, for example, a roughing mold 13 of the forming section, where the parison undergoes a first forming operation, called drilling, carried out by compressed air blowing or by punch penetration. A transfer system (not shown) is then able to take the parison which has undergone the first forming operation, namely the blank (“parison” in English), from the roughing mold 13 to take it to a finishing mold 14 where the blank can undergo at least a second forming operation, the last operation called finishing. Generally, each roughing or finishing mold of a forming section comprises two half-molds respectively (half-molds 13a and 13b are visible in Figure IB) which are movable relative to each other in a direction perpendicular to a parting line by which the two half-molds are in contact in a closed position. In the illustrated example, the parting line extends along the vertical Z direction and the transverse X direction.

[0081] A section 12 may include a single finishing mold 14 receiving a blank from a single roughing mold 13. However, as mentioned above, each of the different forming sections 12 may include at least two distinct finishing molds 14 and as many roughing molds 13. The case of four forming sections 12 offset along a longitudinal direction Y perpendicular to the transverse direction X is illustrated in the Figures. According to this example, each forming section 12 includes three roughing molds 13 respectively front, central, and rear (or external, central, and internal), each associated with a finishing mold 14 respectively front, central, and rear, that is, each receiving the blank from a roughing mold 13. In the illustrated example, the different roughing molds 13 and respectively the finishing molds 14 of the same section are offset from each other along a transverse direction X.In the illustrated example, the finishing molds 14 of the same section are of identical shape, therefore generally intended to form identical containers, but different shapes and weights could be provided.

[0082] It should be noted that each finishing mold 14 is identified in the forming installation in relation to the other finishing molds 14. Similarly, each roughing mold 13 is identified in the manufacturing system. It is thus possible to identify the forming section 12, the roughing mold 13, and the finishing mold 14 from which each container 2 originates.

[0083] In a glass container manufacturing system, each blank mold location 13 in each section bears, according to various possible conventions, an identifier, for example, a number or a letter. These locations can be referred to as forming cavities, which are identified by a forming cavity number.

[0084] Furthermore, the finishing molds can be imprinted with a pattern to emboss the mold number (e.g., from 1 to 99 or from 1 to 128, etc.) onto the containers. A lookup table linking the forming cavity numbers to the mold numbers is permanently available to operators or the plant's information system. In some installations, a laser marker is used as described in the patent. EP 2 114 840 B1 in order to print on each container, still hot immediately after its forming, a code indicating the mold number or the section and cavity numbers of the forming as well as a timestamp, that is to say the precise moment of manufacture.

[0085] Thus, containers generally bear the mold number or forming cavity number either in coded form (barcode, dot matrix code, Datamatrix code) or alphanumeric form. To read these mold or forming cavity numbers on the containers, for example in cold storage environments, various optical reading systems exist for production lines, such as those described in EP 1 010 126, EP 2297 672, or EP 2992315.

[0086] Thus, in this description, it is understood that identifying the finishing mold from which a sample container originates amounts to knowing either the forming cavity number or the mold number. It is understood that identifying the finishing mold allows for the direct identification of the associated roughing mold that produces the blank.

[0087] In forming installations, the control and synchronization of parison forming operations, scissor cutting, mold movements, punch movements, blowing, transfers, etc. are carried out by means of a control device 200 in the general sense, allowing the control of the various mechanisms necessary for the operation of the installation for the implementation of the container forming process.

[0088] Although mechanical and pneumatic controls are still used in older installations, the control system generally has a structure similar to that of an automaton or a computer and includes, for example, a processor and non-volatile memory (in any form, for example arranged within the same semiconductor chip, arranged within separate chips, etc.) in which computer program instructions executable by the processor can be stored.

[0089] The formed containers are removed from the finishing molds, placed on holding trays, and then transferred by means of pivoting articulated arms called "pushers" onto a linear chain conveyor to the entrance of the annealing arch 6. They are pushed onto the slowly advancing arch's conveyor belt. Inside this annealing furnace, the temperature of the containers is gradually raised, and they are then cooled slowly enough for the uniform cooling to release thermal stresses. The time spent in the arch is therefore relatively long, frequently ranging from 45 minutes to 1 hour 30 minutes, depending on the mass of glass.

[0090] Figures IA and IB show the hot sector extending up to the annealing arch 6, and part of the sector extending beyond the annealing arch 6, i.e., the cold sector. Here, a second detection device 7C is shown in the cold sector.

[0091] Inspection in cold environments is generally easier to implement than in hot environments. This is because it is possible to handle the containers and also to get closer to them to acquire images.

[0092] Here, the second detection device is configured to detect defects of the given type.

[0093] The 7C detection device can perform inspections of the body of the containers, or inspections of the rings of the containers, both of which are carried out during the translation of the containers in motion or even during a rotation of the containers.

[0094] These inspections can be carried out using cameras configured to acquire visible light, with a light source illuminating the containers (images can be acquired, relative to this light source, in reflection, downward, transmission, etc.). It should be noted that wall thickness measurements are generally performed using optical sensors that include at least one line-of-sight camera.

[0095] The 7C detection device can detect many defects, but it can detect at least one defect in common with the 7H detection device, for example, a burr defect on the ring, a non-rendered ring, a trapezoid, fins, an inclusion, burst bubbles of deformation, or a thin zone.

[0096] As an example, the 7C detection device can include from 6 to 18 cameras. Preferably, the number of cameras is chosen so that at least one observation direction of a camera is at plus or minus 45° of the observation direction of an image acquired by the 7H hot sector detection device.

[0097] Information can be communicated directly or indirectly via a COM communication method shown in the figure, connecting the first and second detection devices. Specifically, the figure depicts a COM communication method that enables indirect communication between the first detection device 7H and the second detection device 7C. This communication takes place through a controller in the CI system.

[0098] The first detection device 7H communicates with the CI installation controller through a communication means COM_H, and the second detection device 7C communicates with the CI installation controller through a communication means COM_C. The CI installation controller, and the communication means COM_C and COM_H together form the COM communication means.

[0099] Alternatively, communication is direct between the two detection devices, as represented by the COM_D link shown as a dashed line in the figure. Direct communication can be implemented using wired or wireless methods known per se.

[0100] In what follows and with reference to figures 2 and 3, we will describe the emission of messages between the hot sector and the cold sector to improve fault detection and the productivity of the installation. [Figure 2 shows the steps of a process according to an example. This process is implemented here by the first detection device 7H and by the second detection device 7C described with reference to figures IA and IB (but, in the illustrated case, the communication is direct between the two detection devices).

[0102] In the first step DET_D1, the first detection device carries out an inspection of at least one container. If the container has a defect of a given type, the test step CT is carried out in which it is verified whether a given number of detections of this defect of the given type have been carried out (in other words, the number of occurrences of the defect is counted) during a time window (for example a sliding time window).

[0103] If the given number has not been reached, then the DET_D1 step is implemented again to reach another container containing the defect.

[0104] If the specified number is reached, the first detection device generates a message for transmission (TX_M step). This message generation is based on an inspection result from the first detection device. The inspection result may include an indication that the specified number of defects were detected within the time window, acquired images of the containers with the defects, or information on the location of the defects. [0105JA As an example, the given number (of fault occurrences during the time window) can be 1 for a fault of a given type, such as a fin considered a critical fault. In other words, the message can signify a specific alert, such as the hot detection of a critical fin-type fault.

[0106] The message may include, for example, all or part of the inspection result. The message may also, particularly in the example shown in this figure, include instructions for the second inspection device.

[0107] The message is then transmitted to the second inspection device 7C, which receives the message at the RX_M stage. This transmission is implemented through the COM communication method described above. In particular, in the example in Figure 2, the COM communication method can be a direct communication method between the two inspection devices.

[0108] As mentioned above, the message may contain instructions for the second inspection device. This is the case in the example shown in this figure, and these instructions may tell the second inspection device that a fault detection parameter of a given type needs to be modified. In this case, these instructions are implemented directly at the MOD_P step where the parameter is changed. This parameter could be, for example, a detection sensitivity that may need to be adjusted due to the detection of the fault in the hot sector.

[0109] During the parameter modification in the MOD_P step, the parameter, which may have a value, changes from an initial value to a modified value. The initial value has been used for at least one container inspection prior to the implementation of the MOD_P step. In some implementation methods, the initial value is stored by the second inspection device. After a specified period, calculated from the implementation of the MOD_P step, during which the modified parameter value is used for inspections, the parameter can be modified again to revert to the initial value. Subsequent inspections will then be performed using the initial value.

[0110] As an example, sensitivity can be a size criterion (for example, the size of the defect), a threshold criterion beyond which a detection score indicates the presence of the defect, or any other type of sensitivity. For example, if the given type is a fin, then the second device will be made very sensitive to fins, meaning, for instance, that even with a low confidence level fin detection by the second device, the container will be ejected. This can be explained in this example because fins are sometimes mistaken for mold seals. The more sensitive setting could then consist of temporarily rejecting certain heavily marked mold seals that resemble fins. [011 l]The second device being capable of classifying defects according to at least one given type, preferably several given types, the sensitivity of the second device can be different for each type and the modification of the parameter is the modification of the sensitivity for a given type.

[0112] At the DET_D2 step following the modification, for a container (for example the container for which a DET_D1 detection was implemented triggering the emission of the message or another container other than the one for which the DET_D1 detection was implemented for which a message was emitted or not), the modified detection parameter is used to implement an inspection by the second detection device.

[0113] Preferably, the MOD_P step, in the two examples described here, is triggered by the reception of the RX_M message but is not implemented immediately; it is implemented only after a specified delay. This delay corresponds to the time it takes for a container to move from the first inspection device to the second inspection device, passing through the annealing arch (approximately one hour). Therefore, the modification of the detection parameter can be temporary; it can be applied for a given period. This given period can be chosen to encompass the containers surrounding those for which the DET_D1 step was implemented.

[0114] The first detection device can also obtain a mold or forming cavity identifier used for one or more containers in the DET_D1 step. This identifier is then added to the transmitted message. The MOD_P modification can apply only to containers formed by the same mold or within the same forming cavity. This is particularly advantageous because containers formed by other molds or forming cavities, which are assumed not to produce the given type of defect, are not affected by an unnecessary change to the detection parameter, thus improving plant production (fewer containers are scrapped). An inspection is then carried out with the modified parameter during the DET_D2 step.

[0115] Figure 3 shows another example of an implementation similar to that of Figure 2, but differing in that the message does not contain instructions. It does, however, contain all or part of the inspection result.

[0116] After receiving RX_M, the 7C detection device then performs a PROC step to process the received message and determine instructions to be applied, which may lead to the implementation of the MOD_P modification step. For example, the PROC step may correspond to identifying the type of fault identified in the message, or to processing an image or a portion of an image to identify a fault of the given type.

[0117] Figure 4 shows another example of implementation which differs from that of Figure 2 in that the first detection device transmits its message to the CI installation controller described with reference to Figures IA and IB, which receives it at the C_RX_M stage.

[0118] The installation controller can have the structure of a computer system and it can receive messages based on inspection results from several installation inspection devices.

[0119] Here, it implements a C_PROC step analogous to that described above with reference to Figure 3, before implementing a C_TX_M step of transmitting a message to the second detection device 7C, containing instructions to modify the parameter for detecting faults of the given type.

[0120] The second detection device 7C receives this message at the RX_M stage, and the following stages are analogous to the MOD_P and DET_D2 stages described with reference to Figure 2.

[0121] In the example in Figure 4, the means of communication is indirect and passes through the installation controller. This controller and the detection devices can communicate with each other via a wired or wireless local area network. In some embodiments, the installation controller can simply relay the message emitted by the first detection device to the second detection device, but it can also perform processing and store information such as: - Arch times (the time it takes a container to pass through the annealing arch) allowing the determination of delays between emission, reception and modification, - the values ​​of parameters, whether modified or not (product records), - one or more given type(s) of defects for which the process must be implemented.

[0122] Figure 5 schematically illustrates the structure of a first detection device, based on an example. The first detection device, 7H, is traversed by the moving containers 2. A CAMVIS camera is shown in the figure, configured to take downward-facing images of the ring 2B of the container 2 being inspected, illuminated by reflection from a visible light source EC directed downwards towards the ring 2B. In this embodiment, the first inspection device is capable of performing two different inspections. A first inspection is carried out using the lighting system, with the CAMVIS camera configured to acquire the visible radiation from the reflected-illuminated ring. The first device also includes another camera, CAMIR, configured to acquire the infrared radiation emitted by the hot containers. The CAMIR camera is configured to observe the body 2C of the containers.These cameras can be configured to acquire two-dimensional images.

[0123] The CAMVIS camera can be sensitive to visible light and, in some cases, to near-infrared light down to 900nm. The EC light source, used to illuminate containers by reflection or transmission, can emit visible light with wavelengths up to 900nm in some cases.

[0124] The CAMIR camera can be sensitive to radiation from containers and therefore to wavelengths preferably greater than 1.1pm.

[0125] The first detection device 7H has a computer system structure and includes a processor 100H and a memory 101H. In the memory, computer program instructions (INS) are stored for the implementation of the process described above in Figures 2 to 4 when these instructions are executed by the processor 100H. Also, the memory 101H contains inspection results (RI) (typically, these may be images or portions of images, or inspection verdicts such as detected defects or classes of defects, or measurements of compliant or non-compliant containers).

[0126] Communication via COM communication means is carried out by a 102H communication module of the network controller type.

[0127] The first detection device is typically installed near the exit of the manufacturing machine, so that it can inspect the containers moving along the outfeed conveyor. This device is connected to the manufacturing machine and synchronized with its cycles to determine the original section and cavity of each container, as well as its manufacturing time. The first detection device can therefore include a mold identifier, cavity section identifier, and / or a timestamp in its transmitted message.

[0128] It should be noted that the same connection and synchronization methods are used by container marking devices that apply a two-dimensional 2ID code (for example, Datamatrix code) to uniquely identify the containers. These hot-stamping devices are installed at the output of the manufacturing machine, either upstream or downstream of the first inspection device.

[0129] If a hot stamping device is installed upstream of the first inspection device, the containers shown in this figure will have a two-dimensional 2ID code (e.g., Datamatrix code) obtained by marking the containers, which allows for their unique identification. The first detection device may include a RIDH reader capable of extracting a mold number or a cavity (or section) number used in the forming of each container from the 2ID code. This number can be added to the message transmitted by the first detection device.

[0130] Figure 6 schematically represents the structure of a second detection device as an example. The second detection device 7C is traversed by the passing containers 2. A camera CAM1 is shown in the figure, configured to take images of the ring 2B of the container 2 being inspected, either from above or below, with a light source EC1 cooperating with the camera for reflected light operation. [013 l] In this figure, a second camera CAM2 is also shown, configured to acquire images of the container body 2C in transmission with a light source EC2. Such a camera and light source configuration is also conceivable in the hot sector. The images acquired by the CAM2 camera can be close to those acquired by a similar camera within the first detection device. However, using a plurality of CAM1 cameras distributed around the vertical axis passing through the container 2 in the figure makes it possible to obtain an image of the container body at approximately 45 degrees from the observation direction used in the hot sector.

[0132] Here, the light sources and the CAM1 and CAM2 cameras can operate in the visible or near-infrared range.

[0133] The second detection device, 7C, has a computer system structure and comprises a processor, 100C, and a memory, 101H. Computer program instructions (INS) are stored in the memory for implementing the process described in Figures 2 to 4 when these instructions are executed by the processor, 100C. The memory, 101C, also contains a PAR inspection parameter that can be modified. The memory, 101C, can also store inspection results, such as images, or inspection verdicts, such as detected defects or defect classes, or measurements of compliant or non-compliant containers.

[0134] Communication via COM communication means is carried out by a 102C communication module of the network controller type.

[0135] Also, the two containers in this figure have a code that identifies a mold number or a forming cavity number for each container. This code is generally applied by molding during the forming process, as each mold has a unique engraving that can be read to obtain the mold number. The second detection device includes a RIDC reader capable of extracting from the code a mold number or a cavity (or section) number used for forming each container. This number can be compared to a number contained in a message received by the second detection device to implement a modification. The RIDC reader is either a dedicated reader equipped with one or more sensors, usually image sensors, or capable of using a signal or image from a sensor on the second device.In other words, generally speaking, the second device may be able to obtain a mold number or a cavity (or section) number for each container it inspects.

[0136] When containers have a two-dimensional 2ID code that uniquely identifies them, the second detection device includes a RIDC reader capable of extracting the 2ID code. From this code, the mold number or cavity (or section) number used to form each container, or the production timestamp, can be determined. This number can be compared to a number contained in a message received by the second detection device to implement a change. The timestamp can be compared to a production time interval contained in a message received by the second detection device to implement a change.

[0137] Although the present invention has been described with reference to specific embodiments, it is evident that modifications and changes can be made to these examples without departing from the general scope of the invention as defined by the claims. In particular, individual features of the various embodiments illustrated / mentioned can be combined in additional embodiments. Therefore, the description and drawings should be considered in an illustrative rather than restrictive sense. [0138JII is also evident that all the characteristics described with reference to a process are transposable, alone or in combination, to a device, and conversely, all the characteristics described with reference to a device are transposable, alone or in combination, to a process.

[0139]

Claims

Demands

1. A method for inspecting glass containers circulating from a hot section of a glass container manufacturing installation (2) to a cold section of the installation, wherein, for one of said containers, the following is implemented: - an inspection (DET_D1) implemented by a first detection device (7H) configured for the inspection of circulating vessels within the hot sector and configured for the detection of the presence of at least one defect of a given type, - a transmission (TX_M) by the first detection device of a message generated based on an inspection result by the first detection device, - a reception (RX_M) by a second detection device (7C) configured for the inspection of circulating containers within the cold sector, of an alert message, the reception by the second detection device being implemented after said emission by the first detection device, - a modification (MOD_P), by the second detection device, of at least one fault detection parameter of said given type, the modification being triggered by said reception.

2. A method according to claim 1, wherein the message emitted includes a result indicating the presence of at least one defect of the given type within at least one container inspected by the first detection device.

3. A method according to claim 1 or 2, wherein the message emitted includes at least one image of a container acquired by the first detection device.

4. A method according to any one of claims 1 to 3, wherein the detection parameter is a detection sensitivity, and the modification is an increase in said detection sensitivity.

5. A method according to any one of claims 1 to 4, further comprising an inspection configured for the detection of the defect of said given type and implemented by the second detection device using said modified parameter.

6. A method according to any one of claims 1 to 5, wherein said at least one detection parameter is modified for a given period, the given period having a start time separated from said reception by a given delay.

7. A method according to any one of claims 1 to 6, comprising obtaining, by the first detection device, an identifier of a mold used or a forming cavity used for the manufacture of the container, said emitted message further comprising the identifier of the mold or cavity, - the message received by the second detection device, which also includes the mold or cavity identifier, and - the process further comprising obtaining, by the second detection device, an identifier of a mold used or a forming cavity used for the manufacture of a container controlled by the second detection device, said modification being implemented if the identifier of the mold or forming cavity corresponds to the identifier of the mold or forming cavity of the controlled container included in said received message.

8. A method according to any one of claims 1 to 7, wherein the message emitted by the first detection device is said message received by the second detection device.

9. A method according to any one of claims 1 to 8, wherein the message emitted by the first detection device is received by an installation controller, the installation controller processing and emitting the message received by the second control device.

10. A method according to any one of claims 1 to 9, wherein the first detection device is configured to acquire one or more images of a controlled container by means of one or more cameras, in the visible range, in particular by using a light source, or in the infrared.

11. A method according to any one of claims 1 to 10, wherein the second detection device is configured to acquire one or more images of a controlled container by means of one or more cameras, by illuminating the container by means of a light source, in particular in the visible range.

12. A method according to claims 10 and 11, wherein the first detection device and the second detection device are each configured to acquire an image of a container along the same observation direction at plus or minus 45°.

13. A method according to claims 10 and 11, wherein the first detection device and the second detection device are each configured to acquire an image of the same portion of the containers.

14. A method according to any one of claims 1 to 13 wherein the first detection device and the second detection device implement the same classifier or a different classifier but configured for classification according to said given type.

15. A method according to any one of claims 1 to 14, wherein the message is issued if a given number of detections implemented by the first detection device for several containers indicate the presence of at least one defect of the given type, during a given time window.

16. A system for monitoring circulating glass containers from a hot section of a glass container manufacturing facility to a cold section of the facility, the system comprising a first detection device (7H) configured for inspecting circulating containers (2) within the hot section and a second detection device (7C) configured for inspecting circulating containers within the cold section, wherein the first detection device is configured to implement: - an inspection (DET_D1), - a transmission (TX_M) by the first detection device of a message generated based on an inspection result by the first detection device, and in which the second detection device is configured to implement: - a reception (RX_M) of an alert message, the reception by the second detection device being implemented after said transmission by the first detection device, - a modification (MOD_P), of at least one fault detection parameter of said given type, the modification being triggered by said reception.

Images