Method and device for assessing container qualities based on sound recordings

Abstract

Method for assessing the container qualities of containers and in particular of glass bottles, wherein the containers are transported by means of a transport device along a predetermined transport path and at least one sound recording of noises originating from the containers (10) is recorded with a sound recording device (2), characterized in that this sound recording is evaluated and from the evaluation at least one parameter characteristic for a quality of at least one transported container is deduced and preferably at least one quality value is output which is characteristic for at least one quality property of the containers.

Description

[0001] The present invention relates to a device and a method for assessing container quality. In beverage manufacturing plants, it is known that containers such as glass bottles are often transported over long distances. In some cases, containers from different manufacturers and with varying container qualities are processed.

[0002] The containers or bottles used in the systems are often manufactured by different companies and / or differ in quality (air inclusions, inferior glass, pre-existing damage, etc.). Poorer container quality can lead to bottle breakage during the filling process or at a later stage, resulting in losses. In the context of optimizing the weight of glass bottles (lightweighting), bottle quality is becoming increasingly important.

[0003] It would therefore be desirable to have an analysis tool available that can determine the container quality and, in particular, the glass quality both in real-world conditions and in the manufacturing plant.

[0004] Currently, such line optimizations are carried out by service technicians. The technicians' experience is of paramount importance in this process. Reconfiguration is less based on facts and more on experience and intuition.

[0005] The location of a contamination within a container is often not immediately apparent. Therefore, a process of fault isolation is frequently necessary to identify the problem. Furthermore, improvements cannot be documented with numerical data but are largely based on observation and experience. This makes this task more difficult for less experienced technicians to implement.

[0006] Furthermore, widely varying container quality can lead to problems during filling, transport, or storage. Lower-quality containers are generally more susceptible to breakage and therefore require different handling than higher-quality containers. From a lightweighting perspective, it's also important to note that containers are becoming less stable overall, making container quality a crucial factor. Increased waste and breakage are thus the result of poor quality and improper handling.

[0007] Various methods for acoustically detecting defects in glass bottles are known from the prior art. For example, CN 112185419 (A) describes a crack detection method for glass bottles based on machine learning. In this method, a sound is recorded that is produced when a metal rod is struck against the glass bottle being inspected. This sound signal is then transmitted to an inspection machine, which extracts features from the recorded sound signal.

[0008] JPS 6173063 (A) also describes a defect inspection method for glass bottles. In this method, a specific part of the glass bottle is struck with a hammer and a spectral analysis of the frequencies is performed.

[0009] CN 113567552 (A) also describes a glass bottle quality detection method and a device. Here too, a glass bottle is struck with a specific force and the corresponding sound is recorded.

[0010] WO 2015 150 607 A1 describes an automatic acoustic inspection method for defective glass containers. Here, too, the individual containers are struck and the corresponding sound signals are recorded.

[0011] The methods known from the prior art are therefore suitable for testing individual containers. However, they are not suitable for use in ongoing processes or in a manufacturing plant.

[0012] The present invention is based on the objective of providing a method and a device that enable the assessment of the container qualities of individual containers or groups of containers, even in a manufacturing plant. This is achieved according to the invention by the subject matter of the independent claims. Advantageous embodiments and further developments are the subject of the dependent claims.

[0013] In a method according to the invention for assessing the container qualities of containers and in particular of glass bottles, the containers are transported by means of a transport device along a predetermined transport path and at least one sound recording of the noises and / or sounds originating from the containers is recorded with a sound recording device.

[0014] According to the invention, this sound recording is evaluated and from the evaluation at least one parameter characteristic of a quality of a transported container is deduced and preferably a quality value is output that is characteristic of at least one quality property of the containers.

[0015] In contrast to the prior art, it is therefore proposed that instead of recording the sounds produced, for example, when individual containers are struck, the noises or sounds produced by a large number of containers, and especially during the transport of a large number of containers, should be recorded. This creates a way to check the quality of individual containers or groups of containers even during ongoing operations.

[0016] It should be noted that the corresponding sound spectra, which result in particular from the transport of the containers, are more complex than the sound spectra resulting from individual containers.

[0017] The applicant was able to determine that these noises, when properly analyzed, are also suitable for inferring defects or specific parameters, such as quality characteristics. Surprisingly, it has even been shown that analyzing such noises provides a way to identify individual defective containers within a stream of containers.

[0018] The containers are preferably transported in such a way that they come into at least partial and / or temporary contact with each other. For example, the containers can be transported on a belt conveyor or a chain conveyor. Furthermore, with such conveyors, it is inevitable that the containers will collide with other objects, such as edge barriers and the like. The resulting noises can also be recorded and analyzed.

[0019] Therefore, it is preferential to record sounds that result from multiple sound sources, such as the clanging of two consecutive bottles or the striking of a bottle against a side guide.

[0020] The volume of a sound recording or recorded audio signal is also preferably taken into account. This can be relevant, for example, for determining the position of a specific container from which the noise originates. If, for instance, it is determined that a container is emitting a noise that suggests a defect in the container, the volume of this noise can be used to infer the container's location.

[0021] A change in the volume of a sound recording or a recorded audio signal is also given particular consideration. This change can also be relevant to the position of a specific container from which the sound originates. If the volume increases, this can indicate that the container in question is approaching a recording microphone.

[0022] In addition, frequency changes of a recorded signal can also be taken into account. For example, the Doppler effect can lead to different frequency shifts depending on whether the distance between the sound source and a receiving device such as a microphone decreases or increases.

[0023] Preferably, the transport equipment is a belt conveyor or a chain conveyor.

[0024] The containers are preferably transported in an empty, i.e., unfilled, state.

[0025] Container quality, assuming identical geometry, can be determined based on its sound, specifically its resonance or frequency range. Air inclusions or cracks alter the sound. Using a microphone, particularly a noise-isolated microphone, such as one mounted on an unpacking machine or in an empty bottle transport system, the audio signature of the bottle cluster can be recorded. By filtering out the background noise and reducing the signal to the pure sound of the containers (as mentioned below, this can be achieved, for example, through advanced signal processing), the audio signature can be compared to a database.

[0026] Deviations from a standard can be measured and this information can be made available to a line or individual machines. For example, if poorer container quality is detected, this equates to gentler handling, as the containers are categorized as less stable. This information can then be used to adjust the conveyor speed or the overall output of a line accordingly.

[0027] The invention ensures the quality determination of containers. This is primarily applicable to new bottles, but after developing, for example, an AI and collecting sufficiently large datasets, an application for reusable containers is also conceivable. This would allow a specific customer to use containers regardless of the manufacturer, minimizing the risk of breakage. Ideally, the quality information would be captured as soon as the containers are unpacked or removed from a pallet.

[0028] The advantage here is that there is less interference and an earlier detection time. In this way, information can be sent to the respective machines even before the containers arrive, and they can select appropriate configurations.

[0029] Preferably, the recorded noises and / or sound recordings are assigned to individual containers or specific groups of containers.

[0030] Acoustic measurements and analyses are preferred for line optimization. Areas of high noise levels can be more easily distinguished from their surroundings due to a higher sound level or a different frequency range, and can therefore be analyzed in isolation. These measurements also allow for the meaningful demonstration of improvements and provide evidence of the implemented enhancements.

[0031] Measurements can be taken either using a mobile device and / or by a service technician, but the use of a permanently installed microphone is also possible. The measurement points or locations are preferably defined and measured for each line during commissioning.

[0032] This allows for the measurement and documentation of an optimal setting. If, at a later date, a customer requests line optimization, a measurement can determine how much the line's acoustic fingerprint has changed. It can also identify at which points on the machine this fingerprint has changed.

[0033] There, targeted action can be taken, and the effectiveness of the measures can be determined through process-based adjustments to the initial state. With increasing amounts of data, it is possible and advantageous to train artificial intelligence (AI) to such an extent that it can be determined in advance by what percentage improvement is necessary to achieve the optimization effect desired by the customer.

[0034] One advantage of the approach described here is that acoustic measurements can be implemented relatively inexpensively and easily integrated into a production line. The ability to quantify modifications and improvements also saves time and money, as lengthy troubleshooting is unnecessary and problems can be resolved quickly and to the necessary extent.

[0035] In a preferred method, an audio recording is made of sounds originating from several of the transported containers. Unlike in the prior art, the sound analyzed is not that emanating from a single container, for example, one struck with a metal rod, but rather the sounds of several, and in particular several, transported containers are recorded.

[0036] This means that, on the one hand, the evaluation is technically more complex than with the sound of a single container, but on the other hand, statements can be made about the container qualities during operation.

[0037] In another preferred method, the noises emanating from the containers result at least indirectly from a transport movement of the containers.

[0038] In another preferred method, the sound is recorded during the transport movement of the containers. This allows, for example, the recording of noises that occur when containers collide or strike guide elements.

[0039] In another preferred method, a frequency analysis of the sound recording is performed. Based on this frequency analysis, various defects can be identified, such as air inclusions in the containers, cracks, or the like.

[0040] In a further advantageous method, the sound recording is compared with at least one reference value, preferably a plurality of reference values ​​and particularly preferably a reference recording and / or a reference spectrum.

[0041] This comparison with a reference allows conclusions to be drawn as to whether the transported containers, or individual containers, meet specific quality requirements, or whether they are defective, for example. It also allows conclusions to be drawn about fluctuating container quality.

[0042] It is possible that the comparison could simply reveal the presence of one or more defective containers. However, it would also be possible to identify a specific type of container based on the comparison. Artificial intelligence can be used for both purposes, as explained in more detail below. For example, the recorded sounds could be used to determine that a particular type of container is currently being transported.

[0043] A comparison of the recording made with a large number of recordings stored in a database is particularly preferred.

[0044] Preferably, at least one reference value and especially preferably a reference curve and / or a spectrum is generated using artificial intelligence (AI).

[0045] In a further preferred method, the device and, in particular, individual transport units of the device or individual treatment units of the device are controlled and / or regulated and / or maintained depending on a, in particular trainable, container noise model of machine learning, which comprises a set of, in particular trainable, parameters which are set to values ​​learned as a result of a training process, wherein the training process is based on a set of training data.

[0046] Preferably, the training data includes at least one noise level recordable for the individual transport devices under specified operating conditions, a frequency response recordable under specified operating conditions, frequency components recorded under different conditions, peak frequencies recorded under different conditions, and a characteristic parameter for the transport of the containers (such as a transport speed, a number of transported containers, a type of container, a material of the containers, or certain material properties of the containers).

[0047] Preferably, the container noise machine learning model is based on an (artificial) neural network. Preferably, the neural network is a deep neural network (DNN), in which the parameterizable processing chain has multiple processing layers, and / or a so-called convolutional neural network (CNN) and / or a recurrent neural network (RNN).

[0048] Preferably, the container noise model or the (artificial) neural network is fed the (to-be-processed) data, in particular container types, container materials, transport speeds, any material properties such as container wall thickness, any material defects in the containers, (or data derived therefrom), which are preferably specified by a user or recorded by an inspection device, as input variables. Preferably, the container noise model or the artificial neural network maps the input variables to output variables as a function of a parameterizable processing chain.

[0049] Preferably, at least one control variable for controlling and / or regulating the device and in particular the container transport, preferably a variable characteristic of a transport speed or of a post-processing operation, is selected as the output variable.

[0050] Preferably, at least one control variable for controlling and / or regulating the transport process of the containers or a further processing process such as a filling process is selected as the output variable for each transport unit or, more generally, processing unit.

[0051] Preferably, the container noise machine learning model is / was trained using predefined training data, with the parameterizable processing chain preferably being parameterized through training.

[0052] In a preferred method, training data is used in the training process of the container noise model, which includes material sizes or container parameters (preferably material sizes or sizes characteristic thereof) determined by the sound recording device and / or within the framework of the method for assessing container qualities, preferably specific for one or more container types.

[0053] It is also conceivable that training data based on various (preferably identical) devices for transporting containers could be used, ideally also taking environmental data such as temperature or air pressure into account. This offers the advantage of generating a large number of training data points in a short time and enabling the detection of any malfunctions of the transport devices and / or defective containers.

[0054] Preferably, the determined container sizes intended for use as training data are provided with (container and / or closure) type characteristics (such as a composition of the closure or material of the container) and / or classification characteristics (such as a specified target material size, such as a material composition), which indicates whether the respective container size is sufficient to fulfill the transport function, particularly depending on the container type.

[0055] Preferably, the determined container dimensions (such as container weights, material compositions, wall thicknesses, or defects) are stored and / or used as a training data set (particularly on a non-volatile storage device) together with their respective associated container type, closure type, and / or classification characteristics. Preferably, a large number of training data sets are generated in this way.

[0056] By varying the transport parameters, especially the transport speeds or accelerations, it is theoretically possible to determine a correlation between the transport parameters and the respective sound recordings.

[0057] Preferably, a neural network trained in this way (as a container noise model) is used. Training is preferably carried out using supervised learning. However, it would also be possible to train the container noise model or the artificial neural network using unsupervised learning, reinforcement learning, or stochastic learning.

[0058] Additionally or alternatively, it is possible that the training data includes at least one characteristic parameter for the current state of a transport device.

[0059] Preferably, the container noise model or the (artificial) neural network is supplied with the (to be processed) data, in particular a recorded current state of the transport device, preferably at least one recorded current state of a transport unit, and most preferably at least one recorded current state per transport unit (or characteristic or derived data thereof), which were preferably determined by the transport device. Preferably, the container noise model or the artificial neural network maps the input variables to output variables as a function of a parameterizable processing chain.

[0060] Quantities characteristic of a current state can be required manipulated variables and / or control variables (within the framework of a control system).

[0061] Preferably, at least one state variable and / or a variable characteristic of maintenance and / or a fault condition of the transport device is chosen as the output variable.

[0062] Preferably, at least one state variable and / or a variable characteristic of maintenance and / or a fault condition and / or failure probability of the transport device is chosen as the output variable for each transport unit.

[0063] Preferably, the container noise machine learning model is / was trained using predefined training data, whereby the parameterizable processing chain is parameterized through training.

[0064] In a preferred method, the training process of the container noise model uses training data that includes historically recorded data, in particular for limit values ​​and / or states, limit values ​​for repeatedly reaching limit values ​​and / or states, limit values ​​for a temporal change and / or rate of change from the values ​​characterizing the current state, and statistical limit values ​​and / or states of transport devices that are different from the transport device (and / or the individual transport units) (but preferably of identical construction). This offers the advantage that a large number of training data points can be generated in a short time and any malfunctions of the transport device and / or the other working units, such as a filling machine or a capping unit, can be detected.

[0065] Preferably, the training data includes, as a classification feature, information on whether the condition corresponds to (uninterrupted) normal operation and / or whether a malfunction is present and / or whether there is a (progressive) aging state of the transport device and / or the other processing units.

[0066] Preferably, the container noise model is suitable and intended to detect (in the training data) deviations and / or anomalies and / or patterns that indicate a defect condition of containers and / or material deviations and / or (advanced) aging or wear of containers.

[0067] Preferably, the container noise model outputs a characteristic probability value (as an output variable) depending on the input variables.

[0068] Using the container noise model, conclusions can be drawn from recorded sound samples regarding specific parameters such as the material properties of the transported containers, the occurrence of defects in specific containers or groups of containers. As mentioned above, these conclusions can be drawn during the ongoing operation of the device.

[0069] This allows for measures such as excluding certain containers or container groups, or adjusting operating parameters, particularly but not exclusively, transport speed. Subsequent or preceding processes, such as filling or sealing operations, can also be adjusted.

[0070] In another preferred method, the audio recording is filtered. In yet another preferred method, at least one noise-isolated microphone is used for recording. For example, high-pass or low-pass filters can be used to eliminate background noise. This background noise may originate, for example, from drive motors or mechanical elements and therefore does not affect the container quality or is not correlated with it.

[0071] A large number of microphones, and especially a large number of noise-isolated microphones, are preferred.

[0072] In another preferred method, characteristic data for the transport (of containers) are taken into account during the evaluation. Preferably, these data are selected from a group of data that includes the transport speed of the containers, ambient temperature, humidity, pressure, machine types, types of transport equipment, container types, noise from surrounding machinery, and the like.

[0073] In another preferred method, the treatment of the containers, in particular treatment by subsequent container handling equipment, is controlled taking into account the quality value. For example, if a certain reduced container quality is detected, the transport speed of subsequent transport equipment can be reduced.

[0074] In another preferred method, taking into account the quality value, a treatment sequence of at least one machine for treating the containers is adapted and / or modified (or parameters which are characteristic of this treatment sequence).

[0075] For example, the acoustic measurement described here can be used as a tool for line optimization and is particularly available to a service technician for this purpose. Line optimization can relate, for instance, to the materials used for transport vehicles or to transport speeds.

[0076] In another preferred method, at least one transport speed of at least one transport device is controlled, taking into account the quality value.

[0077] Preferably, the transport speeds of several transport devices are controlled, in particular several successive transport devices.

[0078] Preferably, a treatment system described here for treating the containers has at least two, preferably at least three, in particular consecutive transport devices. These can be, for example, conveyor belts.

[0079] Particularly preferred, taking into account the quality value or at least one quality value, is the control and / or modification of the relative speed between a first transport speed, at which a first transport device transports the containers, and a second transport speed, at which a second transport device transports the containers. Particularly preferred are successive transport devices, and especially directly successive transport devices.

[0080] For example, the control and, in particular, the regulation of speed differences between various transport devices, such as different conveyor belts, can be controlled and, in particular, regulated based on quality values ​​or, more generally, on acoustic measurements.

[0081] Recordings are given priority for storage. This allows a large number of recordings to be stored in a database, for example.

[0082] In a further preferred method, the quality value is selected from a group of quality values ​​which are characteristic of a wall thickness of a glass material of the containers, the presence of air inclusions in the material of the containers, the presence of cracks in the material of the containers and in particular the glass material of the containers, a quality and / or composition of the glass material of the containers, the presence of pre-damage to the containers, and the like.

[0083] Comparisons with reference curves, particularly those performed using artificial intelligence, can also be used in this area. For example, training data (as mentioned above) can be created, containing frequency components, frequency intensities, frequency bandwidths, and the like. This training data can be obtained from various types of containers, including those with pre-existing damage. This data can also be collected across multiple machines.

[0084] In another preferred method, the sound recording is made at an unpacking device for unpacking the containers and / or in an area of ​​an empty container transport.

[0085] Ideally, the sound recordings are made using empty containers. New containers are preferred, but the use of recycled or reused containers is also conceivable.

[0086] In another preferred method, the sound recordings are made using at least one mobile microphone. However, it would also be possible to make the sound recordings using at least one permanently installed microphone, and preferably using a plurality of microphones. Furthermore, it is also possible to consider physical parameters of the microphone used, such as frequency responses, or to compare them.

[0087] Preferably, the sound recordings are made in a wavelength range greater than 5Hz, preferably greater than 10Hz, preferably greater than 15Hz and particularly preferably greater than 20Hz.

[0088] Preferably, the sound recordings are made in a wavelength range that is less than 50kHz, preferably less than 40kHz, preferably less than 30kHz and particularly preferably less than 20kHz.

[0089] On the one hand, sounds, tones, and / or noises in the audible frequency range can be recorded. However, the recorded frequency range can also extend to the ultrasound and infrasound ranges, which may allow for more precise statements about the condition of the containers.

[0090] In a further advantageous method, taking into account the quality value, a state of the plant handling the containers is described and / or at least a state value is output that is characteristic of the plant handling the containers.

[0091] This approach proposes that data be determined based on the quality value or, more generally, on acoustic measurements, which allows conclusions to be drawn about the overall condition of the system handling the containers. For example, the quality value can indicate whether maintenance of the system is required or whether certain components need to be replaced.

[0092] The quality values ​​can also be used to predict future commissioning with other containers.

[0093] Preferably, a description of the condition of the machine or line handling the containers, or of the transport routes, is output or can be output based on the quality values ​​or on acoustic paths.

[0094] This can be used, for example, during the commissioning of the machine, potentially for other containers, particularly for calibration. In this way, acoustic measurements can be used to predict how the system will behave when processing different containers.

[0095] In addition, the quality values ​​can also serve as a data basis for handling complaints. Based on the quality values ​​or acoustic measurements, for example, conclusions can be drawn about the causes of malfunctions within the system, or recommendations for system maintenance can be issued.

[0096] The present invention further relates to a device for assessing the container qualities of containers, and in particular glass containers. This device comprises a transport unit that moves the containers along a predetermined transport path. In addition, the device comprises at least one sound recording device, which is suitable and intended for recording at least one sound recording of noises originating from the containers (and in particular also during transport of the containers). In particular, the recording device is suitable and intended for recording sounds that result, at least indirectly, from the transport of the containers.

[0097] According to the invention, the device has an evaluation unit which is suitable and intended to evaluate this sound recording and wherein at least one conclusion can be drawn from the evaluation regarding the quality of at least one transported container and wherein preferably at least one quality value can be output which is characteristic of at least one quality property of the container(s).

[0098] Preferably, the device has a storage device to store the recorded sounds and / or signals.

[0099] Preferably, the sound recording device is suitable and intended for continuous sound recording. For example, a sound recording can be made over a predetermined period. This predetermined period is preferably longer than 10 seconds, preferably longer than 20 seconds, preferably longer than 30 seconds, and preferably longer than 1 minute.

[0100] In a further advantageous embodiment, the sound recording device comprises at least one microphone and preferably a plurality of microphones. Recording by the combined action of several microphones is particularly preferred. For example, sounds can be recorded simultaneously at different positions and then combined accordingly. For instance, the spatial development of these sounds can also be captured. For example, the combined action of several microphones can reveal that a specific defective container has moved. For instance, it can be determined during the recording that a sound from a defective container has changed its position, thus enabling the identification of a specific defective container.

[0101] In a preferred embodiment, the device includes a reject mechanism suitable and designed to reject individual containers from the transport path. This reject mechanism is particularly preferred for rejecting individual containers in response to the output quality value.

[0102] For example, it can be determined that a particular container is defective, preferably based on the sound and / or noise. This container can then be removed from the process.

[0103] In a further advantageous embodiment, the evaluation device includes a comparison device which is suitable and intended for comparing a sound recording with a reference recording. It is particularly preferred that the quality value can be output based on this comparison.

[0104] Preferably, an acoustic camera is used, which employs beamforming algorithms to calculate sound maps. Beamforming is a method for localizing (determining the position of) sources in wave fields such as sound fields.

[0105] Beamforming measurements are typically performed in the far field. The essentially plane sound waves arrive at the individual microphones with a time delay depending on their direction of incidence. Since the microphone positions are known, the time delay can be calculated for each direction of incidence. This allows the measured signal to be corrected accordingly.

[0106] By (subsequently) summing the individual signals, the signal component in phase - coming from the defined direction of incidence - is amplified, while signal components arriving from other directions are attenuated.

[0107] Beamforming is typically used to create an acoustic map. This map provides a visual representation of sound intensity overlaid on the scene's image. Colors or brightness differences on the map indicate the strength and distribution of sound at various points within the image. This allows users to visually pinpoint the exact location of noise sources, which is particularly useful in complex environments such as industrial facilities.

[0108] Preferably, the sound recording device is (at least one) acoustic camera. This can have a plurality of microphones and preferably an image recording device, such as a camera, which is preferably arranged substantially in the center between the microphones. This will be explained in more detail with reference to the figures.

[0109] The present invention further relates to a container treatment plant for treating containers, comprising a plurality of container treatment units, each of which is suitable and intended to treat the containers in a predetermined manner, and comprising at least one transport device for transporting the containers, wherein the transport device preferably comprises a plurality of transport units which transport the containers between the treatment units.

[0110] According to the invention, the container treatment plant has at least one device for assessing the container qualities of containers of the type described above.

[0111] Advantageously, at least one of the treatment devices and preferably several container treatment devices are selected from a group of container treatment devices which includes cleaning devices for cleaning containers, filling devices for filling containers, closing devices for closing containers, labeling devices for labeling containers and packing devices for packing containers into container boxes.

[0112] Preferably, at least one of these transport devices is designed to transport the containers in a single row. In a further preferred embodiment, at least one of these transport devices is designed to transport the containers in two or more rows.

[0113] In another preferred embodiment, at least one of the transport devices has a transition area in which the number of rows in which the containers are transported is changed.

[0114] In another preferred embodiment, at least one of the transport devices has a buffer area for at least temporarily storing the containers.

[0115] In a further preferred embodiment, at least one of the transport devices has a conveyor belt or a transport chain.

[0116] In a further preferred embodiment, the container handling system has at least one optical inspection device which is suitable and intended to at least partially inspect the transported containers optically.

[0117] Further advantages and embodiments can be seen from the attached drawings: It shows: Fig. 1. A rough schematic representation of a plant for treating containers; and Fig. 2 a representation of a sound recording device for a device or method according to the invention.

[0118] Fig. Figure 1 shows an exemplary and roughly schematic representation of a container treatment plant 1. This container treatment plant 1 has a number of container treatment devices 12-18. Reference numeral 12 designates a cleaning device for cleaning the (empty) containers. This cleaning can be carried out using liquid cleaning agents.

[0119] Reference numeral 14 designates an inspection device for inspecting empty containers. Reference numeral 15 designates a filling and closing device for filling and closing containers 10.

[0120] Reference numeral 16 identifies a further inspection device for inspecting the filled containers. Reference numeral 17 identifies a labeling device for labeling the filled and sealed containers.

[0121] Reference numeral 18 designates a packing machine which packs the containers into receiving containers such as bottle crates.

[0122] A depalletizing machine 11 allows boxes filled with (especially empty) containers to be fed to the container handling system. The (empty) containers 10 are removed from the boxes and fed to the aforementioned system components. The boxes 40 themselves are fed to a box cleaning system 44 and then transported to the previously mentioned packing station 46, where they are loaded with the filled and sealed containers 10.

[0123] Reference number 50 indicates a palletizing device.

[0124] The reference symbols Tb refer to transport path sections along which the containers 10 are transported. Along these transport path sections, the containers 10 are transported between the individual treatment facilities. In some cases, the containers 10 are transported in a single row, and in others in multiple rows.

[0125] Reference numeral 2 refers to sound recording devices which are suitable and intended to record sound from the individual transported containers during their movement.

[0126] Fig. Figure 2 shows a representation of a sound recording device in the form of an acoustic camera. This device has a plurality of microphones 22 arranged on a support 26. Reference numeral 24 denotes an image recording device such as a camera. Reference numeral 28 denotes a tripod, which is preferably movable.

[0127] An acoustic camera is a device for visualizing and localizing sound sources in an environment, in this case, container transport. It typically features an array of microphones strategically positioned to capture sound waves from various directions. The data from these microphones are then preferentially synchronized and analyzed to determine the sound's origin. The acoustic camera often combines this acoustic data with an optical image captured by an integrated camera to provide a visual representation of the sound sources.

[0128] It is noted that all features described with reference to the method are also disclosed for the apparatus, which in particular means that the apparatus includes features suitable and intended for carrying out the respective methods. Furthermore, features described with reference to the apparatus are also applicable to the method(s). This means that the methods are carried out using the corresponding apparatus features.

[0129] The applicant reserves the right to claim all features disclosed in the application documents as essential to the invention, provided they are novel individually or in combination compared to the prior art. It is further noted that the individual figures also describe features which may be advantageous on their own. A person skilled in the art will immediately recognize that a particular feature described in a figure may be advantageous even without incorporating other features from that figure. Furthermore, a person skilled in the art will recognize that advantages may also arise from a combination of several features shown in individual or different figures. QUOTES INCLUDED IN THE DESCRIPTION

[0000] This list of documents cited by the applicant was automatically generated and is included solely for the reader's convenience. The list is not part of the German patent or utility model application. The DPMA accepts no liability for any errors or omissions. Cited patent literature

[0000] CN 112185419

[0007] CN 113567552

[0009] WO 2015 150 607 A1

[0010]

Claims

[1] Method for assessing the container qualities of containers and in particular of glass bottles, wherein the containers are transported by means of a transport device along a predetermined transport path and at least one sound recording of noises originating from the containers (10) is recorded with a sound recording device (2), characterized by that this sound recording is evaluated and that, from the evaluation, at least one parameter characteristic of a quality of at least one transported container is deduced, and preferably at least one quality value is output that is characteristic of at least one quality property of the containers. [2] Method according to claim 1, characterized by , that a sound recording is made of one of the sounds originating from several of the transported containers. [3] Method according to at least one of the preceding claims, characterized bythat the noises emanating from the containers result at least indirectly from a transport movement of the containers and / or that the sound recording is made during a transport movement of the containers. [4] Method according to at least one of the preceding claims, characterized by that a frequency analysis of the sound recording is performed. [5] Method according to at least one of the preceding claims, characterized by that the sound recording is compared with at least one reference value and preferably a large number of reference values. [6] Procedure according to the preceding claim, characterized by that at least one reference value is generated using artificial intelligence. [7] Method according to at least one of the preceding claims, characterized by that the audio recording is filtered and / or that at least a noise-isolated microphone is used for recording. [8] Method according to at least one of the preceding claims, characterized by , that, taking into account the quality value, the treatment of the containers is controlled, in particular by means of subsequent container treatment facilities, and / or, taking into account the quality value, a treatment process of at least one machine for the treatment of the containers is adapted. [9] Method according to at least one of the preceding claims, characterized by , that, taking into account the quality value, at least one transport speed of at least one transport device is controlled and in particular a relative speed between a first transport speed, at which a first transport device transports the containers and a second transport speed, at which a second transport device transports the containers, is controlled and / or changed. [10] Method according to at least one of the preceding claims, characterized by , that the quality value is characteristic of a group of quality values ​​which include a wall thickness of a glass material of the container, the presence of air inclusions in the material of the containers, the presence of cracks in the material and in particular the glass material of the containers, a quality and / or composition of the glass material of the containers, the presence of pre-damage. [11] Method according to at least one of the preceding claims, characterized by , that the sound recording is made at an unpacking device for unpacking the containers and / or in an area of ​​empty container transport and / or that the sound recordings are made of empty containers. [12] Method according to at least one of the preceding claims, characterized bythat the sound recordings are made using at least one mobile microphone and / or at least one permanently installed microphone. [13] Method according to at least one of the preceding claims, characterized by , that, taking into account the quality value, a state of the plant handling the containers is described and / or at least a state value is output that is characteristic of the plant handling the containers. [14] Device (1) for assessing the container qualities of containers and in particular of glass bottles, comprising a transport device which transports the containers (10) along a predetermined transport path and comprising at least one sound recording device (2) which is suitable and intended to record at least one sound recording of the noises originating from the containers (10), characterized bythat the device (1) has an evaluation unit which is suitable and intended to evaluate this sound recording and from the evaluation can be deduced at least one quality of at least one transported container and preferably can output at least one quality value which is characteristic of at least one quality property of the containers. [15] Device (1) according to the preceding claim, characterized by , that the sound recording device (2) has at least one microphone and preferably a plurality of microphones and preferably the sound recording is made possible by the interaction of several microphones and / or the sound recording device has an image recording device and / or the sound recording device is an acoustic camera. [16] Device (1) according to at least one of the preceding claims, characterized bythat the evaluation device has a comparison device which is suitable and intended to compare a sound recording with a reference recording [17] Container treatment plant for treating containers, comprising a plurality of container treatment units (12 - 18), each of which is suitable and intended to treat the containers (10) in a predetermined manner, and comprising at least one transport unit for transporting the containers, wherein the transport unit preferably comprises a plurality of transport units (Tb) which transport the containers (10) between the treatment units, characterized by , that the container treatment plant has at least one device (1) for assessing container qualities of containers (10) according to at least one of the preceding claims 12-15.

Images