Method for ascertaining properties of bulk material
By analyzing bulk material image data and user input to obtain material characteristic information, the time-consuming and expensive detection problem in the existing technology is solved, and fast and reliable quantitative distribution equipment selection and parameter setting are achieved, which improves equipment utilization and economy.
Patent Information
- Application Number
- CN202480016809.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2023-03-27
- Filing Date
- 2024-03-26
- Publication Date
- 2025-10-14
AI Technical Summary
Existing technologies require time-consuming and expensive laboratory testing when quantitatively dispensing bulk materials, and require the participation of professionals. It is difficult to quickly and reliably obtain material information to select appropriate quantitative dispensing equipment and parameters.
By acquiring sample images of bulk materials, analyzing the image data using evaluation and processing modules, and combining them with user input, specific information about the bulk material, including its physical and tactile properties, can be obtained, reducing the need for tedious inspections.
It provides a faster and more reliable way to obtain material information, improves the utilization and economy of quantitative distribution equipment, reduces dependence on professional knowledge, and enables the selection of appropriate quantitative distribution equipment and parameters.
Smart Images

Figure CN120787349A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present invention relates to a method for ascertaining at least one specific information about a bulk material. The invention further relates to a method for generating a data item having at least one information about a bulk material, a method for acquiring an information about a bulk material, and a database, a device, a computer program product, a signal sequence and a computer-implemented data structure. BACKGROUND
[0002] A dosing device for dosing a bulk material is composed of a large number of components which have to be appropriately selected and coordinated with one another depending on the bulk material to be dosed. In order to design a dosing device, it is usually necessary to detect the bulk material to be dosed in terms of its physical properties in a laboratory and to test its interaction with different embodiments of mechanical components and different settings of operating parameters in a test assembly. However, a reproducible and thus reliable result of such a detection is only possible with time-consuming and expensive measures which also require the involvement of specially trained personnel. SUMMARY
[0003] It is therefore an object of the present invention to overcome the described disadvantages of the prior art and to provide, in particular, an apparatus with which, in the case of a bulk material to be dosed, information about the bulk material can be ascertained in a simple and reliable manner so that the application of the bulk material in a dosing device can be reliably planned and prepared.
[0004] According to a first aspect of the present invention, the object is achieved by a method having the features of claim 1.
[0005] A method for ascertaining a specific information about a bulk material is proposed; the method comprises acquiring a first captured image of a sample of the bulk material, ascertaining a first bulk material information on the basis of data of the first captured image by means of a first evaluation module, and / or acquiring a second bulk material information by user input, in particular via a human-machine interface, wherein the second bulk material information is a haptic property of the bulk material. Then, the specific information about the bulk material is ascertained on the basis of data of the first bulk material information and the second bulk material information by means of a processing module.
[0006] The present invention is therefore based on the surprising insight that from a captured picture of a sample of a bulk material to be dosed, at least implicitly one or more information about the bulk material can be derived which relate, in particular, to properties of the bulk material with which the dosing behavior of the bulk material can be evaluated. By means of a supporting, additional user input in terms of further bulk material information, therefore, a cumbersome detection of the bulk material in terms of its physical properties can be dispensed with or at least reduced.
[0007] The information ascertained with the method thus provides a solid basis for further preparation of the application of the bulk material in a more fluent and, above all, more reliable manner than previously. For example, an appropriate dosing device can be selected in this context with the aid of the ascertained information. This can include the selection of components and / or operating parameters.
[0008] It has surprisingly been found here that the method delivers information about the bulk material both for previously known bulk materials. But for first-time bulk materials, such as are applied in a production environment, the ascribed information can also advantageously be ascertained.
[0009] It has been particularly surprising in this context that the information acquired with the proposed method can be superior, in part even, to data ascertained by experts in a laboratory. It is thus apparent that the information obtained from the taken pictures can describe the bulk material better, in part even, than conventional characteristic values, which were previously determined by laborious sample detection at complex test benches, such as for physical parameters of the bulk material.
[0010] The method is particularly advantageously able to achieve in this respect that existing dosing devices can be operated with different bulk materials alternately in a non-laborious manner. This is because no material-specific background knowledge is required to acquire information about the respective bulk material. This can increase the utilization of the dosing device and thus the economic efficiency and also the operating safety.
[0011] The invention is thus based on the surprising insight that from a taken picture of a sample of a bulk material to be dosed, one or more information about the bulk material can be derived at least implicitly, which relates, inter alia, to properties of the bulk material with which the dosing behavior of the bulk material can be evaluated. Thus, by means of an additional user input in support of further bulk material information, laborious detection of the bulk material in respect of its physical properties can be cancelled or at least reduced.
[0012] The information ascertained with the method thus provides a solid basis for further preparation of the application of the bulk material in a more fluent and, above all, more reliable manner than previously. For example, an appropriate dosing device can be selected in this context with the aid of the ascertained information. This can include the selection of components and / or operating parameters.
[0013] It has surprisingly been found here that the method delivers information about the bulk material both for previously known bulk materials. But for first-time bulk materials, such as are applied in a production environment, the ascribed information can also advantageously be ascertained.
[0014] In this context, it has been found particularly surprisingly that the information acquired with the proposed method can be superior to the data ascertained by experts in the laboratory, in part even. It is thus apparent that the information obtained from the taken pictures can describe the bulk material in part even better than conventional characteristic values, which were previously determined by cumbersome sample detection at complex test benches, such as for example for physical parameters of the bulk material.
[0015] The method can furthermore be particularly advantageously implemented in order to alternately operate existing dosing devices with different bulk materials in a non-burdensome manner. This is because no material-specific background knowledge is required to acquire information about the respective bulk material. This can increase the utilization of the dosing device and thus the economic efficiency as well as also the operational safety.
[0016] Primarily, even non-experts can ascertain reliable information about the bulk material to be dosed with the proposed method in order to subsequently select components of the dosing device and / or replace components of an existing dosing device and also to appropriately adjust operating parameters of the dosing device, for example, on the basis thereof.
[0017] Examples of advantageous bulk materials are rock, construction materials (in particular topsoil, sand, gravel and / or cement), raw materials (in particular ore, coal, clay and / or de-icing salt), foodstuffs (in particular grain, sugar, salt, coffee and / or flour) and / or powdery materials (in particular pigments, fillers, granules and / or pellets).
[0018] The type of bulk material can be for example one of the following bulk material types: dust, powder, flour, grain, granules, coarse powder, lumps, pellets and / or others.
[0019] The method is advantageously computer-implemented.
[0020] The method can advantageously be provided as a cloud service. It is thus not necessary to perform extensive data processing at the site of the bulk material (for which the information should be ascertained).
[0021] In one embodiment, the method comprises storing at least one specific information as a data item in a database, in particular in a material database for the bulk material.
[0022] Acquiring the taken pictures advantageously comprises acquiring data of the taken pictures, in particular image data. For example, the data are pixel values of the taken pictures. The data can be received in raw format, in a pre-processed image format and / or by a suitable light-sensitive sensor.
[0023] If in the present application information is ascertained "on the basis of" certain data, this is advantageously to be understood as meaning that the information can not be ascertained without the certain data. Here, the certain data, which can be referred to as initial data, Ausgangs-Daten, can be subjected to one or more data processing steps in order to obtain processed data. In these data processing steps, the certain data can be evaluated, further processed and / or merged with other data. By means of the certain data further information can also be obtained, in particular from other sources such as databases, and exist as intermediate results. In the data processing steps, in addition to and / or instead of the certain data and / or the processed data, further information can be subjected to further data processing steps, including processing with further certain data. More generally, the latter means here that there can be a plurality of different certain data "on the basis of" which information is ascertained, wherein these data can be processed individually and separately from one another, and two or more different such initial data, if necessary in processed form individually, can be merged in a determined processing step and optionally used as input data for a subsequent data processing.
[0024] Alternatively or additionally, it can also be provided that the method comprises ascertaining, by means of the processing module, at least one specific information, that the data of the first captured picture are taken into account in ascertaining the specific information by means of the processing module and / or that the first information about the bulk material is ascertained by means of the first evaluation module on the basis of the data of the first captured picture, and wherein the first bulk material information is taken into account in ascertaining the at least one specific information by means of the processing module and / or the first information is the at least one specific information.
[0025] Preferably, a plurality of first bulk material information is ascertained and optionally taken into account in ascertaining the specific information. In one embodiment, the first bulk material information(s) is ascertained by means of a method of image processing and evaluation from the first captured picture. These are, for example, machine learning methods. Preferably, the first bulk material information is ascertained by means of the first evaluation module and / or the first bulk material information, in addition to optionally further data, is used by the processing module for ascertaining the specific information.
[0026] The first bulk material information can be, for example, a first bulk material property, such as a material property of the bulk material.
[0027] The processing module can be implemented, for example, in software, in hardware, or a combination of both. Alternatively or additionally, the processing module can have a memory, a processor, a receiving device, such as for acquiring the taken pictures via, a transmitting device, such as for transmitting the generated signals to another entity, or any combination thereof. Alternatively or additionally, the processing module can provide and / or be available and / or have all its included content, such as in particular all resources required therefor, for example in the form of software and / or hardware resources.
[0028] The first evaluation module can be implemented, for example, in software, in hardware, or a combination of both. Alternatively or additionally, the first evaluation module can have a memory, a processor, a receiving device, such as for acquiring the taken pictures via, a transmitting device, such as for transmitting the generated signals to another entity, or any combination thereof. Alternatively or additionally, the first evaluation module can provide and / or be available and / or have all its included content, such as in particular all resources required therefor, for example in the form of software and / or hardware resources.
[0029] In an advantageous embodiment, the processing module has a first evaluation module. The first evaluation module and the processing module can preferably also be the same and thus a common module.
[0030] In an embodiment, the two modules run spatially separated from each other. This is advantageous when the two modules have different requirements for hardware and / or software resources and, as a result, the module with more computationally intensive operations can be run in an environment with sufficient hardware and / or software resources. In this case, the modules can exchange data with each other, for example via a data connection.
[0031] In an embodiment, data of at least the first taken picture and / or at least part of the output data of the first evaluation module are input data and / or at least the ascertained specific information is output data of the respective processing module. Other input and / or output data, in particular as described elsewhere in this application, are possible.
[0032] In an embodiment, data of at least the first taken picture and / or at least part of the output data of the first evaluation module are input data and / or at least the ascertained specific information is output data of the respective processing module. Other input and / or output data, in particular as described elsewhere in this application, are possible.
[0033] Preferably, the output data of the modules are acquired by the respective module at least in part by processing, i.e. such as evaluating, analyzing, and / or transforming, at least the input data of the respective module into new data.
[0034] Alternatively or additionally it can also be provided that the method comprises acquiring at least one second taken picture of a sample of the bulk material, and wherein (i) the specific information about the bulk material is ascertained by means of the processing module on the basis of the data of the first taken picture and the second taken picture, and / or (ii) the data of the second taken picture are taken into account in ascertaining the specific information by means of the processing module, wherein preferably (i) the first information about the bulk material is ascertained by means of the first evaluation module on the basis of the data of the first taken picture and the second taken picture, and / or (ii) the second information about the bulk material is ascertained by means of the second evaluation module on the basis of the data of the first taken picture and / or the second taken picture and is taken into account in ascertaining the at least one specific information, at least in part, by means of the processing module.
[0035] In addition to the data of the first taken picture, the data of the second taken picture can also be used to ascertain the specific information.
[0036] Preferably, the data of the first taken picture are not used to ascertain the second information. For example, the second information can then be ascertained on the basis of the data of the second taken picture by means of at least directing the data of the second taken picture, but not the data of the first taken picture, to the data input of the respective evaluation module. Preferably, further input data can be provided, wherein these data are not the data of the first taken picture.
[0037] It can also be advantageous to ascertain the first bulk material information and / or the second bulk material information from the data of the first taken picture and the second taken picture accordingly. To this end, the data of the first taken picture and the second taken picture can advantageously be directed to the data input of the respective evaluation module. Preferably, further input data can be provided.
[0038] The second evaluation module can be implemented in software, in hardware or a combination of both. Alternatively or additionally, the second evaluation module can have a memory, a processor, a receiving device, such as for acquiring taken pictures therevia, a transmitting device, such as for transmitting generated signals to another entity, or any combination thereof. Alternatively or additionally, the second evaluation module can provide and / or be available and / or have all that is included by it, such as in particular all resources required therefor, for example in the form of software and / or hardware resources.
[0039] In an advantageous embodiment, the processing module has the second evaluation module. The second evaluation module and the processing module can preferably also be the same and thus a common module.
[0040] In one embodiment, the two modules are operated spatially separate from one another. This is advantageous when the two modules have different requirements with respect to hardware and / or software resources and, as a result, the module with the more computationally intensive operations can be operated in an environment with sufficient hardware and / or software resources. In this case, the modules can exchange data with one another, for example, via a data connection.
[0041] It is also possible that the first evaluation module, the second evaluation module and the processing module are identical and / or the first evaluation module and the second evaluation module are identical.
[0042] In one embodiment, alternatively or in addition, the data of the second taken picture is input data of the first evaluation module. Other input and / or output data, in particular as described elsewhere in this application, are possible.
[0043] In one embodiment, alternatively or in addition, the data of the first taken picture and / or the second taken picture is input data and / or the second information is output data of the second evaluation module. Other input and / or output data, in particular as described elsewhere in this application, are possible.
[0044] In one embodiment, alternatively or in addition, the data of the second taken picture and / or the second information is input data of the processing module. Other input and / or output data, in particular as described elsewhere in this application, are possible. The processing module can advantageously ascertain the specific information, taking into account the respective data.
[0045] Alternatively or in addition, it can also be provided that the method comprises evaluating and / or analyzing the data of the first taken picture and / or the second taken picture using methods of digital image analysis, in particular at least partially in the field of machine learning, in order to ascertain the first information, the second information and / or the at least one specific information, respectively.
[0046] Advantageously, the digital image analysis is performed using methods of machine learning. For this purpose, a machine learning model can be trained on training data, which are of the type of the input data expected later, with respectively categorized, expected output data, and the trained model can be used to calculate the output data, which is the respective specific information or other information, from the respective input data.
[0047] The processing module can thus, for example, implement a corresponding method, in particular at least on the data of the first taken picture and / or the second taken picture and / or on the first information and / or the second information about the bulk material, in order to ascertain the specific information, at least partially, for example.
[0048] Thus, the first evaluation module may, for example, implement a corresponding method, in particular at least on data of the first captured picture and / or the second captured picture, in order to ascertain, for example, at least partially, the first information of the bulk material.
[0049] Thus, the second evaluation module may, for example, implement a corresponding method, in particular at least on data of the first captured picture and / or the second captured picture, in order to ascertain, for example, at least partially, the second information of the bulk material.
[0050] Alternatively or additionally, it can also be provided that ascertaining the at least one specific information comprises ascertaining a classification, in particular relating to a type of the bulk material, a preferred average grain size and / or a grain shape of the bulk material, and / or a respective identifier of the bulk material, and wherein preferably the result of the classification and / or the identifier is the specific information.
[0051] For example, the classification can relate to a bulk material type. Thus, under the aspect of the classification, the bulk material can be classified into at least one of the following type categories: dust, powder, flour, grain, granulate, kibbles, chunks, pellets and / or others. Alternatively or additionally, the classification can also relate to a grain size, in particular an average grain size, of the bulk material. Thus, the classification can advantageously provide the bulk material type and / or the grain size as a result.
[0052] The identifier may, for example, be a unique sign of the bulk material, such as an alphanumeric string. Advantageously, according to the identifier, in particular in the case of a database being taken into account, further information, for example pre-stored, about the bulk material represented by the identifier can be ascertained. Then, for example, information, such as material specifications of the bulk material, can be retrieved from the database according to the identifier of the bulk material. This will be discussed in more detail below.
[0053] Alternatively or additionally, it can also be provided that the method comprises retrieving, at least partially based on the classification and / or the identifier, information, in particular information having the first bulk material information, from a database which is related to a bulk material which is classified into one of the classification and / or the identifier, and which is taken into account, in particular ascertained, at least partially by means of the processing module, when ascertaining the at least one specific information.
[0054] The specific information can be ascertained in a more reliable and secure manner by the additional information. Thus, information about the bulk material which cannot be directly ascertained from the data of the captured pictures but can be obtained from other sources, such as a database, can also be taken into account. For this purpose, the identifier can be used as a connection link (relationship). Thereby, the specific information is obtained based on the data of the captured pictures and with further use of other resources, such as a database or the like.
[0055] The retrieved information may, for example, be or comprise information about the bulk material, such as first bulk material information of the bulk material.
[0056] Here, the information is advantageously acquired according to the identifier, i.e. such as retrieved from a database.
[0057] Thus, in one embodiment, additional information about the bulk material can be ascertained, in particular according to the identifier from the database, and which is taken into account for ascertaining the specific information, in particular at least partially by means of the processing module.
[0058] Preferably, the information relates to material properties of the bulk material. Based on the material properties, the specific information can advantageously be evaluated, for example for the suitability of the bulk material for use in a dosing device of a determined type, in a dosing device having a determined equipment, such as in terms of mechanical components and / or adjustment of operating parameters. It is thus advantageous to acquire one or more material properties based on the data of the first captured picture and by means of the identifier and which are taken into account when ascertaining the specific information.
[0059] The retrieved information can advantageously be supplied to the entity and / or provided to the user, for example on a user interface, such as a screen. Thereby, this additional information can be provided in addition to the specific information.
[0060] Alternatively or additionally, in one embodiment, at least the retrieved information is input data for the processing module. Other input and / or output data, in particular as described elsewhere in this application, are possible.
[0061] Alternatively or additionally, it can also be provided that the method comprises acquiring at least one data set, which preferably implicitly or explicitly represents at least one second information about the bulk material, wherein (i) the at least one specific information is ascertained at least partially also based on the acquired data set, in particular at least partially by means of the processing module, and / or (ii) the acquired data set is taken into account when ascertaining the at least one specific information, in particular at least partially by means of the processing module.
[0062] Thus, in addition to the data of the at least first captured picture and / or the second captured picture, the acquired data set is used for ascertaining the specific information. Here, the data of the first captured picture and / or the second captured picture and the data set may, for example, be processed together, such as as input data for a first evaluation module, in order to acquire at least one common intermediate value, which is further processed, such as as input data for the processing module, in order to ascertain the specific information. Or at least one first intermediate value is acquired from the image data and at least one second intermediate value is acquired from the data set, and the first intermediate value and the second intermediate value are further processed, such as by means of the processing module, in order to ascertain the specific information.
[0063] The data collection (Datenmenge) can for example represent and / or encode one or more information about the bulk material. Alternatively or additionally, the data collection can also be a preferably binary character sequence, wherein each position represents an information about the bulk material. An example of such a character sequence is the binary sequence 0101011101, wherein in this sequence for ten bulk material properties (e.g. "stickiness", "static", "granular",...) the presence ("1") or absence ("0") of the respective property for the bulk material is encoded.
[0064] Thus, the acquired data collection can be used in addition to the data of at least the first and / or second taken picture to find out specific information.
[0065] The information represented by the acquired data collection is advantageously a different information than the first and / or second information mentioned elsewhere.
[0066] Alternatively or additionally, in one embodiment the acquired data collection is at least an input data of the processing module. Other, especially described elsewhere in this application, input and / or output data are possible.
[0067] Alternatively or additionally it can also be provided that the data collection is acquired by a user input, especially via a human-machine interface.
[0068] Alternatively or additionally it can also be provided that (i) the first and second taking is performed from different perspectives, preferably simultaneously or in time succession, and / or (ii) the bulk material sample is in a first sample configuration at the time of the first taking and in a second sample configuration at the time of the second taking, wherein preferably the bulk material sample is transformed from the first sample configuration to the second sample configuration between the two takings.
[0069] For example, the first and second taken pictures can be taken with the same taking device, and the position and / or orientation of the taking device can be changed between the takings in order to take the two taken pictures in time succession from different perspectives.
[0070] For example, the first and second taken pictures can be taken with different taking devices, wherein the taking devices have different positions and / or orientations. Thus, the two taken pictures can be taken from different perspectives at the same time in a particularly advantageous manner.
[0071] In one embodiment the perspectives of the two taken pictures are identical.
[0072] A sample configuration is preferably understood as the manner and method of spatial arrangement of the bulk material of the sample, in particular under existing boundary conditions, for example existing environmental conditions, and / or in the case of the construction of material accumulations, for example lumps, within the arrangement.
[0073] For example, the bulk material sample can be accumulated from a supply pipe from a defined height on a flat base. The manner and method in which the bulk material sample then constructs the pile can be understood as the sample configuration. This sample configuration can differ from another sample configuration achieved when a bulk material sample from another height is piled on a flat base. Instead of the height, it is also possible to change the inclination angle of the base.
[0074] Alternatively or additionally it can also be provided that the first sample configuration and the second sample configuration are two differently constructed piles of the bulk material sample and / or that at least one parameter in the environmental conditions, such as the ambient temperature, the air humidity and / or the ambient pressure, of the bulk material sample changes between the first sample configuration and the second sample configuration.
[0075] Preferably, the bulk material is in both sample configurations in a state of equilibrium with the respective existing physical environmental conditions.
[0076] For example, the sample can be formed into a pile at two different ambient temperatures, preferably each with the same procedure. The different construction of the pile in both cases is a different sample configuration. Here, the bulk material of the sample is each at the respective ambient temperature, i.e. in this case in an equilibrium state.
[0077] Another possible environmental condition is the slope of the base on which the bulk material sample is piled (also mentioned above).
[0078] Alternatively or additionally it can also be provided that a control signal is generated which represents the ascertained specific information and is preferably supplied to an entity and / or output to a user on a human-machine interface, such as a screen, and / or in which the quantitative dispensing device is influenced, configured and / or controlled at least partially on the basis of the ascertained specific information.
[0079] The control signal advantageously enables the feeding of specific information to an entity. For example, the entity can be a high-level quantitative dispensing device monitoring system and / or a quantitative dispensing device management system. Alternatively or additionally, the control signal can be supplied to an entity to initiate monitoring of the quantitative dispensing device and / or for documentation purposes. In this way, real-time information about the respective bulk material can advantageously be continuously acquired even during the ongoing operation of the quantitative dispensing device.
[0080] The entity can be a dosing device with which the bulk material is dosed or should be dosed, or a component thereof, for example a motor or a motor control thereof. The entity can also be a device different from the dosing device, for example.
[0081] In one embodiment, the control signal can be an analog or digital signal. Alternatively or additionally, the control signal can also be an instruction within a software application.
[0082] Alternatively or additionally, it can also be provided that at least one reference object can be at least partially identified in the first recording and / or in the second recording, and wherein the reference object preferably has defined outer dimensions and / or a pattern and / or serves to determine a section of the bulk material, in particular a particle size, in the first recording and / or in the second recording.
[0083] If the outer dimensions of the reference object are known, the reference object can advantageously be used to determine the outer dimensions of the structures contained in the first recording picture and / or in the second recording picture, respectively.
[0084] The reference object can be a substrate or a portion thereof, on which the bulk material can be deposited. The reference object has a pattern, such as a checkerboard pattern, for example. This is advantageous in that a length reference is obtained by means of the reference object even when the area of the reference object is covered, for example by the bulk material.
[0085] The reference object can have a rectangular main side, for example, and can preferably be a planar and / or plate-like element. Such a reference object can be placed near the bulk material sample when the recording pictures are made or be positioned appropriately.
[0086] The reference object can also advantageously deliver information for the ML model to be discussed in detail below, i.e. which ML model is used to evaluate the first recording picture and / or the second recording picture, respectively. In this case, the ML model can learn the correlation between the reference object and the particle size of the bulk material, for example. Thus, the reference object can advantageously also be provided in the recording pictures used to train the ML model.
[0087] Alternatively or additionally, it can also be provided that the first recording is a recording image, which is retrieved from a memory, received by an optical sensor, in particular, and / or received via a data connection, preferably in the visible, infrared or ultraviolet spectral range.
[0088] The first recording picture can advantageously be taken with a camera. This can have a sensor sensitive in the visible, infrared and / or ultraviolet spectral range, for example. By selecting the sensor type, i.e. which spectral portions of the light are used to take the recording picture, different features can advantageously be identified directly on the basis of the data of the first recording picture.
[0089] For example, the temperature of the bulk material can be ascertained by means of an infrared camera, and the specific information of the bulk material can be ascertained at least partially on the basis of the temperature.
[0090] Furthermore, the first photographed picture can also be taken by means of a radar sensor and / or an X-ray device.
[0091] The first photographed picture of the bulk material to be dosed can be taken, for example, before the dosing device is filled with the bulk material to be dosed. The bulk material captured by the first photographed picture can be located, for example, in a preferably closed material package. The bulk material captured by the first photographed picture can be located, for example, in the inlet of the dosing device and / or in a storage container, from which the bulk material is advantageously removed and supplied to the dosing device. In one embodiment, the bulk material captured by the first photographed picture is located outside the dosing device, inside the dosing device and / or in front of and / or behind the discharge mechanism.
[0092] Alternatively or additionally, it can also be provided that the second photographed picture is a photographed image, which is retrieved from a memory, received by an optical sensor, in particular, and / or received via a data connection, preferably in the visible, infrared or ultraviolet spectral range.
[0093] The second photographed picture can advantageously be taken with a camera. This can have a sensor, for example, which is sensitive in the visible, infrared and / or ultraviolet spectral range. By selecting the sensor type, i.e. which spectral portions of light are taken for the photographed picture, different features can advantageously be ascertained directly on the basis of the data of the second photographed picture.
[0094] For example, the temperature of the bulk material can be ascertained by means of an infrared camera, and the specific information of the bulk material can be ascertained at least partially on the basis of the temperature.
[0095] Furthermore, the second photographed picture can also be taken by means of a radar sensor and / or an X-ray device.
[0096] The second photographed picture of the bulk material to be dosed can be taken, for example, before the dosing device is filled with the bulk material to be dosed. The bulk material captured by the second photographed picture can be located, for example, in a preferably closed material package. The bulk material captured by the second photographed picture can be located, for example, in the inlet of the dosing device and / or in a storage container, from which the bulk material is advantageously removed and supplied to the dosing device. In one embodiment, the bulk material captured by the second photographed picture is located outside the dosing device, inside the dosing device and / or in front of and / or behind the discharge mechanism.
[0097] Alternatively or additionally it can also be provided that the processing module has an ML model which has been pre-trained and / or taken into account when ascertaining the specific information.
[0098] Advantageously, the machine learning model (ML model) is trained with data of one or more taken pictures of the bulk material. In this way, a correlation between, on the one hand, the taken pictures (of the bulk material) and, on the other hand, the determined bulk material or information about it, such as an identifier and / or information about the bulk material, and / or the specific information of the bulk material, can be learned. Alternatively or additionally, further or additional information, such as the first and / or second bulk material information, can be provided as input data in addition to the taken picture data, and the training can take these into account accordingly. The learned correlation can thereby be expanded or changed.
[0099] The ML model preferably takes the input data and the output data of the processing module at least partially as input data and output data, respectively. Thus, the ML model advantageously calculates the output data from the input data. Correspondingly, the ML model is advantageously trained on input data corresponding to the input data used later on.
[0100] The ML model can also learn a correlation between, on the one hand, the identifier (of the bulk material) and, on the other hand, the specific information. Optionally, further information can also be provided as input data, and the training can take these into account accordingly. The learned correlation can thereby be expanded.
[0101] In order to learn information from images, convolutional neural networks (CNN), in particular from the field of deep learning, have proven to be advantageous and are therefore preferably used as a basis for the first ML model.
[0102] The first ML model can also be retrained with taken pictures taken again later on. The reliability and quality of the ascertained specific information can thereby be continuously improved.
[0103] Alternatively or additionally it can also be provided that the first evaluation module has an ML model which has been pre-trained and / or taken into account when ascertaining the first and / or second bulk material information and / or the identifier.
[0104] Alternatively or additionally it can also be provided that the second evaluation module has an ML model which has been pre-trained and / or taken into account when ascertaining the second bulk material information.
[0105] Advantageously, the machine learning model of the respective first and / or second evaluation module is trained with data of one or more taken pictures of bulk material known. In this way, a correlation between the taken pictures (of the bulk material) on the one hand and the determined bulk material or information related thereto, such as an identifier or information about the bulk material, on the other hand can be learned. Optionally, a correlation to different sample configurations can also be learned.
[0106] Advantageously, the ML model can be used in the processing module and in the first and second evaluation module, respectively, wherein the ML models are independent. It is advantageous in this case that the ML model of the processing module is trained at least partially with data derived from the output of the ML model of the first and / or second evaluation module and / or with data ascertained based on the output data of the ML model of the first and / or second evaluation module. In this way, a correlation between the respective output data of the first and / or second evaluation module (and, if necessary, further additional information) on the one hand and the specific information on the other hand can be learned in the case of the ML model of the processing module.
[0107] Thus, in the presence of a ML model, the specific information can advantageously be ascertained by means of the ML model of the processing module at least partially from the output data of the ML model of the first and / or second evaluation module and / or based on information ascertained from the output data of the ML model of the first and / or second evaluation module.
[0108] In order to learn information from the images, convolutional neural networks (CNN), in particular from the field of deep learning, have proven to be advantageous and are therefore preferably used as a basis for the ML model of the respective first and / or second evaluation module.
[0109] The ML model of the first and / or second evaluation module can also be retrained with taken pictures taken again later, respectively. Thereby, the reliability and quality of the ascertained specific information can be continuously improved.
[0110] The ML model of the respective evaluation module preferably takes the input data and the output data of the respective evaluation module at least partially as input data and output data, respectively. Thus, the respective ML model advantageously calculates the output data from the input data. Correspondingly, the ML model is advantageously trained on input data corresponding to input data used later.
[0111] Alternatively or additionally, it can also be provided that the first bulk material information is at least one of the following properties of the bulk material or a measure thereof: grain shape, grain size, grain size distribution, angle of repose, angle of discharge, physical properties, chemical properties, moisture content, material type, density, flow behavior, arching tendency, spouting tendency in fluidized state, proportion of lumps, electrostatic build-up, chemical instability, temperature sensitivity, floatability in air, segregation tendency, composition from components, free-flowing property, agglomeration tendency, abrasiveness, corrosiveness, mechanical sensitivity, brittleness, explosiveness, flammability, dustiness, humidity, stickiness, viscosity, hygroscopic behavior, temperature, fluidization tendency, hardening tendency, radioactivity, toxicity, thixotropic behavior, tendency to spoil, tendency to soften, static electricity, presence of oil, sheeting property and / or tackiness.
[0112] Alternatively or additionally, it can also be provided that the second bulk material information is a tactile property of the bulk material or is characterized by or is at least one of the following properties of the bulk material or a measure thereof: grain shape, grain size, grain size distribution, angle of repose, angle of discharge, physical properties, chemical properties, moisture content, material type, density, flow behavior, arching tendency, spouting tendency in fluidized state, proportion of lumps, electrostatic build-up, chemical instability, temperature sensitivity, floatability in air, segregation tendency, composition from components, free-flowing property, agglomeration tendency, abrasiveness, corrosiveness, mechanical sensitivity, brittleness, explosiveness, flammability, dustiness, humidity, stickiness, viscosity, hygroscopic behavior, temperature, fluidization tendency, hardening tendency, radioactivity, toxicity, thixotropic behavior, tendency to spoil, tendency to soften, static electricity, presence of oil, sheeting property and / or tackiness.
[0113] Further information on bulk material properties, in particular on material properties of bulk materials, can be found, for example, in the document "Common Bulk Material Properties and Overview" of "FEDERATION EUROPEENNE DE LA MANUTENTION SECTION II", FEM 2 582, Original D, Ausgabe D, November 1991, in the document "Specific Bulk Material Properties in Mechanical Conveying" of "FEDERATION EUROPEENNE DE LA MANUTENTION SECTION II", FEM 2 181, Original E, Ausgabe D, 1989, and in the document "Bulk Material Properties" of "FEDERATION EUROPEENNE DE LA MANUTENTION SECTION II", FEM 2 581, Original D, Ausgabe D, November 1991.
[0114] This object is achieved by the present invention according to a second aspect in the following manner, namely, a method for generating a data item having at least one information about bulk material, in particular information about at least one characteristic of the bulk material, the method comprising: providing a screen output to a user on a user interface; confirming the user's interaction with the screen output on the user interface in response to the provision of the screen output; and generating at least one characteristic value as a data item based on the user's confirmed interaction, the at least one characteristic value being about at least one information about the bulk material, in particular at least one characteristic of the bulk material, and being storable in a database.
[0115] The present invention is therefore based on the surprising recognition that if information about the bulk material, such as in particular bulk material properties, is requested by means of a screen interaction that is easy for the user to perform, there is no need for manual acquisition of measured values or the like and also no need for complex physical testing of the bulk material sample before the process.
[0116] This eliminates the need for complex material testing in the laboratory (eg in the form of shear tests, etc.) or at least reduces the scope of the material testing. Furthermore, bulk material information can be easily ascertained by non-specialists in this way and generated as a data item.
[0117] With such data items, a knowledge database about the material properties of bulk materials can advantageously be established, constructed and / or supplemented.Thus, in one embodiment, data items are stored in a knowledge database about bulk materials, in particular about the material properties of bulk materials.
[0118] For example, the method can be implemented using a smartphone, thereby enabling mobile generation of data items.
[0119] Preferably, the method is computer-implemented.
[0120] Alternatively or additionally, it can also be provided that the user interaction is a sliding gesture on a user interface, in particular designed as a touch-sensitive screen, and / or a selection of at least one option from a selection menu provided on a screen output.
[0121] For example, in terms of determining a bulk material property, a swipe gesture on the user interface in a first direction may indicate a first value of the bulk material property and / or a swipe gesture on the user interface in a second direction may indicate a second value of the bulk material property. For example, the bulk material property may be stickiness, and the first value may be "sticky," while the second value may be "non-sticky."
[0122] The possible directions of the sliding gesture can be displayed to the user by at least temporarily fading in the auxiliary symbol. This allows the operation performed by the user to be designed to be safer and supports the correct generation of data items.
[0123] Alternatively or additionally it can also be provided that a plurality of screen outputs is provided to the user on the user interface, in particular in temporal succession, and for each screen output the interaction of the user with the screen output on the user interface, which takes place in response to the provision of the respective screen output, is ascertained separately, and one or a plurality of characteristic values is generated as data items on the basis of the ascertained interaction of the user, which can be stored in a database, which relate to two or more than two information about the bulk material, in particular two or more than two properties of the bulk material.
[0124] Thereby a more extensive characterization of the bulk material can also be easily carried out.
[0125] Alternatively or additionally it can also be provided that the specific information is a haptic property of the bulk material or a measure thereof.
[0126] For example, at least one data item can be created for each of a plurality of bulk materials. Alternatively or additionally, it can also be possible that a plurality of users respectively generate data items with at least one information in respect of at least one information about the bulk material.
[0127] Alternatively or additionally it can also be provided that a plurality of such data items is generated, in particular by the interaction of a plurality of users, and wherein preferably the plurality of data items is transformed into a common data item, in which the information about the bulk material is built up from the information of the individual data items.
[0128] It is advantageous if a plurality of data items generated in respect of the same bulk material, in particular when they are generated by more than one user, is replaced by a common data item, in which the information from the plurality of data items can be combined, such as by taking the average, in particular with the aid of a weighting of the individual data items, which is preferably related to the user.
[0129] Alternatively or additionally it can also be provided that one data item or a plurality of data items is respectively associated with a classification of the bulk material and / or is saved in at least one database, in particular in order to build up, supplement and / or update a knowledge database about the bulk material information.
[0130] The data item can also be a common data item.
[0131] The database can advantageously be the knowledge database in question.
[0132] Alternatively or additionally it can also be provided that the data items generated are provided in the method according to the first aspect of the application and are taken there as a data collection, in particular a data collection representing information of the second bulk material.
[0133] Alternatively or additionally it can also be provided that in the method according to the first aspect of the application the ML models of the first evaluation module, the second evaluation module and / or the processing module are each trained at least partially with the generated data items and / or the data stored in the database.
[0134] This object is achieved by the present application according to a third aspect in that a database is proposed which is designed to store the data items from the method according to the second aspect of the application.
[0135] All advantages described in connection with the method according to the second aspect of the application also apply correspondingly to the database according to the third aspect of the application. In this respect reference can thus be made to the previous statements.
[0136] This object is achieved by the present application according to a fourth aspect in that a method for acquiring information about bulk material is proposed, which method comprises taking at least one first photographic picture of at least one sample of the bulk material, in particular in a first sample configuration, and / or taking with a camera, and / or taking at least one second photographic picture of a sample of the bulk material, in particular in a second sample configuration, and / or taking with a camera, and which is provided in the method according to the first aspect of the application and is acquired there as first photographic picture and / or second photographic picture.
[0137] In particular, the method in which the photographic pictures are provided can also be implemented at a remote location, such as on a cloud computer. In order to be able to use the method according to the fourth aspect of the application, it is thus necessary to provide only one camera and a corresponding data connection on the client side.
[0138] For example, the method is implemented with a smartphone. Thereby the mobile acquisition of information about the bulk material can be achieved by an operator.
[0139] Preferably, the method is computer-implemented.
[0140] All advantages described in connection with the method according to the first aspect of the application also apply correspondingly to the method according to the fourth aspect of the application. In this respect reference can thus be made to the previous statements.
[0141] Alternatively or additionally it can also be provided that the method comprises (i) acquiring the specific information identified in the method according to the first aspect of the application and / or acquiring the control signals generated in the method according to the first aspect of the application, and / or (ii) identifying the properties of the bulk material, in particular material properties, such as haptic material properties, in particular manually, and which are provided in the method according to the first aspect of the application and are acquired there as a data set representing second bulk material information.
[0142] This object is achieved by the present application according to the fifth aspect in that a device for data processing is proposed, which has an apparatus which is set up to carry out the method according to the first aspect of the application and / or according to the second aspect of the application.
[0143] The device for data processing can have a processing module, a first evaluation module and / or a second evaluation module and / or be in operative connection with these modules.
[0144] All advantages and options explained in connection with the method according to the first and / or second aspect of the application also apply accordingly to the device for data processing according to the fifth aspect of the application. In this respect, reference is made to the previous statements.
[0145] For example, the device for data processing is a cloud computer system. The device for data processing can have or be a distributed system.
[0146] This object is achieved by the present application according to the sixth aspect in that a dosing device is proposed, which has a device for data processing according to the fifth aspect of the application and / or is in operative connection with the device for data processing.
[0147] Such a dosing device can be particularly advantageously implemented, since real-time bulk material information is continuously acquired during the operation of the dosing device and, for example, on the basis of the bulk material information, an ongoing dosing process is interrupted and / or a new dosing process is not started.
[0148] All advantages and options explained in connection with the method according to the first and / or second aspect of the application and in connection with the device for data processing according to the fifth aspect of the application also apply accordingly to the dosing device according to the sixth aspect of the application. In this respect, reference is made to the previous statements.
[0149] This object is achieved by the present application according to the seventh aspect in that a device for data processing, in particular a smartphone, is proposed, which has a camera and further apparatuses which are set up to provide images taken with the camera in the method according to the first aspect of the application, so that these images are acquired there as first taken pictures and / or as second taken pictures.
[0150] The smartphone can have a screen for outputting specific information. With the device for data processing, the method according to the second aspect of the application can be carried out.
[0151] The device for data processing can optionally acquire the specific information ascertained in the method according to the first aspect of the application.
[0152] All advantages and options explained with respect to the method according to the second aspect of the application apply accordingly to the device for data processing according to the seventh aspect of the application. Thus, reference is made in this respect to the previous statements.
[0153] The object is achieved by the present application according to the eighth aspect by proposing a computer program product comprising instructions which, when the program is implemented by a device for data processing, in particular a device according to the fifth aspect of the application, initiate the device to implement the method according to the first and / or second aspect of the application.
[0154] The object is achieved by the present application according to the ninth aspect by proposing a signal sequence representing instructions which, when implemented on a device for data processing, in particular a device according to the fifth aspect of the application, initiate the device to implement the method according to the first and / or second aspect of the application.
[0155] The object is achieved by the present application according to the tenth aspect by proposing a computer-implemented data structure having in one area of the data structure commands for a data model stored in another area of the data structure for computer-aided learning, such that the data model is calculated by a device for data processing according to the commands when the data model is trained on training data stored in the other area of the data structure, the training data in particular (i) having a photographic image of a sample of bulk material with correspondingly classified bulk material information, (ii) having been at least partially acquired from the method according to the second aspect of the application, and / or (iii) having been at least partially retrieved from the database according to the third aspect of the application, such that at least one information about the bulk material can be ascertained from at least one photographic image of at least one sample of the bulk material by means of the trained data model.
[0156] The data structure advantageously enables the provision of a bundle consisting of the program product and the data model as well as the training data as a unit.
[0157] Preferably, the data structure is stored on a data carrier.
[0158] This object is achieved by the present application according to the eleventh aspect by proposing a computer-implemented data structure having in one area of the data structure a command for a computer machine learning of a trained data model stored in another area of the data structure, such that when the data model is calculated on data of one or more taken images of at least one sample of a bulk material by a device for data processing according to the command, at least one information about the bulk material is ascertained by means of the data model, in particular by the trained data model analyzing the one or more taken images and classifying the bulk material.
[0159] The data structure advantageously enables the provision of a package consisting of the program product and the data model as a unit.
[0160] Preferably, the data structure is stored on a data carrier. BRIEF DESCRIPTION OF DRAWINGS
[0161] Further features and advantages of the present application result from the following description, in which preferred embodiments of the present application are explained with reference to the schematic drawings.
[0162] Herein is shown:
[0163] Figure 1 A schematic view of a dosing device known from the prior art is shown;
[0164] Figure 2a A first taken picture of a sample of a bulk material in a first sample configuration is shown;
[0165] Figure 2b A second taken picture of a bulk material sample in a second sample configuration from Figure 2a is shown;
[0166] Figure 3 A flow chart of the method according to the first aspect of the present application in a first embodiment is shown;
[0167] Figure 4 A flow chart of the method according to the first aspect of the present application in a second embodiment is shown;
[0168] Figure 5 A flow chart of the method according to the first aspect of the present application in a third embodiment is shown;
[0169] Figure 6 A flow chart of the method according to the second aspect of the present application is shown;
[0170] Figure 7 A screen output is shown;
[0171] Figure 8a database according to the third aspect of the application is shown;
[0172] Figure 9 a flow chart of a method according to the fourth aspect of the application is shown;
[0173] Figure 10 a schematic diagram of a device for data processing according to the fifth aspect of the application is shown;
[0174] Figure 11 a schematic diagram of a dosing device according to the sixth aspect of the application is shown; and
[0175] Figure 12 a schematic diagram of a device for data processing according to the seventh aspect of the application is shown. DETAILED DESCRIPTION
[0176] Figure 1 a schematic diagram of a dosing device 1 known from the prior art is shown.
[0177] Bulk material 3 to be dosed is supplied to the dosing device 1 from a storage container 5. From the storage container 5, the bulk material enters a receiving container 9 of the dosing device 1 via a connection section 7 which can be selectively opened and closed, in which receiving container the bulk material exists in a bulk material quantity having a surface 11. By opening the connection section 7, the bulk material can be transferred from the storage container 5 into the receiving container 9. The bulk material 3 is then discharged from the dosing device 1, i.e. from the receiving container 9, in a manner known per se by means of a discharge mechanism 13 and exits therefrom via a vertical discharge section 15. The discharge mechanism 13 is coupled with a motor 17 and can be controlled in a controlled manner by a motor control of the motor 17 at an adjustable, variable rotational speed. During material discharge, the change in weight of the system of the dosing device 1, which is weighed by a weighing unit 19, is used to adjust the rotational speed of the discharge mechanism 13.
[0178] The dosing device 1 must be appropriately set for the bulk material 3 to be dosed. It can therefore be of particular interest, for example, to acquire information about the bulk material prior to a first dosing process in order to appropriately prepare the dosing device 1. For this purpose, the method according to the first aspect of the application can advantageously be employed with which specific information about the bulk material can be ascertained.
[0179] Figure 2a A first captured picture 21a of a sample 23 of bulk material, for example which is to be dosed with the dosing device 1, is shown in which the sample 23 is in a first sample configuration. In order to adjust the first sample configuration, the bulk material sample 23 from a defined height, for example 40 cm, is piled onto a base 25.
[0180] Figure 2b A second taken picture 21 b of the bulk material sample 23 is shown, in which the sample 23 is in a second sample configuration. The bulk material sample 23 is transferred from the first sample configuration into the second sample configuration by again piling the bulk material sample 23 from a defined further height, for example 80 cm, onto the base 25.
[0181] Both taken pictures are taken from the side at the same perspective. The shown sample configurations are here two differently configured piles 27a, 27b of the same bulk material sample 23. Due to the properties of the bulk material of the sample 23, in particular the material properties, two differently configured piles 27a, 27b with different heights Hl and H2 and different widths Bl and B2 are derived for the two different dumping heights, which are shown in both taken pictures by the marked double-headed arrows.
[0182] In both taken pictures 21a, 21 b a reference object 29 can also be seen. The reference object 29 is configured plate-like and is fastened on a holder 31 in such a way that a main side of the reference object 29 is taken in the taken pictures 21a, 21 b in a frontal view. The outer dimensions of the reference object 29 and a chessboard pattern on the reference object 29 are known. Thereby, the reference object 29 can be used in the respective taken picture 21a, 21 b for determining the outer dimensions of an interval, for example a structure of the bulk material.
[0183] Figure 3 A flow chart 100 of the method according to the first aspect of the application in a first embodiment is shown.
[0184] In 101 a first taken picture 21a is acquired, that is a taken picture 21a of a sample 23 of bulk material in a first sample configuration to be dosed with the dosing device 1. For this, the data of the taken picture 21a are received, for example via a data line or recalled from a memory.
[0185] In 103, based on the data of the first taken picture 21a, specific information about the bulk material of the sample 23 is ascertained. For this, the image data are processed with a processing module. The processing module is in particular a machine learning model (ML model) which has learned specific information about the bulk material in the scope of a training from a number of taken pictures of different bulk materials with classified specific information. (The taken pictures for the training show here each a sample of bulk material in a first sample configuration as described with regard to Figure 2aThe sample of the depicted sample configuration, including the correspondingly placed reference objects. The image data acquired is thus supplied to the processing module and thereby to the ML model, and the ML model is calculated on this image data. At the output of the ML model, output data is thus obtained, which represents specific information about the bulk material.
[0186] In 105, the specific information ascertained is output to the user on a user interface, such as a screen, and / or stored in a database.
[0187] The specific information about the bulk material is thus ascertained on the basis of the data of the first captured picture 21a.
[0188] Figure 4 A flowchart 200 is shown in a second embodiment of the method according to the first aspect of the application.
[0189] In 201, a first captured picture 21a is acquired, which is a captured picture 21a of a sample 23 of the bulk material to be dosed with the dosing device 1 in a first sample configuration. For this, the data of the captured picture 21a are received, for example via a data line or called up from a memory.
[0190] In 203, a second captured picture 21b is acquired, which is a captured picture 21b of a sample 23 of the bulk material to be dosed with the dosing device in a second sample configuration. For this, the data of the captured picture 21b are received, for example via a data line or called up from a memory.
[0191] In 205, specific information about the bulk material of the sample 23 is ascertained on the basis of the data of the first captured picture 21a and the second captured picture 21b. For this, the image data are processed with a processing module. The processing module is in particular a machine learning model (ML model), which has learned specific information about the bulk material in the scope of a training from a number of pairs of first captured pictures and second captured pictures of different bulk materials with classified specific information (wherein they are correspondingly in the above-described first sample configuration and second sample configuration in the respective captured pictures of the respective samples of the bulk materials). (The captured pictures for the training also show here correspondingly placed reference objects 29). The image data acquired are thus supplied to the processing module and thereby to the ML model, and the ML model is calculated on this image data. At the output of the ML model, output data is thus obtained, which represents specific information about the bulk material.
[0192] In 207, the specific information ascertained is output to the user on a user interface, such as a screen, and / or stored in a database.
[0193] From this data of the first captured picture 21a and the second captured picture 21b certain information about the bulk material is ascertained.
[0194] Figure 5 A flowchart 300 is shown in which the method according to the first aspect of the present application is implemented in a third embodiment.
[0195] In 301 a first captured picture 21a is acquired, which is also a captured picture 21a of a sample 23 of the bulk material in a first sample configuration to be dosed with the dosing device 1. For this purpose, the data of the captured picture 21a are received, for example via a data line or recalled from a memory.
[0196] In 303 the image data are processed with a first evaluation module. The first evaluation module is in particular a machine learning model (ML model) which has learned bulk material properties from a number of captured pictures of different bulk materials with classified properties in the scope of a training. (The captured pictures for the training show here samples of bulk materials in sample configurations as described with regard to Figure 2a to the first aspect of the present application, including correspondingly placed reference objects 29). Thus, the acquired image data are supplied to the first evaluation module and thereby to the ML model, and the ML model is calculated on this image data. At the output of the ML model output data are obtained which represent a first property of the bulk material, i.e. a first bulk material information. This is here a material property of the bulk material.
[0197] In 305 a data set representing a second property of the bulk material, i.e. a second bulk material information, is acquired via a user input. This is here another material property of the bulk material.
[0198] In 307 the first property and the second property are processed with a processing module to ascertain certain information about the bulk material of the sample. The processing module is in particular a machine learning model (ML model) which has learned certain information about the bulk material from a number of combinations of the first property and the second property with classified certain bulk material information in the scope of a training. Thus, the data, i.e. the first property and the second property, are supplied to the processing module and thereby to the ML model, and the ML model is calculated on this data. At the output of the ML model output data are obtained which represent the certain information about the bulk material. Other data can here also be set as input data at the ML model, although this is not necessary at present.
[0199] In 309 the ascertained certain information is output to a user on a user interface, such as a screen, and / or stored in a database.
[0200] For example, the second property is a material property of the bulk material which is not ascertained from the taken pictures or can not be ascertained at all.
[0201] Thereby, based on the data of the first taken picture 21a and the data collection, specific information about the bulk material is ascertained. Here, the data of the first taken picture 21a is not directly processed together with the data collection, but the first bulk material property ascertained based on the data of the first taken picture 21a is then processed together with the second bulk material property by supplying this data as input data to the ML model of the processing module.
[0202] In each of the above-explained embodiments of the method according to the first aspect of the application, the ascertained specific information can be, for example, a result of a classification and / or an identifier of the bulk material. For this purpose, an ML model can be provided in the processing module which at least processes the respectively mentioned input data with the ML model and delivers the classification result and / or the identifier as output data. Ascertainment of the classification and / or ascertainment of the identifier then respectively refers to ascertainment of at least a part of the specific information. The classification can advantageously be made in terms of bulk material type and / or (especially average) grain size of the bulk material. It goes without saying that in the preceding explanations the first bulk material property respectively represents the first bulk material information and the second bulk material property respectively represents the second bulk material information.
[0203] It should be noted that the ML model can also learn and exploit the association between the bulk material portion and the reference object by reference to the object 29, which can enable better specific information. However, in alternative embodiments, no reference object can be placed in the taken pictures, that is to say, those taken pictures which are acquired in the method and those taken pictures which are used to train the respective ML model.
[0204] Figure 6 A flowchart 400 of the method according to the second aspect of the application is shown. With this method, a data item with information about a bulk material can be generated. This data item can be provided, for example, as the data collection in the method according to the first aspect of the application. In this way, a knowledge database with information about bulk materials can also be built up. This information can relate, for example, to material properties of the bulk material.
[0205] In 401, a screen output on a touch-sensitive screen is provided to a user.
[0206] Figure 7An exemplary screen output 33 is shown. In a region 35 of the screen output 33, a question is displayed about information of the current bulk material ("Lorem?"). In a region 37, two possible answers to the question are displayed ("Ipsum" and "Dolor"). Furthermore, two symbols 39 are shown in the screen output 33. These symbols 39 guide the user through an interaction with the screen output 33 by the user implementing a corresponding swipe gesture on the touch-sensitive screen to move the screen output 33 optionally to the left or to the right. That is, it depends on which of the two answers the user considers to be correct. The interaction of the user with the screen output can be designed to be safer and more reliable by means of the symbols 39. Thereby, the generation of the data item can be made more reliable. As indicated by the six circles in the lower region 41 of the screen output 33, Figure 7 the screen output 33 is the third of a total of six screen outputs which are successively provided to the user for different bulk material information.
[0207] In 403 (see Figure 6 ), the interaction of the user with the screen output 33 in the form of a swipe gesture is confirmed.
[0208] In 405, a feature value is generated for the asked information of the bulk material depending on whether the interaction is confirmed as a swipe gesture to the left or to the right, for example "0" (if the swipe gesture is to the left, the question "Lorem?" is answered with "Ipsum") or "1" (if the swipe gesture is to the right, the question "Lorem?" is answered with "Dolor").
[0209] By providing six such screen outputs for six different information about the bulk material (i.e. with different questions and possible answers) and by confirming the interaction of the user with the respective screen output, in 407 a feature value is generated for the information, for example "110011", wherein each digit is the feature value generated in 405 for the respective screen output. This feature value then exists as a generated data item.
[0210] In 409, the data item is saved in a database. In this database, a knowledge database with information about the bulk material can be built up by saving the generated data items related to the classification of the bulk material, which is currently optionally also the case.
[0211] Figure 8 A database 43 according to the third aspect of the application is shown, which is designed to store data items from the method according to the second aspect of the application. For example, the database mentioned in 407 in the method according to the second aspect of the application can be the database 43, which has been described above with regard to the flowchart 400 according to Figure 6 the method.
[0212] The data within the database are also particularly advantageously suitable for use as training data for an ML model in an alternative or supplementary manner in the method according to the first aspect of the invention.
[0213] Figure 9 A flow chart 500 is shown of a method according to a fourth aspect of the invention. With this method specific information about bulk material can be obtained.
[0214] In 501, a first photograph of a sample of bulk material in a first sample configuration is taken with a camera (e.g., from Figure 2a a first picture 21a of the bulk material) and / or a second picture of a sample of the bulk material in a second sample configuration, taken with a camera (e.g., from Figure 2b A corresponding reference object 29 can also be provided in the picture.
[0215] In 503, in the method according to the first aspect of the invention, a first picture or two pictures (in particular their image data) are provided and acquired there as a first picture and / or a second picture. For example, this can be as described in reference Figures 3 to 5 The method explained.
[0216] At 505 , specific information about the bulk material is obtained.
[0217] With the aid of the specific information, the metering device 1 can be very easily adjusted appropriately for the bulk material to be metered. The method provided with the image data can advantageously be carried out on a cloud server.
[0218] Figure 10 A schematic diagram of an apparatus 45 for data processing according to a fifth aspect of the present invention is shown.
[0219] The device 45 for data processing has means which are configured to carry out the method according to the first and / or second aspect of the invention.
[0220] Figure 11 A schematic diagram of a dosing device 47 according to a sixth aspect of the invention is shown.
[0221] The dosing device 47 may be Figure 1 The apparatus 49 for data processing according to the fifth aspect of the present invention (such as referring to Figure 8 The dosing device of the device 43) described above.
[0222] Figure 12 A schematic diagram of an apparatus 51 for data processing according to a seventh aspect of the present invention is shown.
[0223] The device 51 for data processing is a smartphone having a camera 53. The device 51 provides a flexible possibility for an operator of the dosing device 1, for example for ascertaining information about the bulk material to be dosed before the bulk material 3 is loaded into the storage container 5.
[0224] To this end, the device 51 for data processing has an apparatus which is set up in such a way that the images (such as the captured pictures 21a and / or 21b) taken with the camera 53 are provided in the method according to the first aspect of the application in such a way that they are acquired there as first captured pictures and / or second captured pictures.
[0225] The features disclosed in the foregoing description, in the claims and in the accompanying drawings may, both separately and in any combination thereof, be necessary for the realization of the application in its various embodiments.
[0226] Explanation of reference signs
[0227] 1 dosing device
[0228] 3 bulk material
[0229] 5 storage container
[0230] 7 connecting section
[0231] 9 receiving container
[0232] 11 surface
[0233] 13 discharge mechanism
[0234] 15 vertical discharge section
[0235] 17 motor
[0236] 19 weighing unit
[0237] 21a, 21b captured picture
[0238] 23 bulk material sample
[0239] 25 base
[0240] 27a, 27b pile
[0241] 29 reference object
[0242] 31 holder
[0243] 33 screen output
[0244] 35 area
[0245] 37 area
[0246] 39 guide symbol
[0247] 41 region
[0248] 43 database
[0249] 45 device for data processing
[0250] 47 dosing device
[0251] 49 device for data processing
[0252] 51 device for data processing
[0253] 53 camera
[0254] 100 flow chart
[0255] 101 taking a first photographic picture of a sample of bulk material
[0256] 103 ascertaining specific information about the bulk material
[0257] 105 outputting and / or storing the ascertained specific information
[0258] 200 flow chart
[0259] 201 taking a first photographic picture of a sample of bulk material
[0260] 203 taking a second photographic picture of a sample of bulk material
[0261] 205 ascertaining specific information about the bulk material
[0262] 207 outputting and / or storing the ascertained specific information
[0263] 300 flow chart
[0264] 301 taking a first photographic picture of a sample of bulk material
[0265] 303 processing data of the first photographic picture with a first evaluation module and ascertaining a first property of the bulk material
[0266] 305 taking a set of data
[0267] 307 ascertaining specific information about the bulk material
[0268] 309 outputting and / or storing the ascertained specific information
[0269] 400 flow chart
[0270] 401 providing a screen output for a user
[0271] 403 confirming an interaction of the user with the screen output
[0272] 405 generating a characteristic value from the interaction
[0273] 407 generating a feature value as a data item from a plurality of interactions with a plurality of screen outputs
[0274] 409 storing the data item in a database
[0275] 500 flowchart
[0276] 501 taking a first and / or a second picture of a sample of a bulk material
[0277] 503 providing a picture in a method according to the first aspect of the invention
[0278] 505 obtaining information about the bulk material
[0279] Bl, B2 width
[0280] Hl, H2 height
Claims
1. A method for ascertaining specific information about bulk materials, the method comprising: - acquiring a first captured image of a sample of the bulk material, - determining first bulk material information based on the data of the first recorded image by means of a first evaluation module, and - obtaining second bulk material information by user input, in particular via a human-machine interface, wherein the second bulk material information is a tactile property of the bulk material, The specific information about the bulk material is thereby ascertained using a processing module based on data of the first bulk material information and the second bulk material information.
2. The method of claim 1 , wherein the method comprises acquiring a second captured image of the sample of the bulk material, and The specific information about the bulk material is thereby ascertained by means of the processing module based on the data of the first and second recorded images.
3. A method according to any one of the preceding claims, wherein the method comprises evaluating and / or analyzing the data of the first captured image and / or the second captured image in the field of machine learning using a digital image analysis method to correspondingly ascertain the first information, the second information and / or one of the specific information.
4. The method according to claim 1 , wherein ascertaining one of the specific information comprises ascertaining a classification relating to the type of the bulk material, the average particle size and / or particle shape of the bulk material, and / or a corresponding identifier of the bulk material, and wherein a result of the classification and / or the identifier is the specific information.
5. A method according to any one of the preceding claims, wherein a control signal representing the ascertained specific information is generated and supplied to an entity and / or output to a user on a human-machine interface, and / or wherein a quantitative dispensing device is influenced, configured and / or controlled based on the ascertained specific information.
6. A method according to any one of the preceding claims, wherein (i) The first bulk material information is one of the following properties of the bulk material or a measure thereof: Particle shape, Particle size, particle size, Angle of repose, discharge angle, Physical properties, Chemical properties, moisture content, material type, density, flow behavior, tendency to arch, tendency to burst in the fluidized state, fraction of lumps, electrostatic accumulation, chemical instability, temperature sensitivity, airborne floatability, tendency to separate, composition, free flowability, tendency to agglomerate, abrasiveness, corrosiveness, mechanical sensitivity, brittleness, explosiveness, flammability, dustiness, humidity, adhesion, viscosity, hygroscopic behavior, temperature, fluidization tendency, tendency to harden, radioactivity, toxicity, thixotropic behavior, tendency to spoil, softening tendency, static electricity, presence of grease, flaking and / or stickiness; and / or (ii) the second bulk material information is at least one of the following properties of the bulk material or a measure thereof: Particle shape, particle size, particle size, angle of repose, discharge angle, physical properties, chemical properties, moisture content, material type, density, flow behavior, arching tendency, tendency to burst in the fluidized state, lump content, electrostatic accumulation, chemical instability, temperature sensitivity, airborne floatation, tendency to separate, composition, free flowability, agglomeration tendency, abrasiveness, corrosiveness, mechanical sensitivity, brittleness, explosiveness, flammability, dustiness, humidity, adhesion, viscosity, hygroscopic behavior, temperature, fluidization tendency, hardening tendency, radioactivity, toxicity, thixotropic behavior, putrefaction tendency, softening tendency, static electricity, presence of grease, flaking and / or stickiness.
7. The method according to any one of the preceding claims, comprising providing a screen output to a user on a user interface, confirming an interaction of the user with the screen output on the user interface in response to the provision of the screen output, and generating a characteristic value as a data item based on the confirmed interaction of the user, the characteristic value relating to one of the specific information of the bulk material and capable of being stored in a database.
8. The method according to claim 7, wherein a plurality of screen outputs are provided to the user on the user interface, and for each screen output, the interaction of the user with the screen output on the user interface in response to the provision of the corresponding screen output is respectively confirmed, and one or more characteristic values are generated as data items based on the confirmed interaction of the user, and the characteristic values can be stored in a database and relate to two or more information about the bulk material, in particular, two or more characteristics of the bulk material.
9. The method according to any one of claims 7 to 8, wherein (i) a plurality of such data items are generated by interaction of a plurality of users, wherein the plurality of said data items are preferably converted into a common data item, wherein the specific information about the bulk material is constructed from the information of the individual said data items, and / or (ii) one said data item or a plurality of said data items are respectively associated with a classification of the bulk material and / or stored in a database, in particular in order to construct, supplement and / or update a knowledge database about bulk material information.
10. A device for data processing, in particular a smartphone, having a camera and other devices, which are configured to provide data in the method according to any one of claims 1 to 9, so that the image captured by the camera is obtained there as a first captured image and / or as a second captured image and thus as first bulk material information, and the tactile properties of the bulk material input via user input are obtained as second bulk material information.
