Method for operating a cleaning device and cleaning device

Abstract

The invention relates to a method for operating a cleaning device (100) with an interior (105), a door (110) closing the interior (105), and at least one basket (115) for receiving load (200), wherein the method comprises a step of reading sensor data via an interface to a detection sensor (130) of the cleaning device (100), wherein the sensor data represent mechanical vibrations, basket information, and / or a door opening angle of the door (110), and a step of processing the sensor data to provide a control signal for controlling a device function of the cleaning device (100).

Description

[0001] The invention relates to a method for operating a cleaning device and to a cleaning device.

[0002] Devices and methods for determining the loading status and loading progress in dishwashers are known, either based on image acquisition systems that capture images of the loading status in the baskets, or by vibration sensors that evaluate mechanical vibrations during the loading of the baskets in order to generate information about the loading status or the loading progress.

[0003] WO 2020 002 395 A1 therefore describes the monitoring of the loading and unloading of dishwashers using acceleration sensors.

[0004] DE 10 2018 009 311 A1 discloses a method for operating a dishwasher in which the door opening angle is read in order to determine the load quantity together with a camera and to operate the machine accordingly.

[0005] DE 10 2019 003 958 A1 discloses a method for operating a dishwasher in which basket information is read in order to determine the load quantity and to operate the machine accordingly.

[0006] US2020 / 397216 A1 discloses a method for operating a dishwasher in which basket information is read in order to determine the load quantity together with a camera and to operate the machine accordingly.

[0007] The approach presented here aims to create an improved method for operating a cleaning device as well as an improved cleaning device.

[0008] According to the invention, this problem is solved by a method for operating a cleaning device and by a cleaning device having the features of the main claims. Advantageous embodiments and further developments of the invention are described in the following dependent claims.

[0009] The approach presented here can advantageously create a way to provide users with appropriate information through improved load detection, or, for example, to start a program.

[0010] At the same time, the probability of issuing incorrect information is reduced. Furthermore, it is advantageous that limited detection areas or processes located outside the detection range can also be captured and evaluated within the washroom. In addition, the approach presented here can reduce the required storage capacity.

[0011] A method for operating a cleaning device with an interior compartment, a door closing the interior compartment, and at least one basket for receiving loads is presented. The method comprises a step of reading sensor data via an interface to a sensor on the cleaning device, where the sensor data represents mechanical vibrations, basket information, and additionally or alternatively, the door opening angle; and a step of processing the sensor data to provide a control signal for activating a function of the cleaning device.

[0012] The cleaning device can be designed, for example, as a dishwasher and thus as a household appliance, but it can also be used as a professional device. Advantageously, the method can detect the load of the cleaning device and thus recognize when at least one basket is completely full. Advantageously, the at least one basket and at least one other basket can be implemented as a single insert. The basket can, for example, be a lower basket of the cleaning device. The load can be, for example, dishes in typical household use and, in a professional context, medical instruments. The sensor data can include at least one and advantageously a plurality of pieces of information.Accordingly, the detection sensor can advantageously be a sensor already integrated into the cleaning device, which can, for example, be used for multiple functions. This can reduce additional manufacturing costs. In the processing step, the sensor data can be analyzed, compared, or otherwise processed to obtain the control signal. For example, the process steps can be repeated.

[0013] According to one embodiment, the sensor data can be processed during the processing step to obtain an analysis result that can represent a pulse pattern for a receiving basket that is at least partially withdrawn from the cleaning device or the pulse pattern for a detected loading process. The control signal can then be provided to an interface for an image acquisition device of the cleaning device, responding to the analysis result, in order to activate the image acquisition device. This means that initially only the detection sensor can be activated, and the image acquisition device can only be controlled once it has been detected that the loading process will take place. In this context, a pulse pattern is understood to be a pattern or temporal progression of a physically detectable parameter, such as at least one movement, acceleration, vibration, or sound pressure level.

[0014] The image acquisition device can advantageously be designed as a camera capable of capturing the interior of the cleaning device. By enabling delayed activation of the image acquisition device, user data protection and privacy can be advantageously ensured. In short, this means that the image acquisition device can only be switched on when the loading of the at least one collection basket is imminent. Furthermore, the method can include a step of providing an additional control signal to the interface of the image acquisition device to deactivate it again, for example, when the end of the loading process has been detected.

[0015] Furthermore, during the processing step, the sensor data can be compared with predefined reference data to obtain a reference result, and the control signal can be provided using this reference result. The reference data could, for example, be training data implemented in the cleaning device within a training environment, based, for instance, on previously acquired real-world data. Thus, a neural network within the cleaning device could, for example, independently determine the reference result using this reference data.

[0016] In the acquisition step, in addition to the sensor data, image data can be acquired via an interface to an image acquisition device of the cleaning machine. This image data can represent at least one piece of visual information about the interior of the machine. In the processing step, the image data and the sensor data can be processed to provide the control signal. Acquiring the image data advantageously allows for clear detection of the loading process. This means that, for example, both the image data and the sensor data can be evaluated to confirm or refute one of the signals, thereby preventing misinterpretations by the cleaning machine. Overall, this approach allows for a more accurate result.

[0017] According to one embodiment, the image data and sensor data can be processed during the processing step to obtain an evaluation result that can represent load information for the cleaning device. In particular, the load information can represent the load status of the at least one receiving basket and, additionally or alternatively, the loading type of the load. The load status can, for example, represent the quantity of the load in the receiving basket. Furthermore, it is advantageously possible to detect whether loads have been obscured by other items. The loading type can, for example, refer to the type of dishes, thus advantageously allowing for a classification of the load.The image capture device can therefore be advantageously activated as needed, thus advantageously reducing the required storage capacity and, secondly, protecting the privacy of a user.

[0018] According to one embodiment, the method can include a step of storing the evaluation result in order to compare the evaluation result in a repeated processing step with repeatedly acquired sensor data and, additionally or alternatively, image data. This means, for example, that a current loading status can be determined, since cleaning equipment is typically filled with material piece by piece.

[0019] In one step of the setup process, the sensitivity value of the detection sensor for the collection basket can be adjusted using the image data. This advantageously allows the sensor signals for all baskets of the cleaning appliance to be reliably detected. For example, if the detection sensor is positioned closer to a lower basket of the cleaning appliance, it can more easily detect vibrations that might indicate the lower basket is being loaded than, for example, the loading of an upper basket. Therefore, the sensitivity value for the upper basket can be advantageously adjusted. Similarly, this can be advantageously done for a cutlery drawer.

[0020] Furthermore, the method can include a step of providing the control signal to an interface with a display device of the cleaning device, wherein the control signal represents the current load level of the collection basket. The load level can, for example, represent a maximum load and thus advantageously alert a user when the collection basket and, additionally or alternatively, the cleaning device are partially or fully loaded. For example, the user can be informed when the cleaning device can be switched on.

[0021] The control unit can be configured to read input signals and use these input signals to determine and provide output signals. An input signal can, for example, be a sensor signal readable via an input interface of the control unit. An output signal can be a control signal or a data signal that can be provided at an output interface of the control unit. The control unit can be configured to determine the output signals using a processing instruction implemented in hardware or software. For example, the control unit can include a logic circuit, an integrated circuit, or a software module and may be implemented as a discrete component or comprised of a discrete component.

[0022] A computer program product or computer program according to claim 13 is also advantageous.

[0023] If the program product or program is executed on a computer or control unit, the program product or program can be used to carry out, implement and / or control the steps of the procedure according to one of the embodiments described here.

[0024] Furthermore, a cleaning device according to claim 9 is presented.

[0025] The cleaning device can be designed as a household appliance, for example. Alternatively, the approach described here can be used in conjunction with a commercial or professional device, such as a medical device, a cleaning or disinfection unit, a small sterilizer, a large-capacity disinfector, or a container washing system. Preferably, however, it can be a dishwasher. The interior of the device can also be referred to as the receiving chamber or washing chamber, which can be closed with, for example, a hinged door.

[0026] Advantageously, the appliance interior can accommodate three baskets, for example, a lower basket, an upper basket, and a cutlery tray. The control unit can be located, for example, on a control panel of the cleaning appliance.

[0027] According to one embodiment, the detection sensor can be designed as a vibration sensor, a knock sensor, an accelerometer, or a microphone.

[0028] The detection sensor can advantageously be used for a variety of device functions, thus potentially saving costs. Additionally or alternatively, the detection sensor can, for example, be integrated into the control unit.

[0029] Furthermore, the cleaning device has an image acquisition device for capturing image data, wherein the image acquisition device and, additionally or alternatively, the detection sensor can be arranged on the device door. The image acquisition device can advantageously be designed as a camera, which can, for example, be arranged on the device door. Alternatively, the image acquisition device can also be arranged inside the device.

[0030] The cleaning device can also have a display device for showing user instructions. Advantageously, the display device can be in the form of a screen.

[0031] An embodiment of the invention is shown purely schematically in the drawings and is described in more detail below. It shows Figure 1 is a schematic representation of a cleaning device according to an embodiment; Figure 2 is a schematic representation of an embodiment of a cleaning device; Figure 3 is a schematic representation of an embodiment of a receiving basket; Figure 4 is a diagram of an embodiment of a pulse pattern; Figure 5 is a schematic representation of an embodiment of a cleaning device; Figure 6 is a diagram of an embodiment of a pulse pattern; Figure 7 is a diagram of an embodiment of a pulse pattern; Figure 8 is a schematic representation of an embodiment of a cleaning device; Figure 9 is a schematic representation of an embodiment of a cleaning device; Figure 10 is a schematic representation of an embodiment of a cleaning device; Figure 11 is a flowchart of an embodiment of a method for operating a cleaning device;Figure 12 is a block diagram of a control unit according to an embodiment; Figure 13 is a flowchart for an embodiment of a method for operating a cleaning device; and Figure 14 is a flowchart for an embodiment of a method for operating a cleaning device.

[0032] Figure 1Figure 1 shows a schematic representation of a cleaning appliance 100 according to an exemplary embodiment. This is a dishwasher that can be used as a household appliance as well as a professional appliance. The cleaning appliance 100 has an interior compartment 105, an appliance door 110 closing the interior compartment 105, and at least one lower basket 115 arranged in the interior compartment 105 for receiving items. According to this exemplary embodiment, the cleaning appliance 100 additionally has a second upper basket 120 and a third basket 125 as a cutlery tray. The baskets 115, 120, and 125 are arranged one above the other in the interior compartment 105. The cleaning appliance 100 also has a sensor 130 for acquiring sensor data and a control unit 135. According to this exemplary embodiment, the sensor 130 is integrated into the control unit 135.Both are located at the device door 110.

[0033] The detection sensor 130 is configured, for example, as a vibration sensor, a knock sensor, an accelerometer, or a microphone, and is optionally usable for a variety of functions of the cleaning device 100. According to this embodiment, the cleaning device 100 also has an image capture device 140 for capturing image data. The image capture device 140 is also arranged on the device door 110. Alternatively, the image capture device 140 could be positioned differently than shown here. Furthermore, the cleaning device 100 optionally has a display device 145 for displaying a user message, according to this embodiment. The position of the display device 145 is chosen here only as an example. Alternatively, it is conceivable that the user message could be provided, for example, to a mobile device connected to the cleaning device 100 via a wireless connection.

[0034] According to this embodiment, the cleaning device 100 additionally has a lighting device 150, which is arranged in the interior 105 of the cleaning device 100. The lighting device 150 comprises, for example, a light-emitting diode (LED) shaped to illuminate the interior 105 of the device so that the image acquisition device 140 can acquire the image data with the device door 110 closed.

[0035] According to this embodiment, an optimized load detection system is described for the cleaning device 100, comprising the detection sensor 130, which can also be implemented as an acceleration sensor, and the image acquisition device 140, which is designed as a camera.

[0036] In Figure 1 A load sensor 155 is shown as optional on each of the receiving baskets 115, 120, 125, but this is for the aforementioned and in Figure 11The described method is not strictly necessary. More precisely, this embodiment depicts a cleaning appliance 100, also referred to as a standard dishwasher, which has a wash chamber 105 that is closed by the appliance door 110 and can be opened by means of a door handle 160. The cleaning appliance 100 is, or is usually, installed as a built-in appliance in a kitchen unit below a worktop. Inside the appliance door 110 is the control unit 135, also referred to as the appliance control unit, which is designed, for example, to control an actuator (not numbered, not shown) of the cleaning appliance 100. Furthermore, load sensors 155 are optionally arranged on the baskets 115, 120, 125, which, for example, are designed as acceleration sensors according to the prior art. On the wash chamber side inside the appliance door 110 is the camera 140, which captures images of the wash chamber 105.To aid image acquisition, the lighting device 150 illuminates the wash chamber 105. Inside the dishwasher 100 is the basket system, consisting of the baskets 115, 120, and 125, for holding the items to be washed. Furthermore, the detection sensor 130, located in the appliance door 110 within the appliance control unit 135, is positioned here, as previously described. This sensor optionally activates the appliance door 110 with a special actuator (not shown) and / or opens it a crack when the door is tapped.

[0037] To enable reliable load detection, a combination of image acquisition systems known from the prior art is presented here. In particular, the functionality is implemented without the additional vibration sensors 155 on the baskets 115, 120, and 125. The signals from the existing detection sensor 130 are used for evaluation, significantly increasing the accuracy of load detection for dishwashers.

[0038] This is supplemented by image-based loading information from camera 140 and the vibration signals from the detection sensor 130, which are generated, for example, when dishes are placed in at least one of the receiving baskets 115, 120, 125. Both pieces of information are evaluated to improve the reliability of the load detection to such an extent that even smaller dishes obscured by larger dishes or dishes not located within the detection range of camera 140 are still reliably detected.

[0039] Furthermore, the presented approach makes it possible to trigger the image acquisition device 140 using the detection sensor 130 whenever a loading event is imminent or has occurred. A loading event would be imminent if, preferably, the receiving basket 115 has been at least partially extended. The loading event has occurred when a dish has been placed, preferably, into the receiving basket 115. This event is also detected by the detection sensor 130.

[0040] Figure 2 Figure 1 shows a schematic representation of an exemplary embodiment of a cleaning device 100. The cleaning device 100 shown here corresponds to the one described in Figure 1. Figure 1The cleaning device 100 is described. In this embodiment, the device door 110 is shown open and the receiving basket 115 is pulled out. In this embodiment, load material 200 is placed in the receiving basket 115. Here too, the cleaning device 100 has the control unit 135 with the detection sensor 130, the image acquisition device 140, and the lighting device 150.

[0041] In other words, this embodiment illustrates the setup of the dishwasher 100 with baskets 115, 120, 125, the camera 140, and the detection sensor 130, which is located in the appliance control unit 135 or the appliance door 110. At least one lighting device 150 is arranged within the wash chamber 105. The loading process of the load 200, for example, dishes, is detected by means of the camera 140 and the knock sensor 130, and a classification into dish categories is performed. The image from the camera 140 and the signal from the knock sensor 130 are used to classify the load into dish categories, such as large plates, small plates, and cups.

[0042] Figure 3Figure 1 shows a schematic representation of an embodiment of a receiving basket 115. According to this embodiment, the receiving basket 115 is equipped with rollers 300 to move it out of and into the interior of the cleaning device 105, as symbolically represented by a double arrow 305. The receiving basket 115 moves over a threshold 310, which forms a connection point between the door 110 and the body 315 of the cleaning device.

[0043] According to this embodiment, the receiving basket 115 is fully extended so that it abuts a stop point 320. This stop point 320 is, for example, ramp-shaped and thus inclined. If the receiving basket 115 is fully extended to the stop point 320, the detection sensor 130, for example, detects a pulse pattern with three pulses. A first pulse is generated when the front rollers 300 cross the threshold 310. A second pulse is generated when the rear rollers 300 cross the threshold 310. The third pulse is generated when the front rollers 300 abut the stop point 320. Based on this, the control unit 135 determines, for example, how far the receiving basket 115 has been extended.

[0044] The detection sensor 130, which can also be designed as a vibration sensor or a microphone, detects not only load-specific information from the receiving basket 115 caused by the insertion of the load, but also mechanical vibrations (structure-borne sound) that occur when the at least one receiving basket 115, preferably the lower basket, is moved back and forth on the appliance door 110. When the receiving basket 115 is pulled out or pushed in, the rollers 300 on the wash chamber side run over two structurally defined profiled edges in the area of ​​the threshold 310, which is also referred to as the door hinge, and in the front area in the immediate vicinity of the vibration sensor 130.

[0045] Figure 4 Figure 1 shows a diagram representation of an embodiment of a pulse pattern 400, as is characteristic when the receiving basket 115 is pulled out and as it appears in Figure 3was described. The impulses, also referred to as oscillation frequencies f1, represented the movement of the rollers over the threshold of the cleaning device. The oscillation frequency f2 represents an impulse when the rollers strike the in Figure 3 described anchor point.

[0046] In other words, the pulse pattern 400 is characterized by the times Δt1, Δt2, the oscillation frequencies f1, f2, and the signal amplitudes A1, A2. If the detection sensor, which also simultaneously measures the door opening angle of the device door, detects this characteristic pulse pattern 400 or parts of it, indicating that the basket is not fully extended, loading events are very likely to follow. Furthermore, the number of detected rolling pulses provides information about how far the receiving basket has been extended.

[0047] The sensor also detects loading events, which likewise exhibit typical signal patterns. For example, heavy dishes show a higher signal amplitude, a lower frequency, and a longer signal duration than lighter dishes. Each signal peak represents a loading event; the sum of all signal peaks represents the current load status of the lower basket, starting from a value of zero after the program has finished.

[0048] Figure 5 Figure 1 shows a schematic representation of an exemplary embodiment of a cleaning device 100. The cleaning device 100 shown here corresponds, for example, to the one described in one of the Figures 1 to 2The cleaning device 100 is described. However, according to this embodiment, the upper basket, i.e., a second receiving basket 120, is shown pulled out. The device door 110 is shown open by a door opening angle α, which is detected, for example, by the detection sensor 130 in addition to a pulse pattern.

[0049] Figure 6 Figure 6 shows a diagram representing an exemplary embodiment of a pulse pattern 600. The pulse pattern 600 shown here represents an exemplary loading sequence based on the sensor data of the detection sensor for the lower basket, which is implemented as a knock sensor, and is similar, for example, to the one in Figure 600. Figure 4 described pulse pattern. Each signal peak 605 represents a loading event; the summation of all signal peaks 605 represents the current loading state of the lower basket starting from the value zero after the program has finished.

[0050] A classification of the signal peaks 605 is relative to the associated dishware and is preferably carried out via a signal amplitude A1, A2, the duration of the oscillation pulse Δt1, Δt2 and the frequency f1, f2, as in the following Fig. 7 shown.

[0051] Figure 7 This shows a diagram representing an exemplary embodiment of a pulse pattern 700. The pulse pattern shown here is similar, for example, to the one in Figure 4 and / or 6 described impulse patterns, whereby the in Figure 7 The depicted pulse pattern 700 refers to a loading process. This means that each pulse is generated by a dish placed in a receiving basket and detected by the sensor.

[0052] A heavy piece of tableware, such as a large plate or a pot, is therefore characterized by a high amplitude A1, a long signal duration Δt1, and a relatively low frequency f1, since the mechanical impulse is more energetic than, for example, that of a cup or a small plate (A2, Δt2, and f2). A repetitive loading process is further characterized by repetitions of the time interval Δt3 and its duration.

[0053] The sensitivity of the detection sensor is set via the appliance control so that, in addition to the load signals in the lower basket, the upper basket and the cutlery drawer are also detected by the knock sensor. The camera identifies which basket is extended and adjusts the knock sensor's sensitivity accordingly. The cutlery drawer, for example, is assigned the highest sensitivity.

[0054] Figure 8Figure 1 shows a schematic representation of an embodiment of a cleaning device 100. The cleaning device 100 shown here corresponds to or is similar to, for example, the one described in at least one of the Figure 1 , 2 , 5 described cleaning device with at least one receiving basket 115, as used, for example, in Figure 3 The receiving basket 115 is shown extended and loaded with goods 200. A large dish is positioned in front of a smaller dish in such a way that the image acquisition device 140 would not detect the smaller dish without the detection sensor 130.

[0055] In other words, the approach is described by way of example, or preferably, using the loading of the lower basket 115 as an example, which also represents a preferred embodiment. When loading the dishwasher 100, or the lower basket 115, it can happen that large dishes obscure smaller items 200, as shown schematically in this embodiment.

[0056] Figure 9 Figure 1 shows a schematic representation of an exemplary embodiment of a cleaning device 100. The cleaning device 100 shown here corresponds, for example, to the one described in Figure 100. Figure 8 described cleaning device. According to this embodiment, the only additional modification is that the second receiving basket 120 as well as the receiving basket 115 are filled with load material 200 and the device door 110 is closed.

[0057] The camera typically detects changes in the loading status of the baskets (115, 120, 125) after the door (110) is closed. Smaller, hidden items of food (200) added to the load are not detected by the camera (140) after the door (110) is closed. A neural network or signal processing algorithm recognizes the loading progress based on the signal from the knock sensor (130) when the door (110) is open.

[0058] Figure 10 A schematic representation of an exemplary embodiment of a cleaning device 100. The cleaning device 100 shown here is similar, for example, to the one described in at least one of the Figure 1 , 2 , 5 , 9 described cleaning device. According to this embodiment, the image acquisition device 140 has only one alternative position, namely in the interior of the device 105 facing the device door 110.

[0059] In addition to the embodiment as a knock sensor (structure-borne sound sensor), the detection sensor 130 can also be designed as a microphone (airborne sound). According to this embodiment, the camera 140 is arranged in the upper area of ​​the wash chamber 105 and detects the loading progress with the appliance door 110 open. If, for example, smaller dishes are placed behind larger dishes in the loading basket 115 during loading and are not detected by the camera 140, the detection sensor 130 is used.

[0060] Figure 11 Figure 1 shows a flowchart of an embodiment of method 1100 for operating a cleaning device. Method 1100 is used for a cleaning device such as those found, for example, in at least one of the Figure 1 , 2 , 5 , 9 , 10The procedure 1100 comprises a reading step (1105) and a processing step (1110). In reading step 1105, sensor data is read via an interface to a detection sensor of the cleaning device. This sensor data represents mechanical vibrations, basket information, and / or the door opening angle of the device door. In processing step 1110, the sensor data is processed to provide a control signal for activating a function of the cleaning device. This control signal can be provided to various device components, for example, to issue instructions to a user, control motors, or start a program. Furthermore, steps 1105 and 1110 of procedure 1100 can be repeated.

[0061] For example, in step 1110 of the processing, the sensor data is processed to obtain an analysis result that represents a pulse pattern for a collection basket that has been at least partially removed from the cleaning device or the pulse pattern for a detected loading process. The control signal is then provided to an interface for an image acquisition device, such as a camera or a so-called dishcam of the cleaning device, in response to the analysis result, in order to activate the image acquisition device. Optionally, according to this embodiment, the sensor data is compared with predefined reference data to obtain a reference result. This means that the control signal is provided using the reference result.According to this embodiment, in addition to the sensor data, image data is read in step 1105 of the acquisition process via an interface to an image acquisition device of the cleaning device. The image data represents at least one piece of image information relating to the interior of the device. In step 1110 of the processing process, the image data and the sensor data are processed to provide the control signal. For example, the image data and the sensor data are processed to obtain an evaluation result that represents load information for the cleaning device. The load information represents, for example, the load status of the at least one receiving basket and / or the loading type of the items being loaded. The load status refers, for example, to the quantity of the load and / or whether one item of dishware is obscured by another. The loading type refers, for example, to the type of dishware.This means that, for example, in step 1110 of the processing, the cargo is classified. Furthermore, the sensor data is either confirmed or refuted by the image data.

[0062] According to this embodiment, method 1100 additionally includes a step 1115 for storing the evaluation result in order to compare the evaluation result with repeatedly acquired sensor data and / or image data in a repeated processing step 1110. This means that the evaluation result is used as new reference data when method 1100 or individual steps thereof are repeated. Furthermore, method 1100 optionally includes a step 1120 for setting a sensitivity value of the detection sensor for the receiving basket using the image data, so that, for example, sensor signals from all baskets are acquired and / or each basket is assigned its own sensitivity value.According to this embodiment, method 1100 further comprises a step 1125 of providing the control signal to an interface to a display device of the cleaning device, wherein the control signal represents an achieved load level, for example, a maximum fill level of the receiving basket and / or the cleaning device. This, for example, alerts a user that the device is ready for a washing cycle.

[0063] Figure 12 Figure 1 shows a block diagram of a control unit 135 according to an exemplary embodiment. The control unit 135 corresponds, for example, to the control unit that is described in the Figure 1 , 2 , 5 , 9 , 10as described as part of the cleaning device. The control unit 135 includes a reading unit 1205, which is configured to read sensor data 1210 via an interface to a detection sensor 130 of the cleaning device. The sensor data 1210 represent mechanical vibrations, basket information, and / or a door opening angle of the device door. The control unit 135 also includes a processing unit 1215, which is configured to process the sensor data 1210 in order to provide a control signal 1220 for controlling a device function of the cleaning device, for example, to an interface to a display device 145.

[0064] Furthermore, the reading unit 1205 is designed to read image data 1225, in addition to the sensor data 1210, via an interface to an image acquisition device 140 of the cleaning device. The image data 1225 represent at least one piece of image information relating to the interior of the device. Consequently, the processing unit 1215 is also designed to process the image data 1225 and the sensor data 1210 in order to provide the control signal 1220.

[0065] In other words, the control unit 135, for example, includes a neural network or a classifier, which is located, for example, in the processing unit 1215 (alternatively, the processing unit, in particular the neural network and / or the classifier, could also be located outside the control unit and, in particular, outside the cleaning device, for example, in a cloud, whereby the sensor data and / or image data are transmitted from the control unit to the processing unit via an internet connection, and the control signal is subsequently sent back to the control unit after processing). The neural network or the classifier is fed by the image acquisition device 140 and the detection sensor 130, whereby the neural network or the classifier acquires, evaluates, and analyzes load-specific information and transmits the information about the load status.

[0066] Figure 13Figure 1300 shows a flowchart of an exemplary embodiment of a method for operating a cleaning device, such as the one described in Figure 1300. Figure 11 The described procedure is similar. However, flowchart 1300 represents an algorithm and thus a procedure of the neural network. According to this embodiment, it is a flowchart with a multitude of sub-steps, leading from closing the device door to starting a program. Flowchart 1300 includes branching sub-steps and intermediate steps, which are described in more detail below.

[0067] After the device door is closed (starting condition 1302), the first sub-step 1305 is executed. In this first sub-step 1305, it is checked whether the image capture device has detected a load. If the answer is yes, the first intermediate step 1310 follows, incrementing a counter for the load, and the final sub-step 1315 of the flowchart is executed. At this point, in the final sub-step 1315, it is determined whether the cleaning device is fully loaded or not. If yes, a program starts (1320). If not, the first sub-step 1305 is repeated.

[0068] If the image capture device does not detect any loading in the first step (1305), a second step (1325) follows, in which it is determined whether the detection sensor has detected a loading process. If not, a second intermediate step (1330) follows, and the counter for a dishware category is not incremented. If, however, the detection sensor detects a loading process, a third step (1335) follows. This determines whether the image capture device has detected a rearrangement. If so, the second intermediate step (1330) follows again. If, on the other hand, the camera does not detect a rearrangement, a third intermediate step (1340) initially assumes that the loading took place outside a detection area or that the loading location is obscured. In this case, the first intermediate step (1310) is repeated, and the counter is incremented.

[0069] The final sub-step 1315 follows, as previously described.

[0070] For this purpose, the neural network is trained in a training environment, for example in a laboratory, experimental kitchen or test households in the field, using the images from the camera and the signals from the vibration or knock sensor to recognize user-specific loading behavior, typical or household loading of the baskets and to identify the dishes.

[0071] In summary, this means, for example, the following: If the camera clearly detects an addition to the basket after the door is closed, the neural network processes this change as loading progress, classified according to the type of dishware. If the camera does not register any additions, the neural network determines whether the knock sensor detected any additions when the door was open. If no vibration signals were detected in the history, no loading progress has occurred. If the knock sensor emitted signals and the camera detects a rearrangement when the door is closed, there is also no loading progress. If the camera does not detect any rearrangement, the dish must have been placed in the basket outside the camera's field of view, or the dish must have been positioned in a location in the basket that is obscured by a larger dish.During the process, the processing algorithm and / or the neural network accesses stored image data (referencing) from past loading events as well as the signals from the knock sensor. When a washing program is started, all image data from the last loading cycle is deleted and a new loading measurement begins.

[0072] Figure 14 Figure 1400 shows a flowchart of an exemplary embodiment of a method for operating a cleaning device, such as the one described in Figure 1400. Figure 11The described procedure is similar. However, flowchart 1400 represents an algorithm and thus a procedure of the neural network, relating, for example, to whether and when the image acquisition device is activated and / or deactivated, and thus to a triggered load detection for a cleaning device. According to this embodiment, it is a flowchart with a multitude of sub-steps, leading from closing the device door to starting a program. Flowchart 1400 includes branching sub-steps and intermediate steps, which are described in more detail below.

[0073] With the triggering event (start premise 1405), the first step (1410) uses the vibration sensor to check if the device door is open. If the device door is closed, the first intermediate step (1415) determines that the image capture device, i.e., the camera, remains deactivated. Otherwise, a second step (1420) checks if a basket movement pattern has been detected. If so, a second intermediate step (1425) determines whether the collection basket has been fully extended. If a loading signal is then detected in a third step (1430), the camera is activated, the load is recorded, and / or a counter is incremented in a third intermediate step (1435) before the camera is deactivated in a repeated first intermediate step (1415). If, however, no loading signal is detected in the third step (1430), the flowchart (1400) ends or begins again.

[0074] If no basket movement pattern is detected in the second step (1420), a fifth intermediate step (1445) assumes that the basket was not fully extended, and the third step (1430) is repeated. If no loading signal is detected, the algorithm depicted in the flowchart terminates or restarts. If the loading signal is detected, the camera is switched on in a sixth intermediate step (1450), the camera focus and / or exposure time are adjusted if necessary, the load is detected, and / or the counter is incremented. The camera is then deactivated in a repeated first intermediate step (1415). In summary, this means:

[0075] Responding to the start condition 1405, the vibration sensor detects how far the basket has been extended. The camera is switched on, and the counter for the detected dish type is incremented. The camera is then switched off. The neural network also recognizes a partially extended lower basket by its movement profile, which can also be described as an impulse pattern. Unlike a fully extended lower basket, only the impulse—that is, only the contact between a basket roller and the profile edge facing the wash chamber—is detected, thus identifying a partially extended lower basket. If a loading signal is detected, the camera is switched on, the camera focus and / or exposure time are adjusted if necessary, image capture is started, and / or the counter for the corresponding dish type is incremented. The camera is then switched off.If all counters for the dishware categories reach their maximum value, the operator receives a notification to start the appliance. This notification is displayed either on the machine's screen and / or via a mobile phone. According to the invention, the camera is only activated when the basket is at least partially extended and a loading signal has been detected. This significantly reduces storage capacity and protects privacy by storing non-load-specific image data. All content described for the lower basket also applies analogously to the upper basket and the cutlery drawer.

Claims

1. Method (1100) for operating a cleaning appliance (100) comprising an appliance interior (105), an appliance door (110) which closes the appliance interior (105), and comprising at least one receiving basket (115) for receiving a load (200), the method (1100) comprising the following steps: - reading-in (1105) sensor data (1210) via an interface to a detection sensor (130) of the cleaning appliance (100), the sensor data (1210) representing mechanical vibrations, basket information and / or a door opening angle of the appliance door (110); and - processing (1110) the sensor data (1210) in order to provide a control signal (1220) for controlling an appliance function of the cleaning appliance (100), characterised in that, in the processing step (1110), the sensor data (1210) are processed in order to obtain an analysis result which represents a pulse pattern (400; 600; 700), a pulse pattern being understood to be a pattern or temporal progression of a physically detectable parameter such as at least of a movement, an acceleration, a vibration or sound pressure, the control signal (1220) being provided on an interface to an image capture device (140) of the cleaning appliance (100) in response to the analysis result in order to activate the image capture device (140).

2. Method (1100) according to claim 1, characterised in that the analysis result represents a pulse pattern (400; 600; 700) for a receiving basket (115) which has been at least partially pulled out of the cleaning appliance (100) or the pulse pattern (700) for an identified loading process.

3. Method (1100) according to any of the preceding claims, wherein, in the processing step (1110), the sensor data (1210) are compared with predetermined reference data in order to obtain a reference result, and wherein the control signal (1220) is provided using the reference result.

4. Method (1100) according to any of the preceding claims, wherein, in the reading-in step (1105), in addition to the sensor data (1210), image data (1225) are read-in to an image capture device (140) of the cleaning appliance via an interface (100), wherein the image data (1225) represent at least one piece of image information relating to the appliance interior (105), wherein, in the processing step (1110), the image data (1225) and the sensor data (1210) are processed in order to provide the control signal (1220).

5. Method (1100) according to claim 4, wherein, in the processing step (1110), the image data (1225) and the sensor data (1210) are processed in order to obtain an evaluation result that represents loading information relating to the cleaning appliance (100), in particular wherein the loading information represents a loading state of the at least one receiving basket (115) and / or a load type of the load (200).

6. Method (1100) according to claim 5, comprising a step (1115) of storing the evaluation result in order to compare the evaluation result in a repeated processing step (1110) with repeatedly read-in sensor data (1210) and / or image data (1225).

7. Method (1100) according to any of claims 4 to 6, comprising a step (1120) of adjusting a sensitivity value of the detection sensor (130) for the at least one receiving basket (115) using the image data (1225).

8. Method (1100) according to any of the preceding claims, comprising a step (1125) of providing the control signal (1220) on an interface to a display device (145) of the cleaning appliance (100), wherein the control signal (1220) represents a reached loading level of the receiving basket (115).

9. Cleaning appliance (100) comprising the following features: - an appliance interior (105); - an appliance door (110) which closes the appliance interior (105); - at least one receiving basket (115) arranged in the appliance interior (105) for receiving a load (200); - a detection sensor (130) for detecting sensor data (1210); - an image capture device (140) for capturing image data (1225); and - a control unit which is designed to perform and / or control the steps (1105, 1110, 1115, 1120, 1125) of the method (1100) according to any of the preceding claims in corresponding units (1205, 1215).

10. Cleaning appliance (100) according to claim 9, wherein the detection sensor (130) is in the form of a vibration sensor, a knock sensor, an acceleration sensor or a microphone.

11. Cleaning appliance (100) according to any of claims 9 to 10, wherein the image capture device (140) and / or the detection sensor (130) is or are arranged on the appliance door (110).

12. Cleaning appliance (1100) according to any of claims 9 to 11, comprising a display device (145) for displaying a user instruction.

13. Computer program product comprising program code for carrying out the method (1100) according to any of claims 1 to 8 when the computer program product is run on a cleaning appliance (100) according to any of claims 9 to 12.

Images