Fitting, fitting arrangement and method of triggering action of a sanitary element

Abstract

The invention relates to a fitting (1) for a sanitary device, in particular an outlet fitting. The fitting (1) comprises a control unit (3) and a sanitary element (4) which can be controlled by the control unit (3). In addition, the fitting (1) has at least one imaging sensor (5). The control unit (3) is adapted to trigger an action of the sanitary element (4) depending on an object recognition based on at least one output signal from the at least one imaging sensor (5). Furthermore, the invention relates to a fitting arrangement having a fitting according to the invention and a method for controlling a fitting according to the invention.

Description

Technical Field

[0001] The present invention relates to accessories for sanitary equipment (especially export accessories), accessory arrangements having accessories according to the invention, and methods for controlling accessories according to the invention. Background Technology

[0002] In sanitary engineering, components used to alter and control the flow of materials (e.g., fluids, i.e., liquids or gases) are called fittings. They are particularly used for plumbing and boilers. Depending on the application, so-called outlet fittings are also known in homes as faucets, shower heads, or bathtub inlet fittings. Hot and cold water outlet combination valves are often called mixing faucets. For example, depending on where they are used, fittings are more specifically called kitchen faucets or bathroom faucets. Fittings typically include an operating unit and / or a control unit, as well as sanitary elements controlled by the control unit, such as pipes with valves for regulating fluid flow.

[0003] It is well known that sanitation systems are operated using electronic control units. It is desirable for users to be able to trigger predetermined functions without physical contact. Infrared or capacitive sensors are commonly used for this purpose, reacting to user proximity. However, recently, voice-controlled faucets have also become available, allowing control via voice commands using voice assistants such as Amazon, Alexa, Apple Siri, or Google Assistant. This means users always have their hands free and do not need to touch any controls, offering practical and hygienic advantages. However, in noisy environments (such as public restrooms), voice recognition is often impaired, making voice control unreliable. Improved automatic support is also needed when using accessories in very different applications. In particular, it would be desirable if accessories could automatically recognize and trigger or execute a variety of different actions desired by the user. Summary of the Invention

[0004] Therefore, the object of the present invention is to provide an accessory for sanitary equipment that automatically assists the user in using the accessory, enabling it to be controlled, particularly without contact. Furthermore, such an accessory should be usable in many different application areas. According to the present invention, this object is achieved by the accessory defined herein.

[0005] Another object of the present invention is to disclose an accessory arrangement including sanitary equipment and accessories.

[0006] Another object of the present invention is to provide a method that automatically assists a user in using an accessory, enabling the accessory to be controlled, particularly without contact. According to the present invention, this object is achieved by the method proposed herein.

[0007] The specification provides specific embodiments and variations according to the present invention.

[0008] The accessory for a sanitary device according to the invention includes a control unit and a sanitary element controllable by the control unit. Additionally, the accessory has at least one imaging sensor. The control unit is adapted to trigger action of the sanitary element based on object recognition based on at least one output signal from the at least one imaging sensor. The imaging sensor is used to detect a scene in a spatial area of ​​the accessory (e.g., an outlet area of ​​the accessory).

[0009] For example, the accessory can be a faucet, shower or bathtub inlet accessory, but it can also be part of a soap, liquid or disinfectant applicator (e.g., a mouthpiece), or part of a hand dryer or hair dryer (e.g., a spray nozzle).

[0010] In one embodiment variation of the accessory, at least one imaging sensor is a 2D or 3D / stereo (optical) camera (for the visible light range), a thermal imaging camera (for the infrared / IR light range), an ultrasonic sensor, a radar sensor, or a laser rangefinder, particularly a lidar (light detection and ranging) or a ToF (time-of-flight) sensor or a laser scanner, or a combination of said sensor types. If only one sensor (element) is used to detect a scene in the spatial region of the accessory, it is particularly adapted to continuously (and periodically) scan (i.e., spatially scan) the spatial region, for example, by changing the orientation in which sensor signals are picked up over time, thereby capturing a two-dimensional (or three-dimensional) image (composed of multiple pixels or voxels) of the spatial region. For example, a combination of a 2D optical camera and a ToF sensor (which can be oriented in different spatial directions) can also be used for additional distance measurements. In particular, multiple ToF sensors can also be used to simultaneously determine multiple distances in different directions.

[0011] In another embodiment variation, the accessory additionally includes an illumination unit for illuminating an object, wherein the illumination unit particularly includes one or more LEDs (light-emitting diodes) or laser diodes or ultraviolet / UV or infrared / IR light sources.

[0012] In another embodiment variation, the accessory additionally includes an image processing unit to perform object recognition. The image processing unit may include, for example, a microprocessor / controller (MPU / MCU) or a signal processor (DSP), wherein these are programmable. Alternatively (or additionally), the image processing unit may include application-specific hardware such as an ASIC or an FGPA (Field Programmable Gate Array). The image processing unit may also be implemented as part of a control unit. The image processing unit and / or control unit may generally be referred to as the processor(s) of the accessory.

[0013] In another embodiment variant, the accessory additionally includes a communication unit for transmitting at least one output signal from at least one imaging sensor to a remote image processing unit (e.g., a server, particularly in the cloud) for object recognition and for receiving the results of object recognition.

[0014] In another embodiment variation of the accessory, the image processing unit is adapted to perform object recognition by means of a neural network, particularly by means of a neural network trained before the accessory (e.g., by the manufacturer) goes into production.

[0015] In another embodiment variation of the accessory, the neural network is a self-learning neural network, in which so-called "reinforcement learning" can be performed. In this case, the user provides feedback to the control unit, image processing unit, or object recognition unit to indicate whether the action triggered by the control unit is correct. The feedback from the user can be picked up by the imaging sensor; for example, if the triggered action is incorrect, the user briefly covers the imaging sensor with his hand.

[0016] In another embodiment variation, the accessory additionally includes a microphone that a user can use to provide feedback, in particular, to a control unit, wherein the control unit is adapted to detect clapping and / or whistling noises, particularly based on the output signal from the microphone.

[0017] In another embodiment variation of the accessory, the image processing unit is adapted to perform object classification as part of object recognition, in order to assign objects to specific classes from a plurality of predefined classes. In this context, each object class contains objects of a specific type.

[0018] In another embodiment variation of the accessory, object classification can be used to identify different categories of kitchen utensils, such as plates, glassware, cups, tableware, cookware, frying pans, etc., or the user's limbs, such as hands (including individual fingers), arms, feet, or legs, of accessories or cleaning tools (such as cleaning brushes, cleaning sponges, steel wool, or cleaning cloths).

[0019] In another embodiment variation of the accessory, the image processing unit is adapted to determine at least one attribute of the object, such as transparency, color, size, or degree of contamination.

[0020] In another embodiment variation of the accessory, the image processing unit is adapted to determine the position of the object (particularly relative to a reference position) and / or move it.

[0021] In another embodiment variation of the accessory, the control unit is adapted to trigger specific actions of the sanitary element based on object classification and / or object attributes and / or object position and / or object movement.

[0022] In another embodiment variation of the accessory, one or more of the following actions may be triggered: - Select a preset for the sanitary element and execute the behavior of the sanitary element as defined by that preset; - Apply a specified amount of fluid; - Apply a specified amount of fluid per unit time; - Apply the fluid at a preset maximum, minimum, or preferred temperature; - Apply a specified fluid from multiple (i.e., two or more) different fluids, specifically depending on the location of the object (such as the user's hand); - To turn on and / or off fluid delivery.

[0023] Fluids are understood to be gases (e.g., air or ozone), liquids (e.g., water or disinfectants), and / or gas / liquid mixtures (foaming water). Alternatively, the applicator may be adapted to apply fine solids (such as powders or sand) rather than fluids. However, in the context of this invention, fluids may also include viscous substances such as soaps, lotions, creams, and pastes.

[0024] According to another aspect of the invention, the fitting arrangement includes a sanitary device and a fitting according to one of the embodiments listed above, wherein (at least one) imaging sensor is arranged such that the inlet area (particularly for applying fluid) and / or outlet area (particularly for removing the applied fluid) of the fitting can be detected by (at least one) imaging sensor.

[0025] In one embodiment variation of the fixed installation arrangement, the sanitary fixtures are a sink or washbasin, a bidet, a shower, a bathtub, a soap applicator, a detergent applicator, a disinfectant applicator, a hand dryer, a hair dryer, a toilet, a shower toilet, a urinal, or a washing facility.

[0026] According to another aspect of the present invention, a method for triggering the action of a sanitary element of an accessory according to any of the above embodiments includes the following steps: - Detecting a scene at a sanitary element having at least one imaging sensor, particularly at the inlet and / or outlet area of ​​the fitting, wherein the scene is detected by means of one or more images as the output of at least one imaging sensor, wherein, particularly in the case of multiple images, these images are recorded in a time-staggered manner, for example as a movie or video, or (simultaneously) from different perspectives. - Perform object recognition to identify at least one object, such as a fixture, kitchen utensil, or cleaning tool, or a user's hand; and - The action of the sanitary element is triggered based on object recognition.

[0027] In one embodiment of the method, the scene is illuminated by means of an illumination unit, specifically one or more LEDs or laser diodes or UV or IR light sources.

[0028] In another embodiment variation of the method, the scene recorded by at least one imaging sensor is sent as the output of at least one imaging sensor to a remote image processing unit, such as a server, particularly in the cloud (i.e., an IT infrastructure provided via the Internet), for object recognition, and the result of object recognition is received by the remote image processing unit.

[0029] In another embodiment variation of the method, object recognition is performed by means of neural networks, particularly by means of neural networks trained before the parts are put into operation (e.g., by the manufacturer).

[0030] In another embodiment of the method, object classification is performed as part of the object identification process to assign objects to specific classes from a plurality of predefined classes.

[0031] In another embodiment variation of the method, different categories of kitchen utensils, such as plates, glassware, cups, tableware, cookware, frying pans, etc., or limbs, such as accessories or cleaning tools (such as cleaning brushes, cleaning sponges, steel wool, or cleaning cloths), are identified by means of object classification.

[0032] In another embodiment variation of the method, at least one attribute of the object is determined, such as transparency, color, size, or degree of contamination.

[0033] In another embodiment variation of the method, the position (particularly relative to a reference position) and / or movement of the object are determined.

[0034] In another embodiment variation of the method, specific actions of the sanitary element are triggered depending on the object classification and / or the object's attributes and / or the object's position and / or the object's movement.

[0035] In another embodiment variation of the method, the action includes at least one of the following: - Select a preset for the sanitary element and execute the behavior of the sanitary element as defined by that preset; - Apply a specified amount of fluid; - Apply a specified amount of fluid per unit time; - Apply the fluid at a preset maximum, minimum, or preferred temperature; - Apply a specific fluid from multiple (i.e., two or more) different fluids, depending on the location of the object (such as the user's hand); - To turn on and / or off fluid delivery.

[0036] In another embodiment variation of the method, the method additionally includes at least one of the following steps: - Use the (at least one) imaging sensor or a separate camera (independent of accessories) (such as a smartphone camera) to capture images (or movies or videos) of new (i.e. previously unknown) objects. - Transmit training data (such as previously recorded images (or movies or videos) of new (i.e. previously unknown) objects) to a (remote) server, optionally with an indication of the (object) category to which the training data is assigned; - Process training data to create trained neural networks, for example, in the form of TensorFlow Lite models; - Transfer the trained neural network (e.g., in the form of a TensorFlow Lite model) from the (remote) server to the accessory; - Assigning a specified object (or a gesture about that object) to a specified action of the control unit, especially by a user, such as through a web app or mobile app on a smartphone (application software for mobile devices).

[0037] It should be noted that combinations of variations of the above embodiments are possible, which in turn lead to more specific embodiments of the present invention. Attached Figure Description

[0038] The following describes in further detail, with reference to the accompanying drawings, non-limiting exemplary embodiments of the present invention, wherein: Figure 1 A schematic representation of a first exemplary embodiment of an accessory for a sanitary device according to the present invention is shown; Figure 2 a) shows a representation in a side view of a second exemplary embodiment of an accessory (such as a washbasin or sink) for sanitary ware according to the present invention; b) A representation in a bottom view of a second exemplary embodiment is shown; Figure 3 A schematic representation of a third exemplary embodiment of an accessory according to the invention for a washbasin in a sanitary facility, such as a bathroom or public restroom, is shown; Figure 4 A schematic representation of a fourth exemplary embodiment of an accessory according to the invention for a washbasin in a sanitary facility, such as a bathroom or public restroom; Figure 5 A representation illustrating a second exemplary embodiment for explaining use cases for handwashing is shown; Figure 6This illustrates a second exemplary embodiment for explaining a use case of washing dirty dishes; Figure 7 This illustrates a second exemplary embodiment for explaining a use case of filling a teacup with hot water; Figure 8 This illustrates a second exemplary embodiment of a use case for filling a measuring cup with cold water; Figure 9 A schematic representation of a fifth exemplary embodiment of an accessory according to the invention for a shower in a sanitary facility, such as a bathroom or public shower room; and Figure 10 A schematic representation of a sixth exemplary embodiment of an accessory according to the present invention for a sanitary device having two outlets for two different liquids is shown.

[0039] In the accompanying drawings, the same reference numerals refer to the same elements. Detailed Implementation

[0040] To explain the basic principles of this invention, Figure 1 A schematic diagram illustrates a first simple exemplary embodiment of an accessory 1 for sanitary appliances according to the present invention, such as a washbasin for a bathroom or a sink for a kitchen. In this case, accessory 1 is, for example, a spout accessory (also called a faucet). Accessory 1 includes a sanitary element 4 in the form of a pipe with an outlet 17 for a liquid 11 (e.g., water). The flow rate of the liquid 11 flowing from the outlet 17 is controlled by a valve ( Figure 1 (Not shown in the diagram) The valve is adjusted. This valve is then adjusted by the control unit 3. Typically, the valve is controlled by a manual control element that must be operated by the user. Alternatively, the valve is automatically triggered, for example by means of a proximity sensor, such as an infrared / IR sensor. According to the invention, an imaging sensor 5, such as a camera, is now used to trigger the valve via the control unit 3. For this purpose, the output signal of the imaging sensor 5 is analyzed, for example, one or more 2D images or movies or videos (i.e., a time series of images), and object recognition is performed, for example, to detect the presence of a hand under outlet 17. Depending on the result of the object recognition, the valve is opened via the control unit 3 to allow the user to wash their hands. Once the user removes their hand from the inlet area 13 detected and monitored by the imaging sensor 5, the object recognition system detects this and again closes the valve via the control unit 3, so that no more liquid 11 escapes from outlet 17.

[0041] The advantage of object recognition is that different operations can now be performed depending on the detected object and its movement. For example, the flow rate of water can be controlled by the imaging sensor 5 based on the distance of the object 7 (e.g., a hand); for example, the farther the hand is from the camera, the more water is applied per unit time. Additionally, for example, in the case of a mixing faucet, i.e., a combined hot and cold water outlet valve, the water temperature can be adjusted based on the movement of the object 7; for example, moving / displaced the hand to the right results in colder water, while moving / displaced the hand to the left results in warmer water.

[0042] Figure 2 A side view and a bottom view of a second exemplary embodiment of an accessory 1 for sanitary ware (such as a washbasin or sink) according to the present invention are shown. Figure 1 In this example, the imaging sensor 5 (e.g., a camera) is also positioned near the outlet 17 of the sanitary element 4 (in... Figure 2 Above Exit 17, in Figure 1 The inlet area 13 is located on the outlet pipe (below the outlet 17) so that the inlet area 13 can be completely covered by the imaging sensor 5 (e.g., a camera).

[0043] Figure 3 A schematic representation of a third exemplary embodiment of an accessory 1 for a sanitary appliance 2 (such as a washbasin in a bathroom) according to the present invention is shown. Signals (e.g., images, movies, or videos) recorded by an imaging sensor 5 (e.g., a camera) are transmitted to an image processing unit 8 for object recognition. The image processing unit 8 may be located directly at the imaging sensor 5 (e.g., in a common electronic module), or it may be a separate unit located, for example, away from the outlet pipe. Signal / data transmission between the imaging sensor 5 (e.g., the camera) and the image processing unit 8 may be via cable or wirelessly (in... Figure 3 (Illustrated as "radio flash"). For example, the image processing unit 8 may be located at the control unit 3 (e.g., in a common electronic module, or both may be implemented by a common processor). Alternatively, object recognition may be performed, for example, in a remote server such as cloud server 10. The connection there may be partially wireless (e.g., via IR, Bluetooth, WLAN, mobile radio) or wired (e.g., via Controller Area Network (CAN) bus, Ethernet), particularly via the Internet. To enable wireless connectivity to the imaging sensor 5 (e.g., a camera), accessory 1 additionally includes a wireless communication unit 9 (transmitter / receiver, transceiver).

[0044] Figure 4 A schematic representation of a fourth exemplary embodiment of an accessory 1 for a sanitary appliance 2 (such as a washbasin in a bathroom) according to the present invention is shown. Figure 3Compared to the third exemplary embodiment, here the imaging sensor 5 (e.g., a camera) is not arranged on the outlet pipe, but is mounted on a wall and facing the inlet area 13 (and outlet area 14). Additionally, accessory 1 includes an illumination unit 6 arranged on the outlet pipe. This is used to illuminate the inlet area 13 (and outlet area 14) so ​​that the imaging sensor 5 can record an image with the highest possible contrast, which aids in object recognition. The light emitted by the illumination unit 6 is adapted to the detection area of ​​the imaging sensor 5. However, the illumination unit 6 can also be used as user work lighting, and for example, as a display / indicator—for example, red light for hot water and blue light for cold water, and a flashing light for malfunctions.

[0045] For example, as part of object recognition, object classification is performed to assign objects to specific classes from a variety of predefined categories. Each object class includes specific types of objects, such as kitchen utensils (e.g., plates, glassware, cups, cutlery, cookware, pans, etc.) or accessories or cleaning tools (e.g., cleaning brushes, sponges, steel wool, or cleaning cloths) and user limbs (e.g., hands, fingers, arms, feet, or legs). Depending on the detected object, various actions of the hygienic element can then be triggered; that is, a specific action is assigned to each object (or each type of object).

[0046] Specifically, object recognition can be performed using neural networks or networks. The neural network is trained before the accessory is put into operation (e.g., by the accessory manufacturer). This means that the settings of the neural network are determined with the help of training data. The training data used to train the neural network includes the output signal / data of the imaging sensor 5 and the assignment of specified (known) objects (referred to as "data tags" in machine learning). As mentioned above, the training of the neural network is done offline and is typically performed using powerful computers (e.g., in the cloud) and specialized software tools. Local object recognition at the accessory (e.g., through "inference" using a neural network) has the advantage that the recorded images do not need to be transmitted to an external server, which is particularly preferred for data protection reasons and to protect privacy. It is conceivable that a user records training data for new objects to be identified using the imaging sensor and transmits it to a (automatic) service that sends back to the user new settings for the neural network (e.g., in the form of a "TensorFlow Lite" model) or new firmware for object recognition or classification.

[0047] Figures 5 to 8 Now it is shown how to use according to Figure 2 The illustrated embodiment uses a modified accessory to detect various objects, which subsequently leads to the corresponding assignment of hygienic elements.

[0048] exist Figure 5In this scenario, a user places their hand under a faucet. Once the camera detects the hand in inlet area 13, the water turns on. When the user removes their hand from inlet area 13, the water turns off again. Hand movements (e.g., certain gestures) can also be used to trigger various additional actions, such as setting the water heater / cooler, increasing / decreasing the water supply, and turning the shower on / off. Specifically, gesture assessment identifies certain movement patterns (direction / speed). For example, opening (one or both) a hand (from a clenched fist to extended fingers) can trigger the application of a portion of soap or disinfectant.

[0049] exist Figure 6 In this scenario, the user places a dirty plate under the tap. Once the camera detects the dirty plate in inlet area 13, the warm / hot water is turned on instead of the cold water. Additionally, dishwasher detergent can be added to the water, especially if (heavily) soiled plates are detected, as object recognition also allows for the determination of object attributes such as the degree of soiling. Similarly, once the user removes the (now clean) plate from inlet area 13, the water is turned off.

[0050] exist Figure 7 In this system, the user places a teacup under the tap. Once the camera detects the teacup in inlet area 13, hot / boiling water is turned on. Object recognition also makes it possible to determine the object's attributes, such as transparency, color, and / or size. On one hand, this makes it possible to distinguish cups of different sizes and colors (e.g., whether they are made of glass or ceramic) so that the appropriate amount of hot / boiling water is also added. Filling can also be interrupted by pulling the cup out of the water jet. Once accessory 1 records a change in the distance between the water jet and the edge of the cup, or if that distance falls below a certain value, the water automatically shuts off. For safety reasons, the boiling water can be immediately shut off once the user's hand is detected in inlet area 13 (-> emergency shut-off safety function).

[0051] exist Figure 8 In this process, the user places the measuring cup under the faucet. Once the camera detects the measuring cup in inlet area 13, cold water is turned on. Alternatively, the user can indicate how much water the measuring cup should be filled with by extending their hand (e.g., extending one to four fingers) (e.g., one-quarter full, half full, three-quarters full, or full). If the user places the cooking pot under the faucet inlet area 13 with both hands, it is filled with water until the cooking pot is pulled out of the spray again.

[0052] Figure 9 The application of accessory 1 according to the invention in a shower is illustrated. Water flows out once the camera identifies the user 15 under the shower in the inlet area 13, or specifically his head or upper body. The user 15 can adjust the temperature and the intensity of the water jet, for example, by gestures. The position (e.g., above head height) and / or orientation of the shower head can also be adjusted, and they can be tracked if necessary.

[0053] Similar to the embodiments for shower applications, the accessories according to the invention can also be used in wall-mounted hand dryers or hair dryers, wherein the airflow can be turned on and off, and the temperature and intensity (as well as the direction) can be adjusted based on object recognition.

[0054] Figure 10 A schematic representation of a sixth exemplary embodiment of the accessory 1 for sanitary devices according to the present invention is shown. Figure 1 In an extension of the first exemplary embodiment, the variant has two liquid outlets 17, 17' for two different liquids 11, 11', where the first liquid may be, for example, water, and the second liquid may be, for example, soap or (alcohol) hand sanitizer. If a user holds their hand at position P1 below the first outlet in the inlet area 13, which is detected and monitored by the imaging sensor 5, this is detected by the object recognition system, and the valve for the first liquid is opened via the control unit 3, allowing water to flow out, for example, to wash the hand. If the user now moves their hand from position P1 to position P2 below the second outlet within the inlet area 13, this is detected by the image processing unit, and the control unit 3 immediately closes the valve for the first liquid and briefly opens the valve for the second liquid, for example, to apply a specified amount of soap or sanitizer. The second liquid may also be applied, for example, via a second pipe having a second liquid outlet, which is, for example, arranged laterally (and rearwardly) offset relative to the pipe for the first liquid. Thus, the image processing unit is able to determine the positions P1 and P2 of the object 7 (e.g., the hand), and the control unit 3 performs different actions based on the corresponding positions P1 and P2 of the object 7 (e.g., the hand).

[0055] To identify a (new, previously unknown) object, the user can use the accessory's camera (or another camera, such as a smartphone camera) to capture the corresponding image (or film or video) as training data. As mentioned above, this data can be sent to a server, which then returns an appropriately trained neural network to the accessory (e.g., in the form of a "TensorFlow Lite" model). Assigning an object (or a gesture about that object) to an action can be done or modified by the user themselves, for example, through a web app or an app on their smartphone.

[0056] Further applications of the accessories according to the invention are conceivable, such as in beverage machines or automatic cleaning systems. Depending on the detected container (large, small, transparent, a certain color), different beverages, such as purified water, sparkling water, tea, coffee, soup, etc., are dispensed in the beverage vending machine.

[0057] The proposed accessory can be used in both the private sector and public health facilities, although private sector applications may involve a greater degree of personalization and a wider variety of actions. However, for the public sector, the proposed accessory offers the following advantages: installers, plumbers, or maintenance technicians can easily configure the system using built-in imaging sensors (such as cameras) without the need for control elements, and no special configuration / programming equipment is required to do so.

[0058] Reference Symbol List

[0059] 1. Accessories

[0060] 2. Sanitary facilities

[0061] 3 Control Unit

[0062] 4 Sanitary components

[0063] 5 Imaging Sensors

[0064] 6 lighting units

[0065] 7 Objects

[0066] 8 Image Processing Units

[0067] 9. Communication Unit

[0068] 10 servers

[0069] 11, 11' Liquid

[0070] 12. Component Arrangement

[0071] 13 Entrance Area

[0072] 14 Export Area

[0073] 15 users

[0074] 16 valves

[0075] 17, 17' Exit

[0076] Positions P1 and P2

Claims

1. An accessory (1) for a sanitary device (2), the accessory (1) comprising a control unit (3) and a sanitary element (4) controllable by the control unit (3), wherein the accessory (1) comprises at least one imaging sensor (5) for detecting a scene at the sanitary element (4), wherein the at least one imaging sensor (5) is a 2D or 3D camera, a thermal imaging camera, an ultrasonic sensor, a radar sensor or a laser rangefinder, a LIDAR or ToF sensor or a laser scanner, or a combination thereof, and the control unit (3) is adapted to trigger an action of the sanitary element (4) based on object recognition for identifying at least one object (7) in the scene, the object recognition being based on at least one output signal from the at least one imaging sensor (5), wherein the object recognition is performed by means of a neural network, characterized in that, The neural network is a self-learning neural network, and the accessory (1) is adapted to receive feedback from the user (15) to the control unit (3), through which the user (15) indicates whether the action triggered by the control unit (3) is correct.

2. The accessory (1) as described in claim 1, characterized in that, The feedback from the user (15) can be received by the at least one imaging sensor (5), wherein the feedback from the user (15) is that if the triggered action is incorrect, the user (15) covers the at least one imaging sensor (5) with his hand for a short time.

3. The accessory (1) as described in claim 1 or 2, characterized in that, The accessory (1) additionally includes a microphone, which the user can use to provide feedback to the control unit (3), wherein the control unit (3) is adapted to detect clapping and / or whistling noise based on the output signal from the microphone.

4. The accessory (1) as described in claim 1, characterized in that, Reinforcement learning is feasible.

5. The accessory (1) as described in claim 1, characterized in that, The accessory (1) further includes an image processing unit (8), wherein the image processing unit (8) is adapted to perform the object recognition by means of the neural network.

6. The accessory (1) as claimed in claim 1, characterized in that, The accessory (1) further includes a remote image processing unit (8) for sending the at least one output signal from the at least one imaging sensor (5) to the accessory (1) in the server (10) for object recognition and a communication unit (9) for receiving the result of the object recognition, wherein the image processing unit (8) is adapted to perform the object recognition by means of the neural network.

7. The accessory (1) as claimed in claim 1, characterized in that, The neural network was trained before the accessory (1) was put into operation.

8. The accessory (1) as described in claim 5 or 6, characterized in that, The image processing unit (8) is adapted to perform object classification as part of the object identification in order to assign the object (7) to a specified class from a plurality of predefined classes.

9. The accessory (1) as described in claim 8, characterized in that, By means of the object classification, the user's (15) limbs or different categories of cleaning utensils or accessories of different types can be identified.

10. The accessory (1) as claimed in any one of claims 5, 6 and 9, characterized in that, The image processing unit (8) is adapted to determine at least one attribute of the object (7).

11. The accessory (1) as claimed in any one of claims 5, 6 and 9, characterized in that, The image processing unit (8) is adapted to determine the position of the object (7) relative to a reference position and / or move it.

12. The accessory (1) as claimed in claim 8, characterized in that, The control unit (3) is adapted to trigger a specified action of the sanitary element (4) depending on the object classification and / or the attributes of the object (7) and / or the position of the object (7) relative to the reference position and / or the movement of the object (7).

13. The accessory (1) as claimed in claim 12, characterized in that, One or more of the following actions can trigger the following: - Select the preset of the sanitary element (4) and perform the behavior of the sanitary element (4) according to the definition of the preset; - Apply a specified amount of fluid; - Apply a specified amount of fluid per unit time; - Apply the fluid to a preset maximum temperature, a preset minimum temperature, or a specified temperature. - Depending on the location of the object (7), a specified fluid from a variety of different fluids is applied; and - To turn on and / or off fluid delivery.

14. An accessory arrangement (12) comprising a sanitary device (2) and an accessory (1) according to any one of claims 1 to 13, wherein an imaging sensor (5) is arranged such that an inlet region (13) of the accessory (1) for applying fluid and / or an outlet region (14) for removing the applied fluid can be detected by the imaging sensor (5).

15. The accessory arrangement (12) according to claim 14, characterized in that, The sanitary equipment (2) is a sink or washbasin, a sitz bath, a shower, a bathtub, a soap applicator, a liquid applicator, a disinfectant applicator, a hand dryer, a hair dryer, a toilet, a urinal, a beverage machine, or a laundry station.

16. The accessory arrangement (12) according to claim 14, characterized in that, The sanitary equipment (2) is a shower toilet.

17. A method for triggering the action of a sanitary element, wherein the sanitary element is a sanitary element (4) of the accessory (1) according to any one of claims 1 to 13, the method comprising the steps of: - Detect a scene at the inlet area (13) and / or outlet area (14) of the accessory (1), at the sanitary element (4) having at least one imaging sensor (5), wherein the scene is detected by means of one or more images as the output of the at least one imaging sensor (5); - Perform object recognition for identifying at least one object (7) in the scene; as well as - The action of the sanitary element (4) is triggered based on the object identification. The object recognition is performed using a neural network, characterized in that the neural network is a self-learning neural network, and the user (15) provides feedback on whether the triggered action is correct.

18. The method as described in claim 17, characterized in that, It also includes at least one of the following steps: Using the at least one imaging sensor (5) or a separate camera independent of the accessory (1), capture images of new, previously unknown objects (7); The previously recorded images of new, previously unknown objects (7) are transmitted to the server (10) as training data. Process training data to create a trained neural network; The trained neural network is transmitted from the server (10) to the accessory (1). Assign the specified object (7) or a gesture about the object (7) to a specified action for the control unit (3).

Images