Method for controlling a domestic refrigeration appliance

Abstract

The invention relates to a method for controlling a domestic refrigeration appliance (1), comprising a coolable interior (3) provided for storing foodstuffs, a door (5) provided for closing off the interior (3), a control device (20) and a wireless interface (21), wherein the method comprises the following steps: - establishing a communication channel via the wireless interface (21) to an external device (30) in the vicinity of the domestic refrigeration appliance (1); - collecting channel state information of the communication channel at the wireless interface (21), transmitted by the external device (30); - monitoring a communication parameter read out via the channel state information; - identifying a person (40) in the vicinity of the domestic refrigeration appliance (1) on the basis of the monitoring, wherein the person (40) can be identified on the basis of a change in the communication parameter; and - controlling a function of the domestic refrigeration appliance (1) by means of the control device (20) on the basis of a result of the identification of the person (40).

Description

[0001] 202503111

[0002] 1 / 91

[0003] Method for controlling a household refrigeration appliance

[0004] The present invention relates to the control of a household appliance. In particular, the invention relates to the contactless control of a household appliance with respect to a person. The invention specifically relates to a method for controlling a household refrigeration appliance, a control device for the household refrigeration appliance, the household refrigeration appliance itself, and a system comprising the household refrigeration appliance and another household appliance.

[0005] A household appliance is generally installed in a household and designed to perform a predetermined, household-related task. For example, a household appliance might include a kitchen appliance, such as a stove, an oven, or a household refrigerator. To operate the appliance, a person can use a physical interface designed on the appliance, such as a switch, a knob, a touch-sensitive button, a display panel, or a door. If the person is more than arm's length away from the household appliance, they must move to control it.

[0006] It has been proposed to equip a mobile device, particularly a smartphone, to enable wireless control of the household appliance. A user interface for the appliance could then be displayed on the mobile device. However, remote control over a greater distance is not permitted for household appliances whose operation could potentially lead to an accident, such as a stove.

[0007] It was further proposed to control a household appliance via gesture control. For this to work, the person must be at a predetermined distance from the appliance and perform a predetermined gesture to trigger an associated function. The appliance would include a sensor to scan the person, such as a radar or LiDAR sensor. The sensor typically identifies numerous points on the person, meaning the collected data could be used to draw conclusions about the individual, potentially violating their privacy. Furthermore, such a sensor is expensive, and analyzing its signals can be complex.

[0008] 2 / 91

[0009] There are household refrigeration appliances known to have presence detection. This presence detection is typically implemented using dedicated sensors, such as radar or infrared sensors. These sensors are capable of detecting the approach or presence of a user in front of the appliance, in order to control certain appliance functions, such as external lighting. However, integrating such dedicated hardware sensors involves additional costs and increased design complexity. Furthermore, the previously described issue of privacy also applies here.

[0010] Modern household appliances are increasingly equipped with wireless communication interfaces. These primarily serve to network the device, for example, to connect it to a home network or to control it via a mobile application. In wireless communication, signals are exchanged between a transmitter and a receiver via a communication channel. The properties of this channel, and thus the received signals, are influenced by the physical environment in which the signals propagate. Information about the state of the transmission channel, often referred to as channel state information (CSI), can describe these influences. Changes in the environment, such as the presence or movement of a person, lead to changes in this channel state information, which can then be analyzed.

[0011] It is an object of the present invention to provide an improved method for controlling a household refrigeration appliance, in which the aforementioned problems and disadvantages are at least partially overcome. It is further an object of the present invention to provide an improved control device, an improved household refrigeration appliance, and an improved system, in which the aforementioned problems and disadvantages are at least partially overcome. One of the underlying objectives of the present disclosure is to provide an improved technique for controlling a household appliance by a person.

[0012] This problem is solved by the inventive method according to claim 1, the inventive control device according to claim 13, the inventive household refrigeration appliance according to claim 14, and the inventive system according to claim 15.

[0013] 3 / 91

[0014] Further advantageous embodiments of the invention will become apparent from the dependent claims and the following description of preferred embodiments of the present invention. Persons skilled in the art will understand that the different embodiments and exemplary embodiments of the present disclosure can be combined with one another in a technically meaningful manner that is not explicitly illustrated or described. Persons skilled in the art will also understand that various features illustrated or described with respect to one of the figures can be combined with features illustrated in one or more other figures to produce embodiments that are not explicitly illustrated or described, particularly with respect to features in the further disclosure.

[0015] A first aspect of the invention relates to a method for controlling a household refrigeration appliance, comprising a coolable interior space for storing food, a door for closing the interior space, a control device, and a wireless interface. The method comprises the following steps: establishing a communication channel via the wireless interface to an external device within the vicinity of the household refrigeration appliance; collecting channel status information of the communication channel at the wireless interface, transmitted by the external device; monitoring a communication parameter read from the channel status information; and identifying a person within the vicinity of the household refrigeration appliance based on this monitoring. The person can be identified based on a change in the communication parameter.Controlling a function of the household refrigeration appliance by means of the control device depending on a result of the determination of the person.

[0016] A household refrigerating appliance can be, in particular, a refrigerator, a freezer, or a fridge-freezer combination. Depending on the design, a household refrigerating appliance may have one or more doors.

[0017] Another aspect of the present disclosure includes a method for controlling a household appliance with a wireless interface, including steps for establishing a communication channel via the interface to an external device in the area 202503111

[0018] 4 / 91

[0019] of the household appliance; of monitoring a communication parameter of the channel; and of identifying a person in the vicinity of the household appliance based on the monitoring.

[0020] Features and advantages described in relation to one claim category can be applied equally to all other claim categories. This means, for example, that the disclosure for all embodiments of the method should include corresponding embodiments of the control device, the household refrigeration appliance, and the system, which include means for carrying out the steps of the respective method.

[0021] Advantageously, the wireless interface for communication with the external system can already be integrated into the appliance and used to transmit information. For example, the appliance can be wirelessly connected to a gateway or router, allowing contact with a central system that manages the appliance's operating parameters or provides additional information about its operation.

[0022] The term "wireless interface" can refer in particular to a hardware and / or software component designed for wireless data communication. The interface can be integrated directly onto a circuit board of the control device or implemented as a separate, connected module. The wireless interface can include a transceiver configured to send and / or receive data. The transceiver establishes a communication channel between the control device and another device, enabling the transmission of information. The wireless interface can be, or include, a wireless receiver. The household refrigeration appliance can include the wireless receiver.

[0023] In one embodiment, the interface operates in the 60 GHz range. In this range, increased resolution allows for more informative measurements of the individual. Experience and established procedures from radar technology can be used to evaluate the surveillance data. In particular, a synthetic aperture radar technique can be employed to enable rapid and

[0024] 5 / 91

[0025] Perform high-resolution scanning of the person. The interface preferably continues to operate using OFDM (orthogonal frequency-division multiplexing).

[0026] The interface preferably operates according to one of the IEEE 802.11 standards. These standards define WLAN norms, also known as Wi-Fi. The wireless interface can be implemented as a Wi-Fi module or a Wi-Fi chip. A person can be scanned using a technique called Wi-Fi sensing. Interfaces for communication based on these standards are widespread and well-understood. Transceivers can be used cost-effectively. Some of the standards use a high transmission frequency, which is particularly suitable for person scanning. In one embodiment, the interface supports the IEEE 802.11bf standard, for which a draft exists. Alternatively or additionally, other technologies such as Bluetooth, Zigbee, or other suitable near-field communication standards can also be used as the wireless interface.

[0027] The term "control device" can, in particular, refer to a central control unit of the domestic refrigeration appliance. This can be configured, for example, as a main processor, a microcontroller, or a dedicated module such as a system master module. The control device can perform both the control tasks necessary for the basic operation of the domestic refrigeration appliance and the process steps according to the invention. It is also conceivable that the control device consists of several interacting logical or physical units. The control device comprises a memory and a processor. The processor can comprise one or more processors. The memory can comprise one or more memory devices.

[0028] The "external device" can be any network-enabled device within range of the wireless interface and capable of establishing a communication channel with it. Typically, this could be a Wi-Fi router or access point of the home network. A repeater, a powerline adapter with Wi-Fi capability, or a node in a mesh network are also conceivable. In some configurations, the external device could also be another household appliance or a mobile device. The external device can be a wireless transmitter or include a wireless transmitter.

[0029] 6 / 91

[0030] The term "channel state information" or "CSI" can refer specifically to a set of data that describes the characteristics of the communication channel between the wireless interface and the external device. This can include, for example, information about the amplitude and phase shift of individual subcarriers of an OFDM signal. More general values ​​such as the received signal strength indicator (RSSI) or signal-to-noise ratios can also be considered part of the channel state information. A wireless signal can consist of one or more pieces of channel state information received in packets.

[0031] The "communication parameter" can be understood, in particular, as a quantitative value derived from the raw channel state information to represent a channel property relevant for presence detection. For example, the parameter could be the variance of amplitude or phase values ​​across multiple subcarriers, an average of the phase shifts, or even a more complex index calculated from multiple CSI values. In a simple embodiment, the signal strength value (RSSI) can also serve as the communication parameter. The communication parameter can be a parameter.

[0032] The communication parameter can be assigned to a physical layer. The physical layer, layer 1 in the OSI model, provides mechanical, electrical, physical, and other functional means to activate and deactivate physical connections, maintain them, and transmit bits over them. Wireless communication, in particular, uses radio waves, which are converted into electrical signals and vice versa by a transceiver (a combination of transmitter and receiver). An electrical parameter can be sampled and, in particular, digitized to support further processing.

[0033] The communication parameter can be assigned to a data link layer. The data link layer can include a Media Access Control (MAC) and / or a Logical Link Control, and the communication parameter can be located in one of these parts.

[0034] 7 / 91

[0035] This layer is layer 2 in the OSI model and abstracts the physical layer.

[0036] The term "monitoring" can refer specifically to the process of repeatedly or quasi-continuously recording the value of a communication parameter over a period of time. This can be done, for example, by sampling the parameter value at predefined, regular time intervals. The result of this step can be a time series of parameter values, the progression of which can then be analyzed for significant changes or patterns. In other words, monitoring creates a dynamic data foundation upon which the subsequent identification of the individual is based.

[0037] In a further development of the invention, communication channels of several household appliances are monitored, and the person within the vicinity of one of the household appliances is identified based on the monitoring data. The necessary exchange of observations or monitoring data can advantageously take place via wireless interfaces used for monitoring. In practice, monitoring data from any number of household appliances can be collected and evaluated together. The evaluation can be performed by one of the household appliances or by a wirelessly connected external device. Results of the evaluation can be forwarded to one or more household appliances. Advantageously, the direct wireless exchange of information between the household appliances ensures that they are located in a common area, particularly within a household, and more preferably within a room of a household.A person can influence communication parameters of multiple channels while moving through the area.

[0038] The communication channel can support one-to-many communication with multiple participants. Such a connection is also called point-to-multipoint. For example, a household refrigerator or appliance can communicate with the external device and additionally with one or more other household appliances. Alternatively, one or more independent one-to-one (point-to-point) channels can be used.

[0039] 8 / 91

[0040] A "change in the communication parameter" can be understood, in particular, as a significant deviation of the current value from a reference value. The reference value could be, for example, a previously measured value, a moving average over several measurements, or a stored baseline for an empty space. A change can be considered detected when a predefined threshold is exceeded or when the rate of change of the parameter exceeds a certain value.

[0041] The communication parameter can be observed over a predetermined period. Changes can be characteristic of specific events, such as a person approaching or moving away, a characteristic of the person, or a gesture performed by the person. Multiple communication parameters can also be evaluated for each communication channel.

[0042] The step of "determining a person" can refer specifically to the logical process of inferring the presence of a person or user based on a detected change in a communication parameter. This can be achieved, for example, through threshold comparison, pattern recognition of the temporal signature of the change, or by using a machine learning algorithm trained to recognize characteristic patterns of human presence. The person-determination step can also be referred to as "presence detection." The person could be a user. The user could be a user of the household refrigeration appliance. The person-determination step can be preceded by preprocessing of channel state information. This channel state information can be received from packets.

[0043] "Within the vicinity of the household refrigeration appliance" means that the external device does not necessarily have to be located in the same room as the household refrigeration appliance. As a minimum requirement, it can be understood that the external device, for example, a Wi-Fi router, is positioned in such a way as to allow a communication channel to be established with respect to the wireless interface of the household refrigeration appliance. This can already be the case if both devices are located in the same household or apartment, so that the 202503111

[0044] 9 / 91

[0045] The communication channel crosses the area relevant for presence detection.

[0046] The "result of the determination" can, in particular, refer to the information provided for function control after the determination step. The result of the determination can be output as a signal. The result can be simple binary information, such as "person present," "person absent," or other information that may be relevant for controlling a function. In embodiments, the result can include more differentiated information, for example, an estimated distance of the person from the device, the detection of a movement or gesture, or even a rough identification of the person.

[0047] The household appliance, or household refrigerator, can be controlled depending on the specific person. In particular, an operating parameter of the appliance can be controlled depending on the presence, distance, identity, and / or gesture of the person.

[0048] "Controlling a function" can include, in particular, the targeted triggering, modification, suppression, or adjustment of various functions or operating functions of the household refrigeration appliance. Examples include suppressing a door alarm, delaying noisy actions such as compressor start-up, activating lighting or a door opening device, switching between an energy-saving mode and a normal operating mode, or sending a notification to an external device. Other controllable functions known to a qualified technician are also conceivable. Controlling a function can be implemented physically by one or more actuators. The actuator can be controlled by the control device to perform the corresponding function or action.Examples of actuators on household refrigeration appliances include: a lighting device, a control device, a door opening device, a sound transducer for a door alarm, or a compressor.

[0049] The term "control" in the context of the invention can encompass controlling and / or regulating a function. This depends on the specific function to be modified. 202503111

[0050] 10 / 91

[0051] Controlling the function based on the result of the determination means, in particular, that a control function can be selected or is selected based on the result. For example, different control functions can be selected depending on the result. In other words, for example, if result "A" is obtained, function "A" is selected and controlled. Conversely, if result "B" is obtained, function "B" is selected and controlled.

[0052] It is proposed to use the wireless interface to wirelessly detect the person. The appliance would not need to be equipped with an additional sensor for person detection, resulting in cost savings. By monitoring the communication parameters, the person can be identified with just enough precision to allow the appliance to be controlled based on their identity. Over-detection, which could potentially infringe on the person's privacy rights, is virtually eliminated.

[0053] This allows the existing wireless interface for presence detection to be used, eliminating the need for an additional sensor. This reduces component costs and design effort. The approach according to the inventive method offers high flexibility, as further or improved functions for context-sensitive control of the device can be implemented directly through software updates without requiring hardware modifications. This enhances the user experience and improves user comfort in the home, particularly in the kitchen.

[0054] There are embodiments in which a person's presence, absence, distance, identity and / or gesture is determined and output as a result of the person's determination.

[0055] To identify the person, a signal strength or a related parameter can be combined with further information regarding the channel, such as the direction in which the external location with which the household appliance communicates is situated.

[0056] The determination of the "distance of the person" can be based, in particular, on the fact that the magnitude of the change in the communication parameter correlates with the person's proximity to the household refrigeration appliance. For example, a strong change in parameter 202503111

[0057] 11 / 91

[0058] A change in position can be interpreted as "close to the device," while a weaker change is interpreted as "present in the room." Various threshold values ​​can be defined for the communication parameter, representing different distance zones. For example, it can be determined whether a person is directly in front of the household refrigerator. These threshold values ​​can be made available to the control device, for example, in a memory. They can be factory preset or adapted to the specific environment through a calibration process. The memory can be, for example, built into the household refrigerator or stored in the cloud.

[0059] Determining the "identity of a person" can be based, in particular, on recognizing person-specific patterns in the temporal changes of communication parameters. For example, different people can generate unique signatures in the channel state information due to their height, gait, or typical movement patterns. An algorithm trained on this, especially a machine learning algorithm, can then assign these signatures to different users. "Identity" can be understood as a distinction between "User A" and "User B," or as a classification, for example, as "adult," "child," or "pet."

[0060] The recognition of a person's gesture can be based, in particular, on the analysis of short-term, dynamic changes in communication parameters. Unlike simple presence detection, this method evaluates characteristic temporal patterns or signatures generated by specific movements. Examples of such gestures include a swiping motion with the hand, a pushing motion towards the door, or a deliberate pause. An algorithm can be trained to distinguish these specific signatures from random movements.

[0061] The "output as a result of the determination" can, in particular, include the internal provision of the determined information for the subsequent step of controlling the function. The output can include an output signal. This can be achieved, for example, by setting a flag in memory, writing a result, or...

[0062] 12 / 91

[0063] representing value into a register or by transferring a data object to a software component of the control device.

[0064] This allows the household refrigerator to react not only to the presence of a person, but also to how they are present. For example, a function can be triggered only when a person approaches the appliance (distance), performs a specific gesture, or if it is a specific user (identity). This improves ease of use and functionality. Furthermore, it reduces the potential for malfunctions.

[0065] In one embodiment, depending on the result of the determination, the control of a function is prevented. This prevention can, for example, last for a predefined period.

[0066] There are embodiments in which controlling one function of the household refrigeration appliance includes controlling various functions of the household refrigeration appliance, each of which is selected depending on the result of the determination of the person.

[0067] The "control of various functions" can, in particular, mean that the control device can access a portfolio or a predefined selection of different controllable functions. This selection can include, for example, functions such as controlling the door alarm, adjusting operating noises, activating lighting, controlling a door opening device, controlling a compressor, or sending notifications. The method according to the invention is not limited to controlling a single function.

[0068] The "selection based on the result" can, in particular, include a predefined assignment logic. This logic can assign one or more functions to be controlled to a specific result of the determination. This logic can be stored in the control device, for example, as an assignment table or a series of "if-then" rules. For instance, the result "person near the device" or the result "person approaching the device" could be assigned the function "activate the lighting," while the result "person in the room" could be assigned the function "activate the lighting."

[0069] 13 / 91

[0070] The function "Suppress noisy actions" is assigned to the "present" setting. A single function can be controlled at the same time, or multiple functions can be controlled simultaneously.

[0071] This enables context-sensitive control of the household refrigerator. The refrigerator can intelligently and independently decide which action is most appropriate in a given situation, instead of triggering the same function every time presence is detected. This further increases user-friendliness and the perceived benefit of the function control.

[0072] There are embodiments in which the control of different functions can be activated and / or deactivated independently of each other by means of an input device.

[0073] The term "input device" can be understood to mean, in particular, any interface that allows a user to configure settings for the functions controlled by presence detection. This could be, for example, a touchscreen display attached to the domestic refrigeration unit itself or a control panel with buttons. In another embodiment, the input device can be an external user device, such as a smartphone or tablet, with input possible via an application installed on the device. The application could, for example, be an app provided by the manufacturer of the domestic refrigeration unit.

[0074] Alternatively or additionally, activation and / or deactivation can also be performed without the input device. For example, activation and / or deactivation can occur after a predefined time window has elapsed.

[0075] The fact that the functions can be "activated and / or deactivated independently" can be understood to mean, in particular, that a user has the option of individually switching each of the various functions controllable by presence detection on or off. For example, a user can activate the function to suppress the door alarm while simultaneously deactivating the function to activate the lighting upon approach. In other words, the user is not forced to use a fixed 202503111

[0076] 14 / 91

[0077] Instead of accepting bundles of function controls, the device's behavior can be individually adapted to personal preferences. In one example, the control of one or more functions can be activated and / or deactivated for a predefined period.

[0078] This allows for a high degree of customizability. Furthermore, it can increase user acceptance. By giving the user full control over which functions should be actively controlled, unwanted automatic behavior can be avoided and the device's behavior can be optimally adapted to individual needs and preferences.

[0079] There are embodiments in which the controlled function is to suppress a door alarm when the door is open, as long as the presence of the person in front of the household refrigeration appliance is determined, and in particular to activate the door alarm when the door is open, upon determination of the absence of the person in front of the household refrigeration appliance for a predefined period.

[0080] The open state of the door can be detected using a method known or commonly used by a qualified professional. For example, the household refrigerator can include a switch and / or sensor designed to detect the open state of the door. In another embodiment, the open state can be determined from channel status information. When the door is opened, a corresponding signal can be transmitted to the control device. This signal can serve as a starting point for determining the door's open time. The signal can also serve as a starting point for determining whether a person is present in front of the household refrigerator.

[0081] The term "suppressing a door alarm" can be understood to mean, in particular, that the regular audible door alarm, which is normally triggered after a certain period of time the door has been open, is not activated or is suppressed. This is especially advantageous when the user is loading or unloading the household refrigeration appliance or cleaning the interior. Instead of a loud, continuous audible alarm signal, either no signal is emitted when a person is detected, or, as described in a further embodiment, a modified signal is emitted. (202503111)

[0082] 15 / 91

[0083] An acoustic signal can be generated using a sound transducer, for example a loudspeaker.

[0084] The phrase "as long as presence is being determined" can specifically mean that the process of determining the person is not a one-time event, but rather a repeated or quasi-continuous process while the door of the household refrigerator is open. This can be achieved, for example, by having the control device re-monitor the communication parameter and perform the person determination at predefined, short intervals, such as every 20, 30, or 60 seconds. In other words, the "presence" status is continuously verified to maintain the suppression of the door alarm.

[0085] The "Activating the door alarm when absent" function means, in particular, that the temporary suppression of the door alarm is lifted. This occurs as soon as it is detected that no one is in front of the appliance, i.e., when the result of repeated detection changes from "presence" to "absence." In other words, if the person leaves the area of ​​the household refrigerator with the door open, the door alarm is reactivated to prevent the door from being accidentally left open.

[0086] The "predefined period" can be understood as a buffer or delay time that elapses after the person's absence has been detected, before the door alarm is actually activated. In other words, the door alarm is not triggered immediately when the person leaves the area, but only after this period has expired. This period, which can be, for example, 15, 30, or 60 seconds, serves to prevent false alarm activation if the person only briefly turns away or if the detection is momentarily interrupted. The value for this predefined period can be factory-set or preset by the user via the input device.

[0087] In another embodiment, the door alarm can be activated after a predefined maximum period, even if a person is present. This predefined maximum period can, for example, be a multiple of the previously described predefined short periods. 202503111

[0088] 16 / 91

[0089] This can increase the ease of use of the household refrigerator. A disruptive alarm sounding while putting away groceries or cleaning is avoided, without impairing the important safety function of the door alarm in case the door is accidentally left open.

[0090] There are embodiments in which, during the suppression of the door alarm, a predefined, in particular acoustic, signal sequence is emitted, which is selected depending on a number of specific presences of the person and on a time component.

[0091] The signal sequence can be output directly by the control device. Alternatively, the control device can issue a control command to another device. This other device can be configured to output the signal sequence depending on the control command. The same applies analogously to the door alarm. The signal sequence and the door alarm can be output by the same device. This device can be a sound transducer.

[0092] The output of the predefined signal sequence is noticeably different from the door alarm for the person in front of the household refrigerator. A "predefined signal sequence" can be understood, in particular, as a discreet and less disruptive form of notification that is emitted instead of the door alarm. Instead of a loud, continuous tone, the sequence could, for example, consist of a single, short beep or a series of several beeps spaced apart. The acoustic properties of the beeps, such as their volume, pitch, or duration, can be chosen so that they are perceived as a gentle reminder rather than a disruptive alarm. In another embodiment, a visual signal can be output as an alternative or additional element to the acoustic signal sequence. A visual signal could, for example, be a light that is activated.

[0093] The dependency on a "number of specific occurrences and a time component" specifically includes the logic according to which the signal sequence escalates over time. In other words, the type of signal sequence, for example, the count 202503111

[0094] 17 / 91

[0095] The tones emitted are not static, but change depending on the duration of the state "door open and person present". The time component can be a predefined interval after which the signal sequence is modified, while the "number of confirmed presences" can refer to the fact that escalation only occurs if presence is confirmed over several confirmation cycles.

[0096] In one embodiment, the signal sequence can be implemented as a kind of escalating reminder mechanism. If the presence of a person is detected when the door is open, a single, short beep is emitted after an initial open time of, for example, 90 seconds, instead of the regular continuous door alarm. At the same time, the door alarm is suppressed. The sequence is then extended over time, for example, by adding an additional beep after every further 60-second interval. Thus, after a total of 150 seconds, a sequence of two beeps would be emitted, and after 210 seconds, a sequence of three beeps. This escalation can continue for a predefined number of stages before the regular, continuous door alarm is finally activated to signal that the door remains open.This logic can be overridden if the person's presence can no longer be determined. In this case, after a predefined period of absence, for example 30 seconds, the regular, continuous door alarm is activated to reliably signal if the door has been accidentally left open.

[0097] This allows the user to be gently reminded that the door is open, without being immediately disturbed by a loud alarm. The escalation mechanism ensures that the reminder becomes more insistent over time if the door remains open for an extended period. This helps maintain adequate food refrigeration.

[0098] There are designs in which, during an energy-saving mode of the household refrigerator, the steps for identifying the person are repeatedly carried out after a predefined time sequence. 202503111

[0099] 18 / 91

[0100] An "energy-saving mode" can be understood, in particular, as an operating state of a household refrigeration appliance in which the control unit and / or the wireless interface are not operated continuously, but only when needed, in order to reduce overall energy consumption. This can be implemented, for example, as a "sleep mode" of the control unit. During energy-saving mode, parts of the control unit or the entire wireless interface are temporarily deactivated or put into a state with reduced power consumption. From this sleep state, the control unit then only wakes up for a short period of time to perform necessary tasks, such as identifying a person.

[0101] The fact that the steps are performed "repeatedly, each time according to a predefined time sequence" can particularly describe the intermittent nature of the determination in energy-saving mode. Instead of continuous monitoring, the control device only wakes up at fixed time intervals to perform the steps of collecting and monitoring the channel state information. In one case, the predefined time sequence could mean that the control device wakes up every 60 seconds for a short time window, for example, 5 seconds, to perform a determination. Between these active phases, the control device is in energy-saving mode.

[0102] This results in significant energy savings. The inventive method can thus be maintained even during periods of inactivity of the control device, without requiring the control device and the wireless interface to be continuously operated at full power. Therefore, the energy efficiency of the household appliance is not affected by the inventive method.

[0103] There are embodiments in which the controlled function is a suppression and / or activation of an operating action of the household refrigeration appliance, in particular a noisy operating action, for a predetermined period.

[0104] An "operating action", in particular a "noise-intensive operating action", can be understood to mean any process of the household refrigeration appliance that is necessary to maintain its cooling function and that potentially produces audible noise.202503111

[0105] 19 / 91

[0106] Noise is generated. This includes, for example, the start-up and operation of the refrigeration compressor, the operation of a fan for air circulation in the interior or in the condenser area, or defrosting system processes. Suppressing and / or activating the operating action encompasses two complementary control strategies for noise management of the domestic refrigeration appliance. Suppression can involve delaying or postponing an upcoming noisy operating action when a person is detected present, in order to reduce the acoustic burden on the user during their presence. The operating action can be delayed or postponed to a period that is not disruptive to the person. A non-disruptive period could be when the person is absent.Furthermore, a non-disruptive period can occur, for example, when a range hood is activated or another noisy operation is underway. Activation, on the other hand, can include the deliberate allowing or prioritizing of such an operational activity when a person's absence is determined in order to schedule necessary but noisy operations into time windows in which they will not cause disturbance.

[0107] In this context, the phrase "for a predetermined period" can refer specifically to a maximum duration for which an operating action is suppressed. This period serves as a safety mechanism to ensure that the technical functionality of the domestic refrigeration appliance, particularly the maintenance of the set indoor temperatures, is not compromised. The period can be set to a value such as one or two hours. For example, the value can be factory preset or configured by the user within technically safe limits.

[0108] In a specific embodiment, particularly in connection with the energy-saving mode, a specific trigger logic for suppressing the operating action can be implemented. If the person's presence is detected intermittently, for example in 60-second cycles, the control device can initiate the suppression of the noisy operating action as soon as a person's presence is confirmed for a predefined number of consecutive cycles. For example, with three consecutive cycles, the confirmation period is three minutes. The predetermined suppression period can, in this case, be set to a maximum value.

[0109] 20 / 91

[0110] The suppression period can be set to, for example, two hours. Conversely, the suppression can be ended prematurely if the person's absence is detected for a corresponding number of cycles, for example, three consecutive cycles, so that normal operation can resume.

[0111] In one embodiment, the operating action can be suppressed when a person is present. When a person is absent, the operating action can be activated.

[0112] This improves acoustic comfort for the user. In situations where quiet is desired, such as in open-plan kitchens, the appliance can adjust its operation to minimize disruptive noise when someone is present. At the same time, food safety is ensured by limiting the duration of the noise suppression.

[0113] There are embodiments in which the controlled function is the activation of a lighting device, an operating device, an energy-saving mode, a door opening device, and / or the control of a function of another device, in particular a household appliance, in the vicinity of the household refrigeration appliance.

[0114] Activating an externally perceptible "lighting device" and / or "control device" can include, in particular, the targeted switching on or waking up of these components on the domestic refrigeration appliance when the approach or immediate proximity of a person is detected. The lighting device can be, for example, external lighting, especially ambient lighting, handle lighting, or backlighting of design or decorative elements. Activating the control device can include waking a display from a standby state or switching on the backlighting of touch-sensitive buttons. In other words, the domestic refrigeration appliance interacts when a person approaches, creating a visually appealing "show effect" or a welcome function.

[0115] 21 / 91

[0116] Activating an energy-saving mode can include, in particular, automatically switching the household refrigeration appliance to an operating state with reduced energy consumption if the result of the determination is that the person will be absent for a longer, predefined period. This could, for example, be the automatic activation of a vacation mode. In vacation mode, the target temperature in the room can be moderately increased. Switching to a general standby state, in which unnecessary components such as a display or certain communication functions are deactivated, also falls under this category.

[0117] The "activation of a door opening device" can include, in particular, the targeted control of an electrical and / or electromechanical device that assists in or fully executes the opening of the door. Activation can occur, for example, when the detection result indicates not only the immediate proximity of a person but also a likely intention to open the door. Such an intention can be inferred, for instance, from deliberately remaining in front of the door or from the recognition of a specific gesture. Such door opening devices can be implemented as motorized push-to-open systems or as fully automatic door operators.

[0118] "Controlling a function of another device" can refer specifically to the ability of a household refrigerator to use the result of its presence detection as a trigger for actions in other networked devices. "In the vicinity of the household refrigerator" can mean that the other device is located in the same household. For example, the detection of a prolonged absence of all persons in the room could cause the household refrigerator to send a control command to activate standby mode to all other networked household appliances in the kitchen. Integration into a higher-level smart home system is also conceivable. For example, the presence of a person detected by the household refrigerator could trigger the initiation of a scenario, such as opening blinds or switching on a coffee machine.In other words, the household refrigerator acts as a central presence sensor for the room, the information from which can be used by a system of multiple devices. 202503111.

[0119] 22 / 91

[0120] This can increase user comfort. This can contribute to energy savings. The present invention can support the function of complex smart home scenarios by using and sharing the environmental information obtained via the inventive method.

[0121] There are embodiments in which controlling the function includes sending a notification to an external user device, the notification containing information about the result of the determination of the person within a predefined time window.

[0122] The term "sending a notification to an external user device" refers specifically to the transmission of information from the control unit of a domestic refrigeration appliance to an external user device. This external user device could be, for example, a smartphone, tablet, or smartwatch. The notification itself can be displayed as a push notification. This display can be implemented, for example, via a specific application, as an email, or as an SMS. Transmission typically occurs via the internet, with the control unit sending the information to a backend server, which then forwards it to the registered user device.

[0123] The "predefined time window" can refer specifically to a user-configurable period during which presence monitoring is active and a notification should be triggered. This period can be defined, for example, as a daily recurring interval, such as "every morning between 9:00 and 10:00 AM," or as a one-time monitoring period. In other words, the time window specifies the particular observation period for which information about a person's activity or inactivity is of interest. This predefined time window can be set via the aforementioned user device.

[0124] The "information about the result" can, in particular, be the specific notification of whether or not the person's presence was determined within the predefined time window. The notification can, for example, be a positive confirmation that movement has been detected. In one implementation, the notification is sent if the person's presence was not detected during the entire time window.

[0125] 23 / 91

[0126] This was determined. In other words, the system can be configured to issue a warning if expected activity fails to occur.

[0127] One example of how this function can be controlled is an application in the field of care for people living alone. A caregiver can, for example, use an app on their smartphone to define a daily time window, such as from 8:00 to 10:00 a.m. The household refrigerator in the home of the person being cared for then actively monitors during this period to see if a person is detected within its range. The caregiver receives a notification on their smartphone if no person is detected within this time window. If the person being cared for is detected within the time window, the caregiver can receive a notification that everything is alright. Alternatively, no notification can be sent if the person being cared for is detected within the time window.In a further development, the determination can also be supported by other devices, in particular household appliances within the apartment.

[0128] This allows for a discreet and privacy-friendly monitoring function. Without the use of cameras, reassuring certainty about a person's activity can be obtained, which is particularly valuable when caring for elderly or vulnerable individuals. It is also conceivable to implement burglary detection in the same way.

[0129] There are embodiments in which, to establish the communication channel, the wireless interface sends a request signal, in particular a ping, to the external device, and the channel status information is collected from subsequently received packets.

[0130] A "request signal" can be understood, in particular, as a data packet actively transmitted by the wireless interface, whose primary purpose is to receive an immediate response from the external device. Unlike random network traffic, targeted and on-demand data packets can be generated. Channel status information can be extracted from these data packets. A "ping," for example, can be considered an ICMP (Internet Control System) message.

[0131] 24 / 91

[0132] The Message Protocol (EMP) is implemented as an echo request packet. In other words, the request signal serves as a tool to actively sample the communication channel and obtain a current snapshot of its properties.

[0133] The fact that channel state information is collected "from subsequently received packets" can be understood to mean, in particular, that the analysis is based on the response packets that the external device sends back as a direct reaction to the previously sent request signal. Since these response packets traverse the communication channel in reverse, they contain the current channel state information relevant for the analysis. This approach ensures that the collected data is not outdated but reflects the state of the channel at the time of the active query.

[0134] This allows data for identifying a person to be provided reliably and promptly. The system does not rely on random network traffic, but can generate the data necessary for presence detection itself, in a targeted manner and at the required intervals. This increases the robustness and responsiveness of the entire system.

[0135] There are embodiments in which the identification of the person is determined by means of a machine learning algorithm that is set up to identify the person based on changes in the communication parameter.

[0136] The machine learning algorithm can comprise one or more algorithms. A person can be identified using an artificial intelligence technique designed to determine the person based on changes. This technique can be trained on a large number of people and their actions. Thus, artificial intelligence (AI) can determine a person's parameter based on monitoring, such as their identity, distance, etc.

[0137] The algorithm for machine learning can be a machine learning algorithm. In particular, an "algorithm for machine learning" can be understood as a computer-based model that is not explicitly designed for the 202503111

[0138] 25 / 91

[0139] The recognition task is not programmed directly, but the algorithm possesses the ability to learn from example data. For instance, the algorithm can use a neural network such as a Convolutional Neural Network (CNN) or a Recurrent Neural Network (RNN). Other models like Support Vector Machines (SVMs), decision trees, or random forests are also conceivable. Preprocessing the channel state information to identify the person can be done using the machine learning algorithm.

[0140] The requirement that the algorithm is "set up" to identify the person can include, in particular, the process of training the algorithm. For this purpose, the algorithm is provided with large amounts of training data (training dataset) that includes measurements of the communication parameter both when a person is present and when a person is absent. In more advanced versions, the training data can also contain more detailed "labels" that describe, for example, the person's distance, their identity, or a gesture performed. During this training process, the algorithm learns independently to recognize the characteristic patterns and signatures in the data that correlate with a person's presence.After training, the algorithm is then able to be applied to new, previously unknown data of the communication parameter in order to make a reliable and differentiated determination that can encompass aspects such as presence, absence, distance, identity, or gesture. Using the training dataset, the machine learning algorithm can achieve a predetermined performance level or convergence. The control device can include the algorithm. The control device can include the algorithm as a software component. The algorithm can be stored in memory within the control device. The algorithm can be updated, for example, via an update of the control device.

[0141] This allows for the recognition of complex relationships within the data that would be difficult to capture using simple threshold comparisons. This enables a more reliable distinction between relevant events, such as the presence of a person, and interfering factors. This significantly reduces the error rate and opens up the possibility of sophisticated recognition tasks such as gesture or identity recognition.

[0142] 26 / 91

[0143] A second aspect of the invention relates to a control device for a household refrigeration appliance, wherein the control device comprises a wireless interface and a processing unit; wherein the processing unit is configured to establish a communication channel via the wireless interface to an external device in the vicinity of the household refrigeration appliance; to collect channel status information of the communication channel at the wireless interface, which has been transmitted by the external device; to monitor a communication parameter read out via the channel status information; to identify a person in the vicinity of the household refrigeration appliance based on the monitoring, wherein the person can be identified based on a change in the communication parameter; and wherein the control device is configured to control a function of the household refrigeration appliance depending on the result of the identification of the person.

[0144] Another aspect of the disclosure relates to a control device for a household appliance, comprising a control device for a household appliance, a wireless interface, and a processing unit; wherein the processing unit is configured to establish a communication channel via the interface to an external device in the vicinity of the household appliance; to monitor a communication parameter of the channel; and to identify a person in the vicinity of the household appliance based on the monitoring.

[0145] The term "processing unit" can be understood to mean, in particular, the logical or physical unit responsible for executing arithmetic operations and control commands. The processing unit may be configured to execute a procedure described herein, either partially or completely. For this purpose, the processing unit may be electronic and may, for example, comprise a programmable microcomputer, a microcontroller, a microprocessor, an application-specific integrated circuit (ASIC), and / or a system-on-a-chip (SoC). The procedure may be in the form of a computer program product with program code means. The computer program product may also be stored on a computer-readable data carrier. The processing unit may constitute the central computing unit of the control device. Alternatively, the 202503111

[0146] 27 / 91

[0147] The processing unit can be designed as a specialized subunit of the control device. In this case, the processing unit can be provided for carrying out the process steps according to the invention, in particular the analysis of the channel state information. Features or advantages of the method according to the invention can be transferred to the device and vice versa.

[0148] With regard to the further terms used in claim 13 and the functions performed by the control device, full reference is made to the explanations already given in connection with the method according to the invention. These apply accordingly to the control device according to the invention.

[0149] The control device according to the invention has the same advantages as already described in connection with the method according to the invention. In particular, as a hardware component, it enables the implementation of cost-effective and flexible presence detection without additional sensors. This allows for a cost-effective and efficient solution for use in modern household refrigeration appliances.

[0150] A third aspect of the invention relates to a household refrigeration appliance comprising the control device according to the invention.

[0151] According to yet another aspect of the present disclosure, a household appliance comprises a control device as described herein. The household appliance can be mobile, for example in the form of an autonomous or manual vacuum cleaner or a food processor; preferably, however, the household appliance is stationary, i.e., a freestanding device, so that its position does not change between measurements. The household appliance can, for example, comprise a kitchen appliance such as a household refrigerator, a stove, an oven, a dishwasher, or a range hood. Alternatively, the household appliance can also comprise a laundry care device such as a washing machine or a clothes dryer.

[0152] The control device can be located, in particular, in an upper section of the household refrigeration appliance. For example, the control device 202503111

[0153] 28 / 91

[0154] The control device is located between an outer casing and the interior or inner container, particularly in the area of ​​an insulating foam. In another embodiment, the control device can be located on the door.

[0155] Regarding the advantages and further embodiments of the household refrigeration appliance, full reference is made to the preceding descriptions of the inventive method and the inventive control device, which apply accordingly to the household refrigeration appliance. The household refrigeration appliance thus benefits in the same way from the cost-efficient and flexible implementation of presence detection and the associated functions.

[0156] A fourth aspect of the invention relates to a system comprising the household refrigeration appliance according to the invention and another household appliance with a wireless interface and a processing unit; wherein the processing unit of the further household appliance is configured to establish a communication channel via the wireless interface to an external device in the vicinity of the household appliance; to collect channel status information of the communication channel at the wireless interface, which has been transmitted by the external device; to monitor a communication parameter read out via the channel status information; and to transmit a result of the monitoring to the household appliance and / or the household refrigeration appliance; and wherein the processing unit of the household refrigeration appliance and / or the household appliance is configured to identify the person based on multiple monitoring operations.

[0157] According to yet another aspect of the present disclosure, a system comprises a household appliance or household refrigeration appliance described herein and another household appliance. The other household appliance comprises a wireless interface and a processing unit configured to establish a communication channel via the interface to an external device within the household appliance; to monitor a communication parameter of the channel; and to transmit a result of the monitoring to the household appliance. The processing unit of the household appliance is configured to identify the person based on multiple monitoring operations. 202503111

[0158] 29 / 91

[0159] The system according to the invention can be understood in particular as a network of interconnected household appliances that collaboratively contribute to presence detection. The additional household appliance, for example an oven, a dishwasher, a washing machine, or a coffee machine, is also equipped with a wireless interface and a processing unit and independently performs the steps of monitoring a communication parameter. It thus functions as an additional, spatially separate sensor point in the room or in the household.

[0160] The result of its own monitoring is then transmitted by the other household appliance to the household refrigeration unit and / or to itself. The final identification of the person is then based on these "multiple monitorings," i.e., through the fusion of data from at least two different locations. As outlined in the claim, the processing unit of the household refrigeration unit and / or the processing unit of the other household appliance can be configured to perform this data fusion. This means that the final identification or calculation can take place centrally in the household refrigeration unit, decentrally in the other household appliance, or collaboratively between the devices.

[0161] This further increases the accuracy and robustness of presence detection. By combining data from different perspectives, "blind spots" in the room can be eliminated and a person's position determined more precisely. For example, the system can more reliably distinguish between a person simply walking through the room and a person standing in front of a specific device. This enables the implementation of even more complex and reliable smart home scenarios without the need for additional sensors.

[0162] Non-limiting embodiments of the invention are now described in more detail with reference to the accompanying figures. These show:

[0163] Fig. 1 shows a schematic representation of a household refrigeration appliance according to the invention;

[0164] Fig. 2 is a schematic representation of a control device according to the invention; Fig. 3A is a flowchart of a method according to the invention;

[0165] Fig. 3B shows a schematic progression of channel state information during the absence and presence of a person;

[0166] Fig. 4 a schematic representation of an external user device;202503111

[0167] 30 / 91

[0168] Fig. 5A is a schematic representation of the household refrigeration appliance according to the invention, during the suppression of a door alarm;

[0169] Fig. 5B shows a flowchart of the inventive method for controlling the suppression of the door alarm;

[0170] Fig. 5C shows a flowchart of the inventive method for controlling the suppression of the door alarm and the output of a signal sequence;

[0171] Fig. 6 shows a schematic representation of the household refrigeration appliance according to the invention, while a lighting device or an operating device is being controlled;

[0172] Fig. 7 shows a schematic representation of a system according to the invention in a household;

[0173] Fig. 8 a system; and

[0174] Fig. 9 shows a flowchart of a process.

[0175] Fig. 1 shows a schematic representation of a household refrigeration appliance 1 according to the invention. The household refrigeration appliance 1 comprises a coolable interior 3, a door 5, a control device 20 and a wireless interface 21. The household refrigeration appliance 1 in Fig. 1 is exemplary designed as a refrigerator-freezer combination with two doors 5.

[0176] The interior compartment 3 is intended for storing food. The doors 5 serve to close the interior compartment 3. The control device 20 and the wireless interface 21 are arranged by way of example in an upper section of the household refrigerator 1 in Fig. 1. The control device 20 comprises a processing unit 23. In one embodiment, the interaction between the household refrigerator 1 or its actuators, the control device 20 and the wireless interface 21 can take place as described below in Fig. 14.

[0177] A separate external device 30 is positioned in the vicinity of the domestic refrigeration unit 1. The external device 30 also includes a wireless interface 31. The wireless interface 21 on the domestic refrigeration unit 1 is configured to establish a communication channel with the external device 30. To establish the communication channel, the wireless interface 21 sends a request signal to the external device 30. The external device 30 transmits 202503111

[0178] 31 / 91

[0179] Channel state information (schematically represented by a dashed arrow) is then collected. This information is gathered from the received packets at the wireless interface 21. A communication parameter is read from the channel state information and monitored by the control device 20. Based on this monitoring, it is determined whether a person (not shown) is within range of the household appliance. The person is identified based on a change in the communication parameter. Depending on the result of this identification, a function of the household appliance 1 is controlled by the control device 20.

[0180] An optional input device 9 is arranged on the front of the door 5 in Fig. 1. The input device 9 on the door 5 is designed as an operating device 9a. The input device 9 is arranged on the door 5 such that it can be operated by a person when the door is closed. The input device 9 is connected to the control device 20 in such a way that various functions for controlling the household refrigerator 1 can be activated or deactivated via this input device 9. Furthermore, the input device 9 or operating device 9a on the door 5, or any associated lighting, can be activated or deactivated by means of the control device 20 depending on the result of a determination by a person 40.

[0181] Figure 1 shows, by way of example, an input device 9, which is designed as an external user device 50. The external user device 50 also includes a wireless interface 51.

[0182] The external user device 50 can communicate with the control device 20 in such a way that various functions for controlling the domestic refrigeration appliance 1 can be activated or deactivated via the external user device 50. The control device 20 can optionally be configured such that controlling the function includes sending a notification 27 to the external user device 50.

[0183] Fig. 2 shows a schematic representation of the control device 20 according to the invention. The control device 20 is exemplary configured as a so-called system master of the household refrigeration appliance 1. In the embodiment shown in Fig. 2, the wireless interface 21 is arranged on the control device 20. The wireless 202503111

[0184] 32 / 91

[0185] Interface 21 is implemented on a circuit board of the control device 20. The processing unit 23 is implemented on the circuit board of the control device 20. In one embodiment, the control device 20 can be configured as described below in Fig. 14.

[0186] Fig. 3A shows a flowchart of the method according to the invention. The flowchart in Fig. 3A visualizes the basic method according to the invention, which runs independently of the controlled function. The method is divided into blocks S10 to S50. Depending on the embodiment, the individual blocks S10 to S50 may include further process steps, which are carried out within these blocks. The further process steps are to be selected according to the method of determining the person or according to the function to be controlled.

[0187] In block S10, the communication channel is established via the wireless interface 21 to the external device 30 in the area of ​​the household refrigeration appliance 1.

[0188] In block S20, channel status information from the communication channel is transmitted from the external device 30. This channel status information is collected at the wireless interface 21.

[0189] Block S30 monitors a communication parameter read from the channel status information.

[0190] In block S40, the presence of a person 40 within the area of ​​the household refrigeration appliance 1 is determined based on monitoring. The person is identified based on changes in the communication parameter. Depending on the configuration, the presence, absence, distance, identity, and / or gesture of the person is determined and output as the result.

[0191] In block S50, depending on the result of the person identification, a function of the household refrigeration appliance 1 is controlled by the control device 20. The selection of the function to be controlled can depend on the result of the identification. Furthermore, the selection of the function to be controlled can depend on whether the respective function is activated or deactivated. 202503111

[0192] 33 / 91

[0193] Figure 3B shows a schematic progression of channel state information during a person's absence 91 and during a person's presence 92. Figure 3B shows two schematic diagrams each, 91 and 92. The ordinate A1 is the amplitude of the channel state information (CSI). The abscissa A2 is the temporal progression with the increasing number of packets. Figure 3B uses schematic progressions to visually illustrate the differences in channel state information during monitoring and determination.

[0194] In diagrams 91 and 92, the course of the channel state information A1 during an absence 91 and during an absence 92 of a person 40 in the area of ​​the household refrigeration appliance 1 is plotted against time A2 for comparison. Diagram 91 shows the temporal course of the channel state information while a person's absence is being determined. Diagram 92 shows the temporal course of the channel state information while a person's presence 40 is being determined. A different course is evident in the comparison of diagrams 91 and 92 when a person is present. By evaluating a read communication parameter using a machine learning algorithm, the presence or absence of the person can be determined based on the change in the communication parameter.

[0195] Fig. 4 shows a schematic representation of the external user device 50. Depending on the embodiment, the external user device 50 serves as an input device 9. Fig. 4 shows an exemplary mobile application 53 opened on the external user device 50. In the application 53, various controllable functions A to F of the domestic refrigeration unit 1 are listed in a left column. A corresponding OFF / ON switch is assigned to each function in a right column. The various functions A to F can be activated or deactivated via the external user device 50. By way of example, in Fig. 4, functions A, B, D, and E are activated, and functions C and F are deactivated. The input is transmitted wirelessly to the domestic refrigeration unit 1 or the control device 20 via the wireless interface 51. The activation or deactivation of the control of individual functions is stored accordingly in the control device 20.

[0196] 34 / 91

[0197] Fig. 5A is a schematic representation of the household refrigeration unit 1 according to the invention, while the suppression of a door alarm 26 is being controlled. The household refrigeration unit 1 can, for example, be the household refrigeration unit 1 described in Fig. 1.

[0198] In Fig. 5A, person 40 is standing in front of the household refrigeration unit 1. Person 40 is positioned between the external device 30 and the wireless interface 21 of the household refrigeration unit 1. The door 5 is open and is detected by dedicated hardware (not shown). The method according to the invention determines the immediate presence of person 40 in front of the household refrigeration unit 1. As long as person 40's presence in front of the household refrigeration unit 1 is detected and the door 5 is open, the door alarm 26 is suppressed.

[0199] During the suppression of the door alarm 26, a predefined signal sequence 25 is output. An example acoustic signal sequence 25 is shown in Fig. 5A. The signal sequence 25 is selected depending on a certain number of times the person has been present and a time component. For example, after a first time component and the presence of person 40, a first signal 25a is output. After a second time component and the presence of person 40, a second signal 25b is output. After a third time component and the presence of person 40, a third signal 25c is output.

[0200] If the absence of person 40 in front of the household refrigeration unit 1 is determined for a predefined period and the door 5 is open, the door alarm 26 is activated.

[0201] Signals 25a to 25c and the door alarm 26 can be configured differently from each other.

[0202] In Fig. 5A, the household refrigeration appliance 5a further comprises a door opening device 11 and a lighting device 7. The lighting device 7 is arranged on the door 5 such that, even with the door closed, the light is perceptible to a person standing in front of the household refrigeration appliance 1.

[0203] 35 / 91

[0204] Fig. 5B shows a flowchart of the inventive method for controlling the suppression of the door alarm in a basic form with blocks B10 to 50.

[0205] In block B10, the opening of door 5 on the household refrigeration unit 1 is initially detected, as well as the presence of a person 40 in front of the household refrigeration unit 1. This condition marks the start of the flowchart.

[0206] In the subsequent block B20, a communication parameter is monitored such that, after a first predefined period, the identification of person 40 is available. This first predefined period corresponds to the period after which the door alarm 26 is normally activated when door 5 is open. The first predefined period can be any preset value. In this embodiment, the first predefined period is 90 seconds.

[0207] In the subsequent block B30, a decision is made based on the result of the determination of person 40. If, after the first predefined period, person 40 is found to be absent with door 5 open, the process continues with block B40. If, after the first predefined period, person 40 is found to be present with door 5 open, the process continues with block B50.

[0208] In block B40, door alarm 26 is activated.

[0209] In block B50, door alarm 26 is suppressed.

[0210] Fig. 5C shows a flowchart of the inventive method for controlling the suppression of the door alarm 26 and the output of a signal sequence 25. The sequence is represented by blocks C10 to C18.

[0211] The flowchart in Fig. 5C shows an embodiment of the previously described method in Fig. 5B. The process sequence shown and described in Fig. 5C can, for example, begin after the suppression of the door alarm 26 has started. In Fig.

[0212] 5B corresponds to block B50.202503111

[0213] 36 / 91

[0214] Block C10 is the starting block of the procedure. The presence of a person 40 in front of the household refrigeration unit 1 was detected, and the first predefined period has expired. The suppression of the door alarm 26 begins. The first predefined period corresponds to the first predefined period described in Fig. 5B.

[0215] At the beginning of the suppression of the door alarm 26, a first signal 25a of the signal sequence 25 is output in block C11 instead of the door alarm 26. The first signal 25a is, for example, a single beep.

[0216] In the subsequent block C12, the communication parameter is monitored such that, after a second predefined period, the identification of person 40 is available. This second predefined period is shorter than the first predefined period described above. In this embodiment, the second predefined period is 30 seconds, but it can be any preset value.

[0217] In the subsequent block C13, a decision is made based on the result of the determination of person 40. If, after the second predefined period, person 40 is found to be absent with door 5 open, the process continues with block C14. If, after the second predefined period, person 40 is found to be present with door 5 open, the process continues with block C15.

[0218] In block C14, door alarm 26 is activated.

[0219] In block C15, the identification of person 40 takes place again. The identification is carried out in the same way as in the previously described block C12.

[0220] In the subsequent block C16, a decision is made. The decision is made in the same way as in block C13. If, after the second predefined period, the absence of person 40 with door 5 open is determined, the process continues with block C14. If, after the second predefined period, the presence of person 40 with door 5 open is determined, the process continues with block C17. 202503111

[0221] 37 / 91

[0222] In block C17, a second signal 25b of the signal sequence 25 is output. The second signal 25b consists, for example, of two beeps.

[0223] Blocks C12 to C16 are then executed repeatedly. These blocks can be executed in a loop. The repetition of blocks C12 to C16 can occur multiple times until a third predefined period (maximum period) is reached. This third predefined period can be preset. Blocks C12 to C16 continue to be repeated until door 5 is closed again or until a door alarm 26 is activated according to block C14. The third predefined period can, for example, be 450 seconds. If the third predefined period is 450 seconds, blocks C12 to C16 are repeated six times. After each repetition, a signal from signal sequence 25 is output again in block C17. For example, another beep can be added after each repetition.

[0224] When the third predefined period is reached, the process continues from block C16 to block C18. In block C18, regardless of whether person 40 is present or absent, the door alarm 26 is activated if door 5 is open.

[0225] Fig. 6 shows a schematic representation of the household refrigeration appliance 1 according to the invention, while a lighting device 7 or an operating device 9a is controlled.

[0226] A person 40 approaches the household refrigeration unit 1. The person 40's location within the vicinity of the household refrigeration unit 1 is determined based on monitoring. In response to the detection of the approach, the lighting device 7 is activated as a function of the household refrigeration unit 1. The lighting device 7 is positioned on the door 5 such that, even with the door 5 closed, the light is perceptible to a person in front of the household refrigeration unit 1.

[0227] Alternatively or additionally, the control unit 9a is activated. For example, the backlight of a display and / or the backlight of the control buttons of the control unit 9a can be activated. Furthermore, a welcome message can be activated on the display of the control unit 9a. 202503111

[0228] 38 / 91

[0229] Fig. 7 shows a schematic representation of a system 80 according to the invention in a household 105. The system 80 comprises the household refrigeration unit 1 according to the invention, the external device 30 and further household appliances 60. The further household appliances 60 also each comprise a control device 20 with a processing unit 23 and a wireless interface 21.

[0230] The processing unit 23 of the other household appliances 60 also establishes a communication channel via the wireless interface 21 to the external unit 30. A communication parameter read from the channel status information is monitored. A result of the monitoring is transmitted to the household refrigeration unit 1 and / or to one of the household appliances 60. Based on the multiple monitoring results, a person 40 in the household 105 is identified. Based on this identification, one or more functions can be controlled. For example, functions of the different household appliances 1, 60 and / or other functions in the household 105 can be controlled.

[0231] System 80 can be used to monitor a person 40, particularly an elderly person 40, in a household 105. A time window is predefined via the monitoring person's external user device 50, during which the identification of the person 40 is to take place. For example, the predefined time window for identification is every day between 8:00 and 9:00 a.m. for two weeks. Within this time window, the identification of the person 40 is carried out using the method according to the invention. The identification of the person 40 can be repeated within this time window. Information about the result of the identification of the person 40 is sent as a notification 27 to the monitoring person's external user device 50.

[0232] Fig. 8 shows a system and Fig. 9 shows a flowchart of a process. The disclosures relating to Fig. 8 and Fig. 9 can be combined equally with the disclosures described above.

[0233] The first household appliance 110 can be the household refrigeration appliance 1 according to the invention. The second household appliance 115 can be the household refrigeration appliance 1 according to the invention. The control device 120 can be the control device 20202503111 according to the invention.

[0234] 39 / 91

[0235] System 100 can correspond to System 80 according to the invention. Processing device 125 can correspond to Processing device 23. Wireless interface 130 can correspond to Wireless interface 21. Wireless interface 140 can correspond to Wireless interface 31. Method 200 can correspond to the method according to the invention.

[0236] Fig. 8 shows an exemplary system 100 in a household 105. The system 100 comprises, by way of example, a first household appliance 110 and a second household appliance 115. For illustrative purposes, the first household appliance 110 is shown as a washing machine and the second household appliance 115 as an oven. It should be noted that the technology presented herein can also be implemented with only one household appliance 110, 115 or with more than two household appliances 110, 115.

[0237] A household appliance 110, 115 comprises a control device 120, which is preferably configured to control the associated household appliance 110, 115. A control device comprises a processing unit 125 and a wireless interface 130. The interface 130 can operate according to various radio technologies. However, it is preferred that the interface 130 operates according to one of the IEEE 802.11 (WLAN) standards. The external location 135 can correspond to the external device 30.

[0238] WLAN is defined in various frequency ranges, all of which are fundamentally suitable for the technology presented herein. For example, a frequency range of approximately 2.4 GHz, 5 GHz, or 6 GHz can be used. In another embodiment, interface 130 operates in a frequency range higher than approximately 10 GHz, preferably higher than approximately 50 GHz. An IEEE 802.11 standard that meets these requirements and is therefore well-suited for the present technology is IEEE 802.11bf. A higher frequency range that might be used for WLAN in the future is also suitable for the technology presented herein.

[0239] An interface 130 includes a transceiver configured to send and / or receive data. The transceiver enables a channel to be established between the control device 120 and another device – for example, a control device 120 of another household appliance 110, 115.

[0240] 40 / 91

[0241] Information can be transmitted via a channel. A channel can be established between one or more transceivers (1:n) or between a first and a second transceiver (1:1). In a simple embodiment, the transceiver comprises only a single antenna. However, a transceiver can also comprise multiple antennas that can be controlled to determine a directional characteristic. The transceiver can sample at least one communication parameter of the channel and provide it to the control device 120. The parameter can, in particular, include a signal-to-noise ratio (SNR), a signal strength, a signal phase, or a modulation parameter. The communication parameter is preferably read from the device via the Channel State Information (CSI). Preferably, the instantaneous Channel State Information, which indicates the current channel conditions, is evaluated.

[0242] Furthermore, a location 135 is provided in household 105, which is external to the household appliances 110, 115 and is also referred to herein as the external facility or external location 135. The external location 135 is preferably also installed in household 105 and may, in particular, comprise a gateway that is optionally connected to another communication network and through which a control device 120 can communicate with a location connected via that gateway. Such a location can manage and communicate with several household appliances 110, 115 of a household 105.

[0243] The external device 135 comprises a wireless interface 140 configured to communicate with a wireless interface 130 of a control device. The transceivers of interfaces 130 and 140 can essentially be of the same design. In one embodiment, the transceiver of the external device 125 can acquire a communication parameter of a channel. The communication parameter can be forwarded to a control device 120, preferably via the communication channel.

[0244] A person 40; 145, located in the household 105, can move within the area of ​​the household appliances 110, 115 and the external location 135. In one embodiment, the person 40; 145 is in direct contact between a transmitter and a receiver, such that the electromagnetic signals are influenced by the body of the person 40; 145. The person 40; 145 can also be in the 202503111

[0245] 41 / 91

[0246] The Fresnel zone between two transceivers of participants 110, 115, 135 of a communication channel is considered. In another embodiment, person 40; 145 is located outside the line of sight between transmitter and receiver, and their body reflects the signals. In both cases, a communication parameter of the channel can be changed, and this change can be detected by the receiving transceiver.

[0247] Observed changes in a communication parameter can be evaluated at a control device 120 associated with the transceiver. Preferably, observations from several transceivers are aggregated and processed together at a single location 110, 115, 135. A pattern in which one or more communication parameters of a channel have been influenced can be detected. A characteristic of the person 40; 145 can be assigned to the pattern, in particular a size, mass, weight, distance, orientation, or physical feature that allows conclusions to be drawn about their identity. Only a predetermined number of persons 40; 145 can regularly be present in the household 105, so it is sufficient to distinguish these persons 40; 145 from one another. In a further embodiment, a gesture performed by the person 145 can be recognized based on an observed pattern.Such a gesture could, for example, include a vertical, horizontal, or circular movement of the hand.

[0248] A household appliance 110, 115 can be controlled based on a recognized characteristic of person 145. For example, a stove could be switched to an energy-saving mode if no person 40; 145 is within a predetermined area. This can mitigate the risk of fire from overheated cookware. Other control methods are also possible. It is also possible to control several household appliances 110, 115 in a coordinated manner based on observations.

[0249] Fig. 9 shows a flowchart of a method 200 for controlling one or more household appliances 1, 110, 115. The illustration is based on the example of controlling the two household appliances 110, 115 shown in Fig. 8, although the method 200 can also be carried out with only one or more than two household appliances 110, 115. Steps related to the first household appliance 110 are shown in the left-hand section, and steps related to the second appliance 110 are shown in the right-hand section.

[0250] 42 / 91

[0251] The second household appliance 115 is assigned to it. Steps with the same reference numbers are performed identically on all household appliances 110 and 115. A step shown in the middle can be performed by one or more of the household appliances 110 and 115.

[0252] In step 205, a household appliance 110, 115 establishes a communication channel to another device, in particular to another household appliance 110, 115 or to an external location 135, using its wireless interface 130. In another embodiment, signals from another device are only received without establishing a communication channel over which information is transmitted in a controlled manner. For example, unidirectional or bidirectional communication between other devices can be observed. Transmitted content does not need to be decoded. In a further embodiment, signals from another device indicating its readiness to communicate (advertising) can be received.

[0253] In step 210, status information is determined about the communication channel. This status information relates to a communication parameter of the communication link. Typically, multiple status updates are determined for several communication parameters. Furthermore, status information can be collected over a predetermined period.

[0254] In step 215, it is determined whether and how the status of the communication parameter has changed during the period under consideration. Steps 210 and 215 are preferably executed repeatedly in a loop. In one embodiment, a data transfer can be specifically initiated between a control device 120 of a household appliance 110, 115 and its communication partner in order to sample improved status information. The data transfer can, for example, include a simple request for a response ("ping").

[0255] In step 220, status information from several household appliances 110, 115 can be compared. For this purpose, the status information can be collected from one of the appliances 110, 115. Status information can be transmitted from one household appliance 110 to another 115 via the existing communication link, either directly or indirectly via the external point 135. Collected 202503111

[0256] 43 / 91

[0257] Information, processed or unprocessed, can be distributed between the household appliances 110, 115.

[0258] In step 225, based on the observations, a person 40; 145 can be identified in the vicinity of one of the household appliances 110, 115 or the external location 135. For this purpose, a predetermined pattern in observed changes in status information can be recognized. This recognition can be achieved, for example, heuristically, statistically, or using artificial intelligence methods or a machine learning algorithm.

[0259] In step 230, a household appliance 110, 115 can be controlled with respect to the detected person 120 or a detected characteristic of the person 120.

[0260] The inventive method, control device, and system are not limited to use in controlling a household refrigerator. Use in other household appliances is also conceivable, such as an oven, cooktop, extractor hood, dishwasher, washing machine, tumble dryer, coffee maker, or other household appliance.

[0261] The following further disclosure relates to the determination or identification of a person within the vicinity of the household appliance using wireless signals. It specifies, among other things, particularly at a protocol level, how the determination or identification of a person can be carried out. For the determination of the person, the method according to the invention can partially or completely incorporate or utilize specific structural or functional features of the further disclosure. Persons skilled in the art will understand that various features illustrated or described with respect to one of the figures can be combined with features illustrated in one or more other figures to create embodiments that are not explicitly illustrated or described, particularly with respect to features in the further disclosure.

[0262] The following further disclosure describes advantageous embodiments and further details of the present invention. The person skilled in the art understands that the invention described herein is

[0263] 44 / 91

[0264] The components and terms described below correspond to or further specify the components and terms introduced previously. In particular, the “system” (e.g., 502, 1000, 2000) or “computer system” described below is to be understood as a specific embodiment of the control device (20). The “wireless receiver” (e.g., 304, 608) is an embodiment of the wireless interface (21), and the “computer system” (e.g., 2002) is an embodiment of the processing device (23). The term “user” is used synonymously with “person (40)”. The process of “person identification” (PID) represents a specific embodiment of the step of “determining a person (40)”.

[0265] The term "person identification" (PID) encompasses determining a person's identity. "Determining the person" can include determining the person's identity. For example, the presence of a person can first be established. In one embodiment, determining the presence of the person or user can be understood as differentiating whether the person is standing directly in front of the appliance or is located in a predefined area surrounding the appliance. In an additional step, the person's identity can then be determined.

[0266] The further disclosure relates to determining the presence of a user of a machine and, in particular, to identifying a person using wireless signals. The machine may be a household appliance. The machine may be a household refrigeration appliance.

[0267] Human presence detection refers to the detection of a human body within a defined area of ​​interest. This area of ​​interest could be the area of ​​a household appliance, particularly a household refrigerator. Person identification (PID) is the process of recognizing and verifying a person's identity. Applications in this area include security and access control for authentication in various domains. Techniques such as facial recognition, fingerprint reading, voice authentication, and retinal verification offer a biometric approach to using biometric data to verify a person's identity.

[0268] 45 / 91

[0269] Such methods utilize physical interactions between the system and the person to be identified.

[0270] Starting from the state of the art, it is the task of further disclosure to provide an advantageous method and / or system for determining the presence of a user of a machine.

[0271] According to one aspect, a method for determining the presence of a user (corresponding to person (40)) of a household appliance, in particular a household refrigeration appliance (1), using wireless signals is disclosed. The method comprises collecting, on a wireless receiver (corresponding to wireless interface (21)), channel state information from packets received from a wireless transmitter (corresponding to external device (30)); preprocessing, using a computer system associated with the wireless receiver (corresponding to processing device (23)), the channel state information; and selecting segments of the channel state information.Determining (corresponding to the step of determining a person (40)), using the computer system and one or more parameters of the selected segments of the channel state information, a user's presence based on additional packets received by the wireless receiver; and controlling (corresponding to the step of controlling a function), by the computer system, an operation of the machine in response to the detection of the user's presence.

[0272] By determining user presence using channel state information gathered by the wireless receiver, no dedicated sensor hardware is required to detect that a user is nearby. This allows for a very cost-effective deployment of presence detection functionality.

[0273] In some embodiments, a personal identity of the user is determined, and the preprocessing of the channel state information preferably includes determining a gait pattern and a variety of biometric features, including body shape and size, associated with the user, based on 202503111

[0274] 46 / 91

[0275] the selected segments of the channel state information. This means that by thoroughly analyzing the channel state information in the manner described, it may even be possible to identify the user.

[0276] In some embodiments, the method comprises the further steps of annotating, using the computer system, the selected segments of the channel state information with a class indicative of personal identity based on at least the user's gait pattern; and determining, using a machine learning model, the user based on at least the user's gait pattern, wherein the machine learning model is trained using classifier training with training data comprising information from the selected segments of the channel state information and the user's gait pattern, wherein the classifier training includes determining one or more parameters of the selected segments of the channel state information. The machine learning model may be, or comprise, a machine learning algorithm.It is preferred that the step of determining the presence of the user based on additional packets received by the wireless receiver be performed after the classifier training has been carried out.

[0277] The classes can include, for example, approaching, standing in front of, or moving away from the wireless receiver or transmitter. In this way, for instance, a coffee machine's display can be switched off to save energy when no one is present, while the display switches on when a user approaches. Similarly, the lighting and / or control panel of a household refrigerator can be controlled. If the user's identity is known to the system, the user can be given a personalized coffee recommendation. Furthermore, lights in a room can be switched off when no one is present and switched on when a user enters the room. The temperature of a room can also be controlled based on the presence of a user.

[0278] Using the wireless signal, it may also be possible to determine if the user is a child and potentially activate a parental control function. (202503111)

[0279] 47 / 91

[0280] Intuition suggests that children typically have a smaller body size than adults. Consequently, the reflections of gait, muscle, and fat tissue will differ between a child and an adult. To address this use case, classes such as "child," "adult," or "empty space" can be used during annotation.

[0281] In some embodiments, the method further comprises the steps of determining one or more usage patterns of the machine by the user, the machine being located in a space that includes the wireless receiver and the wireless transmitter; and using the one or more usage patterns to control the operation of the machine. This can make it possible to customize the user interface of household appliances. For example, personalized recommendations can be given to the user. Controlling the operation of the machine can include controlling one or more functions of the household refrigeration appliance.

[0282] In some embodiments, the preprocessing of the channel state information includes the use of amplitude information from the received packets to determine the gait pattern and preferably one or more of the user's multiple biometric features.

[0283] In some embodiments, the preprocessing of the channel state information includes the use of phase information from the received packets to determine the gait pattern and preferably one or more of the user's multiple biometric characteristics.

[0284] In some embodiments, the preprocessing of the channel state information includes the use of Doppler frequency shift information from the received packets to determine the gait pattern and motion signature of the user.

[0285] In some embodiments, the classifier training of the machine learning model involves the use of a two-dimensional convolutional neural network to train one or more

[0286] 48 / 91

[0287] To determine parameters of the selected segments of the channel state information that identify the user.

[0288] In some embodiments, the classifier training of the machine learning model includes extracting a variety of time-domain features to determine the variability of wireless signals in the selected segments of the channel state information; and extracting a variety of frequency-domain features to determine the spectral bandwidth, spectral flatness, and peak frequency of the wireless signals in the selected segments of the channel state information, including subcarrier correlations. The classifier training further involves the use of a sequence model with a bidirectional gated recurrent unit (BiGRU) incorporating an attention mechanism and a transformer.

[0289] In some embodiments, the classifier training of the machine learning model involves the use of a sequence model to determine a temporal movement pattern of the user.

[0290] In some embodiments, the machine is a household appliance, in particular a household refrigeration appliance, wherein controlling the operation of the machine may include learning one or more user preferences.

[0291] According to a further aspect of the extended disclosure, a system for determining the presence of a user (40) of a machine (1), in particular a household appliance (1), using wireless signals is described. This system can be understood as an embodiment of the control device (20). The system comprises a wireless receiver configured to receive packets transmitted by a wireless transmitter (an embodiment of the external device 30) and further configured to collect channel state information from the packets. The system further comprises a computer system associated with the wireless receiver (an embodiment of the processing device 23). The computer system is configured to preprocess the channel state information; select segments of the channel state information; and determine the presence of the user based on the channel state information.

[0292] 49 / 91

[0293] to determine the presence of additional packets received by the wireless receiver using one or more parameters of the selected segments of the channel state information; and to control the operation of the machine in response to the detection of the user's presence. The system described in connection with the further disclosure may correspond to or comprise the previously described control device and wireless interface. The functionality of this system may be implemented by the previously described control device and wireless interface.

[0294] In some embodiments of the system, the computer system is further configured to determine a personal identity of the user, wherein the preprocessing of the channel state information preferably includes determining a gait pattern and a variety of biometric features, including body shape and size, associated with the user, based on the selected segments of the channel state information.

[0295] In some embodiments, the computer system is further configured to annotate the selected segments of the channel state information with a class indicative of personal identity based on at least the user's gait pattern; and to determine the user based on at least the user's gait pattern using a machine learning model, wherein the machine learning model is trained using classifier training with training data comprising information from the selected segments of the channel state information and the user's gait pattern, wherein performing the classifier training includes determining one or more parameters of the selected segments of the channel state information. It is preferred that the user's presence is determined after performing the classifier training.

[0296] According to another aspect of the further disclosure, a non-transitory, computer-readable medium is described that stores instructions which, when executed by a computer system, cause the computer system to perform operations. These operations include preprocessing.

[0297] 50 / 91

[0298] from channel state information from packets received by a wireless receiver associated with the computer system; the selection of segments of the channel state information; the determination, using one or more parameters of the selected segments of the channel state information, of the presence of the user based on additional packets received by the wireless receiver; and the control of an operation of the machine in response to the detection of the presence of the user.

[0299] Features and advantages described in relation to one claim category can be applied equally to all other claim categories. This means, for example, that the disclosure should include corresponding embodiments of the system for all embodiments of the method, including means for carrying out the steps of the respective method.

[0300] Furthermore, systems and methods for personal identification using wireless signals are disclosed. In one embodiment, a personal identification method includes collecting, at a wireless receiver, channel state information from received packets transmitted by a wireless transmitter. The method can preprocess the channel state information using a computer system associated with the wireless receiver. This preprocessing of the channel state information can determine a gait pattern and other biometric characteristics associated with a specific person (e.g., the machine user), such as body shape, height, etc.The method comprises, based on selected segments of the channel state information, annotating, using the computer system, the selected segments of the channel state information with a class that is indicative based on at least the gait pattern of the identified person. The method may further comprise determining, using a machine learning model, the identified person based on at least the gait pattern of the identified person and other biometric characteristics including body shape, height, etc., wherein the machine learning model is trained using classifier training and training data comprising information from the selected segments of the channel state information and the gait pattern of the identified person, and wherein the execution of the classifier training

[0301] 51 / 91

[0302] The method includes determining one or more parameters of the selected segments of the channel state information, and determining, using the computer system and the machine learning model, one or more usage patterns of a machine by the specific person, the machine being located in the space containing the wireless receiver and the wireless transmitter. For example, after classifier training, the method may include determining, using the computer system and one or more parameters of the selected segments of the channel state information, the presence of the specific person based on additional packets received by the wireless receiver after the classifier training has been carried out, and controlling, by the computer system and using one or more usage patterns, the operation of the machine in response to the detection of the specific person's presence.

[0303] A second embodiment relates to a system for determining a specific personal identity using wireless signals. The system includes a wireless receiver configured to receive packets transmitted by a wireless transmitter and may further be configured to collect channel state information from the packets, as well as a computer system associated with the wireless receiver. The computer system may be configured to preprocess the channel state information, with the preprocessing of the channel state information enabling the determination of a gait pattern and other biometric characteristics such as body shape, height, etc., which are associated with a specific person, based on selected segments of the channel state information, includes annotating the selected segments of the channel state information with a class that is indicative based on at least the gait pattern of the specific person and other biometric characteristics such as body shape, size, etc., using a machine learning model to determine the specific person based on at least the gait pattern of the specific person and other biometric characteristics such as body shape, size, etc., wherein the machine learning model is trained using classifier training and training data that includes information from the selected segments of the channel state information and the gait pattern of the specific person, and wherein the computer system may be configured to assist in performing the 202503111.

[0304] 52 / 91

[0305] The classifier training process determines one or more parameters of the selected segments of the channel state information and, using the machine learning model, identifies one or more usage patterns of a machine by a specific individual, where the machine is located in a space encompassing the wireless receiver and transmitter. Subsequently, the computer system can be configured to determine the presence of the specific individual based on additional packets received by the wireless receiver after the classifier training has been performed, using the one or more parameters of the selected channel state information segments, and to control the operation of the machine in response to the detection of the specific individual's presence, using the one or more usage patterns.

[0306] A third embodiment relates to a non-transitory, computer-readable medium that stores instructions which, when executed by a computer system, cause the computer system to perform various operations. These operations include preprocessing channel state information from packets received by a wireless receiver associated with the computer system, with the preprocessing of the channel state information enabling the determination of a gait pattern and other biometric characteristics such as body shape, height, etc., which may include being associated with a specific person based on selected segments of the channel state information, annotating the selected segments of the channel state information with a class that is indicative based on at least the gait pattern of the specific person, and determining, using a machine learning model, the specific person based on at least the gait pattern of the specific person and other biometric characteristics such as body shape, size, etc., wherein the machine learning model is trained using classifier training and training data comprising information from the selected segments of the channel state information and the gait pattern of the specific person, and wherein performing the classifier training includes determining one or more parameters of the selected segments of the channel state information and determining, using the machine learning model, one or more usage patterns of a machine by the specific person, wherein the machine is located in a room which202503111.

[0307] 53 / 91

[0308] The operation includes the wireless receiver and a wireless transmitter from which the packets were transmitted. The operations may further include determining, using one or more parameters of the selected segments of the channel state information, the presence of a specific person based on additional packets received by the wireless receiver after the classifier training has been completed, and controlling, using one or more usage patterns, the operation of the machine in response to the detection of the presence of that specific person.

[0309] FIG. 10 shows a system 1000 for training a neural network, as it can be used as part of the machine learning algorithm in the control device (20).

[0310] FIG. 11 shows a computer-implemented method 2000 for training such a neural network.

[0311] FIG. 12A shows an example of an embodiment of a system that uses information from wireless signals for personal identification (PID). This system represents an embodiment of the control device (20) implemented in a household refrigeration appliance (1).

[0312] FIG. 12B shows an example of information that can be extracted from wireless signals for use in an embodiment of a method for determining PID. This information is an example of the channel state information collected by the wireless interface (21).

[0313] FIG. 12C shows a workflow diagram for an embodiment of a method according to the further disclosure. This method is a detailed embodiment of the method according to the invention.

[0314] FIG. 13A shows a workflow diagram and signal characteristics for the preprocessing of the amplitude of channel state information, as can be performed by the control device (20) or the processing device (23).

[0315] FIG. 13B shows a workflow diagram and signal characteristics for the preprocessing of the channel state information phase, as can be performed by the control device (20) and the processing device (23), respectively.

[0316] 54 / 91

[0317] FIG. 13C shows a neural network which, in an embodiment of the further disclosure, can be used as part of the control device (20) or the processing device (23).

[0318] FIG. 13D is a workflow diagram illustrating an embodiment of a method for performing inference according to the further disclosure, which can be executed by the control device (20) or the processing device (23).

[0319] FIG. 14 shows a schematic diagram of an interaction between a computer-controlled machine 510 (an embodiment of the household refrigeration appliance 1) and a control system 512 (an embodiment of the control device 20).

[0320] FIG. 15 illustrates an application of an embodiment of the inventive method in which a household refrigeration appliance (1) recognizes a person (40).

[0321] Embodiments of the further disclosure are described herein. However, it should be understood that the disclosed embodiments are merely examples and that other embodiments may take different and alternative forms. The figures are not necessarily to scale; some features may be exaggerated or minimized to show details of certain components. Therefore, specific structural and functional details disclosed herein are not to be interpreted as limiting, but merely as a representative basis to teach a person skilled in the art how to apply the embodiments in a variety of ways. As those skilled in the art will understand, various features illustrated and described with respect to one of the figures can be combined with features illustrated in one or more other figures to generate embodiments that are not explicitly illustrated or described.The combinations of features illustrated represent representative embodiments for typical applications. However, different combinations and modifications of the features, consistent with the teachings of this further disclosure, might be desirable for certain applications or implementations.

[0322] “Ein”, “eine” and “der / die / das”, as used herein, refer to both the singular and the plural unless the context clearly states otherwise.

[0323] 55 / 91

[0324] Others are different. For example, “a processor” programmed to perform various functions refers to a processor programmed to perform each individual function, or to more than one processor programmed together to perform each of the different functions.

[0325] Personal identification, or PID, is the process of uniquely identifying and verifying an individual's identity. It is a field with applications ranging from security and access control to authentication across various domains. Typically, PID involves using various data sources, such as biometrics, facial and audio data, facial recognition, fingerprint reading, voice authentication, and oculo-retinal verification technologies, to positively identify an individual. While these technologies offer a robust approach to verifying human identity, they can significantly infringe on human privacy. The risk of data theft, misuse, and spoofing remains a concern. Users must also accept some inconvenience while interacting with authentication systems by being in close proximity to them.

[0326] Compared to these methods, wireless signal-based personal identification (PID) systems are seamlessly convenient yet secure. These systems operate over a wide area by utilizing multipath radio signals that propagate in all directions. This allows for a high degree of freedom and signal coverage for human identification, even if the user is not within a small detection zone. Furthermore, wireless signal-based PID systems can protect privacy because the signals, which contain identifiable signatures, are not affected by the environment and can be confined to a specific detection zone.While traditional identifiable information such as fingerprint data remains static after capture (which can increase the risk of theft), wireless signals vary across spatiotemporal domains, meaning that wireless biometric data may not be useful for identifying the same person in a different environment. Thus, wireless signal-based identification systems can be very convenient and secure.

[0327] 56 / 91

[0328] Wireless PID systems can also be useful for personalized content delivery. Current systems may not be able to automatically detect human presence and identify a nearby individual, instead delivering content either in response to proactive human requests or based on time-of-day-based, event-driven systems. While such a system can be useful for certain daily routines, a change in a routine can render the system unusable for that user. However, the use of wireless PID, as described in the further disclosure for intelligent assistance systems, can enable the detection and identification of human presence to deliver personalized content once a person has been identified.For example, a smart home appliance, such as a smart coffee maker, could automatically detect the presence of a specific person and dispense coffee according to their personal preferences. In another example, a smart thermostat system could set the desired temperature based on the presence of a specific person. In other examples, the various functions described above could be controlled within a household refrigerator. Accordingly, this mechanism can enable PID-based learning for household appliances, particularly learning-based household refrigerators. Beyond household appliances, wireless PID can enable the use of smart machines for, for example, industrial purposes or in other workplaces, and a variety of other applications where a machine can be controlled based on personal usage patterns and / or personal preferences.

[0329] In addition to the above, the wireless PID systems of this further disclosure can also provide a rough identification and perform functions based thereon. For example, according to the further disclosure, a wireless PID system can determine whether a detected person is an adult or a child and, based on this, restrict access to certain household appliances (e.g., a stove, an oven, a household refrigerator, etc.).

[0330] The wireless PID systems and methods of this further disclosure are privacy-friendly and non-invasive and can be implemented in a variety of embodiments. In one embodiment, the method can utilize an existing Wi-Fi transmitter (e.g., a home Wi-Fi router) and a 202503111

[0331] 57 / 91

[0332] Utilize a wireless receiver configured to receive wireless signals transmitted by the transmitter. The wireless receiver can be integrated, for example, into a household appliance or a security system and continuously monitor wireless signals transmitted by the transmitter. These wireless signals can be reflected and / or partially absorbed by a person near a path between the transmitter and the receiver. By receiving the wireless signals, the receiver can gather information about gait patterns and other movement signatures (e.g., walking patterns). In particular, since human gait patterns can be unique to each individual due to differences in body shape, muscle, and fat tissue, the systems and methods described in this further disclosure regarding the use of a neural network can learn to identify individuals.The neural network can learn other biometric features such as height, body shape, etc. Furthermore, the systems and methods disclosed herein can recognize activities and their patterns for identified individuals. Using artificial intelligence and neural networks, the systems and methods disclosed herein can use Wi-Fi channel state information features for different individuals to solve a multi-class classification task. The methodology can also be extended to other wireless technologies such as Bluetooth, Ultra Wideband (UWB), LTE, 5G, 6G, etc., and is therefore not limited to Wi-Fi implementations. It can also operate with multiple antennas.

[0333] In various embodiments, the workflow for an embodiment of a method according to the further disclosure comprises 1) wireless acquisition and data collection; 2) data preprocessing; 3) an annotation process; 4) classifier training; 5) inference; and 6) updating. Several non-limiting embodiments are now discussed in more detail below.

[0334] FIG. 10 shows a System 1000 for training a neural network, e.g., a deep neural network. The neural network or deep neural networks shown and described are merely examples of the types of machine learning networks or neural networks that can be used. The System 1000 may include an input interface for accessing training data 1002 for the neural network. For example, as shown in FIG. 10, the 202503111

[0335] 58 / 91

[0336] The input interface is formed by a data storage interface 1004, which can access the training data 1002 from a data storage device 1006. For example, the data storage interface 1004 can be a storage interface or a persistent storage interface, e.g., a hard disk or SSD interface, but also a personal, local, or wide-area network interface such as a Bluetooth, Zigbee, or Wi-Fi interface, or an Ethernet or fiber optic interface. The data storage device 1006 can be internal data storage of the system 1000, such as a hard disk or SSD, but also external data storage, e.g., network-accessible data storage.

[0337] In some embodiments, the data store 1006 may further comprise a data representation 1008 of an untrained version of the neural network, which the system 1000 can access from the data store 1006. However, it is recognized that the training data 1002 and the data representation 1008 of the untrained neural network may each also be accessed from a different data store, e.g., via another subsystem of the data store interface 1004. Each subsystem may be of a type such as that described above for the data store interface 1004. In other embodiments, the data representation 1008 of the untrained neural network may be generated internally by the system 1000 based on design parameters for the neural network and therefore may not be explicitly stored on the data store 1006.System 1000 may further include a processor subsystem 1010, which may be configured to provide an iterative function during the operation of System 1000 as a substitute for a stack of layers of the neural network to be trained. Here, each layer of the stack of layers to be replaced may have mutually shared weights and receive as input an output from a previous layer or, for a first layer of the stack of layers, an initial activation and a portion of the stack of layers' input. Processor subsystem 1010 may also be configured to iteratively train the neural network using the training data 1002. Here, an iteration of the training by processor subsystem 1010 may include a forward propagation part and a backward propagation part.The processor subsystem 1010 can be configured to perform the forward propagation part by, among other things, performing operations that define the forward propagation part.202503111.

[0338] 59 / 91

[0339] The system 1000 can determine an equilibrium point of the iterative function at which the iterative function converges to a fixed point. Determining the equilibrium point involves using a numerical root-finding algorithm to find a square root solution for the iterative function minus its input, and providing the equilibrium point as a substitute for an output of the stack of layers in the neural network. The system 1000 can further include an output interface for outputting a data representation 1012 of the trained neural network; this data can also be referred to as trained model data 1012. For example, as also shown in FIG. 10, the output interface can be formed by the data storage interface 1004, which in these embodiments is an input / output ('IO') interface through which the trained model data 1012 can be stored in the data storage 1006.For example, the data representation 1008, which defines the 'untrained' neural network, can be at least partially replaced during or after training by the data representation 1012 of the trained neural network by adjusting the neural network parameters, such as weights, hyperparameters, and other types of neural network parameters, to reflect the training on the training data 1002. This is also illustrated in FIG. 10 by the reference numerals 1008 and 1012, which refer to the same data set in data storage 1006. In other embodiments, the data representation 1012 can be stored separately from the data representation 1008, which defines the 'untrained' neural network. In some embodiments, the output interface can be separate from the data storage interface 1004, but can generally be of the type described above for the data storage interface 1004.

[0340] In various embodiments, the system for training a neural network can be implemented within a personal identification system using wireless signals (e.g., Wi-Fi) received by a wireless receiver. The data obtained from the wireless signals (e.g., CSI) can be used to determine gait patterns and / or other movement characteristics to identify a specific individual. Embodiments in which the data can also be used for a cruder identification (e.g., to distinguish between an adult and a child) using the neural network training system are described in 202503111.

[0341] 60 / 91

[0342] This is also possible and intended. The system can be implemented in a household appliance, especially a household refrigerator, an industrial plant, or any other suitable environment. Based on the training, the neural network can be used to determine the patterns of a specific person and adapt its operation accordingly.

[0343] Figure 11 depicts a System 2000 for implementing the machine learning models described herein, for example, the deep neural networks used to perform personal identification using data from received wireless signals, as described in more detail above and below. Other types of machine learning models may be used, and the DNNs described herein are not the only types of machine learning models that may be used in the System of this Disclosure. For example, if the input image contains an ordered sequence of pixels after conversion of CSI values ​​to pixels in an image, a CNN may be used. The System 2000 may be implemented to perform one or more of the image recognition stages described herein. The System 2000 may include at least one Computer System 2002.The Computer System 2002 may include at least one Processor 2004 operationally connected to a Memory Unit 2008. The Processor 2004 may include one or more integrated circuits implementing the functionality of a Central Processing Unit (CPU) 2006. The CPU 2006 may be a commercially available processing unit implementing an instruction set such as one of the x86, ARM, Power, or MIPS instruction set families. During operation, the CPU 2006 may execute stored program instructions retrieved from the Memory Unit 2008. The stored program instructions may include software that controls the operation of the CPU 2006 to perform the operation described herein.In some examples, the Processor 2004 can be a system-on-a-chip (SoC) that integrates the functionality of the CPU 2006, the Memory 2008, a network interface, and input / output interfaces into a single integrated device. The Computer System 2002 can implement an operating system to manage various aspects of its operation. While one Processor 2004, one CPU 2006, and one Memory 2008 are shown in FIG. 11, of course, more than one of each can be used in a complete system.

[0344] 61 / 91

[0345] The 2008 memory unit can include volatile and non-volatile memory for storing instructions and data. Non-volatile memory can include solid-state storage such as NAND flash memory, magnetic and optical storage media, or any other suitable data storage device that retains data when the 2002 computer system is turned off or loses power. Volatile memory can include static and dynamic random-access memory (RAM) that stores program instructions and data. For example, the 2008 memory unit can store a 2010 machine learning model or algorithm, a 2012 training dataset for the 2010 machine learning model, and a 2016 raw source dataset.

[0346] The 2002 computer system can include a 2022 network interface device configured to provide communication with external systems and devices. For example, the 2022 network interface device can include a wired and / or wireless Ethernet interface as defined by the IEEE 802.11 standard family. The 2022 network interface device can include a cellular communication interface for communication with a cellular network (e.g., 3G, 4G, 5G). The 2022 network interface device can also be configured to provide a communication interface to an external 2024 network or a cloud.

[0347] The external network of 2024 can be referred to as the World Wide Web or the Internet. The external network of 2024 can establish a standard communication protocol between computer devices. The external network of 2024 can facilitate the easy exchange of information and data between computer devices and networks. One or more servers of 2030 can communicate with the external network of 2024.

[0348] The Computer System 2002 can include an Input / Output (I / O) Interface 2020, which can be configured to provide digital and / or analog inputs and outputs. The I / O Interface 2020 is used to transfer information between internal memory and external input and / or output devices (e.g., HMI devices). The I / O Interface 2020 can include associated circuitry or BUS-202503111

[0349] 62 / 91

[0350] Networks are used to transfer information to or between the processor(s) and memory. For example, the I / O interface 2020 may include digital I / O logic lines that can be read from or set by the processor(s), handshake lines to monitor data transmission over the I / O lines, timing and counter devices, and other structures known to provide such functions. Examples of input devices include a keyboard, mouse, sensors, etc. Examples of output devices include monitors, printers, speakers, etc. The I / O interface 2020 may include additional serial interfaces for communication with external devices (e.g., a Universal Serial Bus (USB) interface).The I / O interface 2020 can be described as an input interface (by transmitting data from an external input, such as a sensor) or as an output interface (by transmitting data to an external output, such as a display).

[0351] The Computer System 2002 may include a Human-Machine Interface (HMI) Device 2018, which may include any device that enables the System 2000 to receive control inputs. Examples of input devices may include human-interface inputs such as keyboards, mice, touchscreens, speech input devices, and other similar devices. The Computer System 2002 may include a Display Device 2032. The Computer System 2002 may include hardware and software for outputting graphic and text information to the Display Device 2032. The Display Device 2032 may include an electronic screen, a projector, a printer, or other suitable equipment for displaying information to a user or operator. The Computer System 2002 may also be configured to enable interaction with remote HMIs and remote Display Devices via the Network Interface Device 2022.

[0352] System 2000 can be implemented using one or more computer systems. While the example represents a single Computer System 2002 implementing all the described features, it is intended that various features and functions can be implemented separately and by multiple computer units communicating with each other. The specific system architecture chosen can depend on a variety of factors.

[0353] 63 / 91

[0354] System 2000 can implement a machine learning algorithm 2010 configured to analyze the raw source dataset 2016. The raw source dataset 2016 can include raw or unprocessed sensor data that may be representative of an input dataset for a machine learning system. The raw source dataset 2016 can include video, video segments, images, text-based information, audio or human speech, time-series data (e.g., a pressure sensor signal over time), raw or partially processed sensor data (e.g., a radar map of objects), and wireless signals related to CSI, RSSI, and CIR.Furthermore, the raw source dataset in 2016 can be input data derived from an associated sensor such as a camera, lidar, radar, ultrasonic sensor, motion sensor, thermal imaging camera, wireless receiver, or any other type of sensor that generates associated data with spatial dimensions where there is a notion of a "foreground" and a "background" within those spatial dimensions. References to an input or input "image" here do not necessarily originate from a camera but can come from any of the sensors listed above. In some examples, the machine learning algorithm in 2010 can be a neural network algorithm (e.g., a deep neural network) designed to perform a predetermined function.

[0355] The System 2000, or computer system, can store a training dataset 2012 for the machine learning algorithm 2010. The training dataset 2012 can represent a set of previously constructed data for training the machine learning algorithm 2010. The training dataset 2012 can be used by the machine learning algorithm 2010 to learn weighting factors associated with a neural network algorithm. The training dataset 2012 can include a set of source data containing corresponding outcomes or results that the machine learning algorithm 2010 attempts to replicate through the learning process.

[0356] The 2010 machine learning algorithm can be run in a learning mode using the 2012 training dataset as input. The 2010 machine learning algorithm can be executed for a number of iterations using the data from the 2012 training dataset. With each iteration, the 2010 machine learning algorithm can apply internal weighting factors based on the 202503111

[0357] 64 / 91

[0358] Update the results obtained. For example, the 2010 machine learning algorithm can compare output results (e.g., a reconstructed or augmented image if image data is the input) with those contained in the 2012 training dataset. Since the 2012 training dataset contains the expected results, the 2010 machine learning algorithm can determine when performance is acceptable. After the 2010 machine learning algorithm has reached a predetermined level of performance (e.g., 100% agreement with the results associated with the 2012 training dataset) or convergence, the 2010 machine learning algorithm can be run using data not contained in the 2012 training dataset. It should be understood that in this further disclosure, “convergence” may mean that a specified (e.g., predetermined) number of iterations have occurred, or that the residual is sufficiently small (e.g.,The change in the approximate probability across iterations (change by less than a threshold value), or other convergence conditions. The trained 2010 machine learning algorithm can be applied to new datasets to generate annotated data.

[0359] The 2010 machine learning algorithm can be configured to identify a specific feature in the 2016 raw source data. The 2016 raw source data can comprise a variety of instances or a single input dataset for which supplemental results are desired. For example, the 2010 machine learning algorithm can be configured to identify or determine the presence of a user. The 2010 machine learning algorithm can be programmed to process the 2016 raw source data to identify the presence of specific features. The 2010 machine learning algorithm can be configured to identify a feature in the 2016 raw source data as a predetermined feature. The 2016 raw source data can be derived from a variety of sources. For example, the 2016 raw source data can be actual input data collected by a machine learning system.The raw source data from 2016 can be generated automatically for testing the system.

[0360] FIG. 12A shows an example of an embodiment of a system that uses information from wireless signals to determine the presence of a user or even for PID. In the example shown, wireless signals 325A, 325B and 325C from 202503111 are used.

[0361] 65 / 91

[0362] Wireless transmitter 302 transmits data to wireless receiver 304. The wireless transmitter 302 can, in various embodiments, be a Wi-Fi router in a home, although the scope of the disclosure is not limited to Wi-Fi implementations and thus includes others (e.g., Bluetooth). Similarly, while FIG. 12A depicts a domestic environment, the further disclosure is not limited in this way, and thus the various systems and methods disclosed herein can be implemented in a variety of environments.

[0363] In the illustrated embodiment, the receiver 304 is embedded in a household appliance (e.g., a coffee maker or a household refrigerator). The household appliance can be operated as an Internet of Things (IoT) device. More generally, the receiver can be part of virtually any type of device or equipment capable of receiving and processing wireless signals transmitted by the transmitter 302. Both the transmitter 302 and the receiver 304 are located in a space (e.g., a room in a house) that may include other static surfaces 340, such as cabinets, walls, floor and ceiling, furniture, etc. In this example, a person 310 is located on a path between the transmitter 302 and the receiver 304 and may attenuate or block wireless signals on this particular path.

[0364] The wireless signals 325A, 325B, and 325C transmitted by the transmitter are subject to multipath propagation, as shown in FIG. 12A. The signals can be reflected by various static surfaces 340 before being detected by the receiver 304. Due to the different paths and thus the path lengths, parts of wireless signals transmitted at a given time can be received by the receiver 304 at different times. Accordingly, the receiver 304 can use beamforming techniques to combine received signals, as received from different angles (via different paths), in order to improve the received signal strength.

[0365] The receiver 304 in the illustrated embodiment can use multipath propagation and knowledge about the environment to detect the presence and movement of person 310. Since person 310 can attenuate or block some wireless signals, the presence and movement of person 310 can be detected.

[0366] 66 / 91

[0367] Receiver 304 can be detected. When detecting the presence of person 310 (and more generally, any person), receiver 304 can perform both coarse and fine-grained identification using various artificial intelligence / machine learning (AI / ML) techniques. For example, coarse-grained identification can determine whether the detected person is an adult or a child and can also determine their approximate height. For fine-grained identification, receiver 304 can determine and identify the presence of a specific person 310 based on factors such as gait, other movement signatures, etc.

[0368] Using the AI / ML techniques of this further disclosure, a model can be subjected to multi-classifier training to identify different individuals who can perform different actions in the space containing both the sender 302 and the receiver 304. Using the AI / ML techniques, the receiver 304 can also recognize usage patterns for the different individuals and perform additional actions based on these patterns. For example, if the receiver 304 is embedded in a coffee machine, it can cause the coffee machine to dispense coffee in a specific way (e.g., black, with cream but no sugar, etc.) during the morning in response to the detection of individual 310, having learned corresponding usage patterns.Alternatively, a device such as a coffee machine in which the Receiver 310 is embedded can use audio to ask the person if they would like a specific action to be performed after their presence has been detected and the appropriate identification has been carried out.

[0369] FIG. 12B shows an example of information that can be extracted from wireless signals for use in an embodiment of a method for determining the presence of a user or PID. In particular, FIG. 12B illustrates, for a single packet received at a receiver (e.g., receiver 304 from FIG. 3A), CSI amplitude information 352, CSI phase information 354, and CSI received signal strength indicator (RSSI) information 356.

[0370] 67 / 91

[0371] As explained above with reference to FIG. 12A, wireless signals undergo multipath propagation between a transmitter and a receiver. If interference is introduced into the various propagation paths, the amplitude, phase, and received signal strength of these signals at the receiver can be affected. Accordingly, if a person moves along the various propagation paths, the receiver can detect this movement by observing changes in amplitude, phase, and received signal strength. Furthermore, the changes in amplitude, phase, and received signal strength within a given packet or across multiple packets can be used to determine characteristics of the person's movement. For example, the human gait pattern is unique to each individual due to various factors.Accordingly, changes in amplitude, phase, and received signal strength within a given packet or across multiple packets can be used to determine a specific person's gait pattern during the training of a machine learning model and later for classification and thus identification of that person. This information, combined with other information (e.g., the person's usage patterns and routines), can be used to trigger additional actions by devices associated with the receiver.

[0372] FIG. 12C shows a workflow diagram for an embodiment of a method according to the further disclosure. Method 360 in the embodiment shown can be performed in a variety of implementations contained in or associated with a wireless receiver. Method 360 comprises the collection of wireless signal data (Block 362), data preprocessing (Block 364), an annotation process (Block 366), classifier training (Block 368), and inference (Block 370). These various operations will now be explained in more detail.

[0373] For the collection of wireless signal data (Block 362), a receiver can be used to capture wireless signals and gather information from them. For example, a Wi-Fi receiver can receive wireless signals in packets and collect CSI (e.g., amplitude and phase) from the received packets. Variations in the amplitude and phase detected in the CSI, both within individual packets and across multiple packets, can be used to

[0374] 68 / 91

[0375] To detect human movement in a space (e.g., a room in a house) where both the sender and the receiver are located.

[0376] The transmitter can be, in various embodiments, a Wi-Fi router already present in the room, although other types of transmitters (e.g., Bluetooth) are possible and provided for within the scope of this disclosure. The receiver can, as noted above, be embedded in a device, in particular a household appliance, or other equipment capable of using the information obtained from the CSI of the received packets for identification purposes. In various embodiments, the receiver can be connected to a laptop / tablet / phone to control the CSI collection and to visually inspect this data during this acquisition phase.

[0377] CSI traces within packets can be used to explicitly capture time-domain (e.g., variations in amplitude over time) as well as frequency-domain information. These traces can be particularly sensitive to human movement in the environment. Accordingly, amplitude, phase, and received signal strength indicator (RSSI) values ​​can be determined from each CSI trace to gain different perspectives on the channel frequency response caused by human movement.

[0378] Data preprocessing (Block 364) can perform various processing tasks to minimize the effects of noise in the received packets. As noted above, CSI can be particularly sensitive to human movement and, more generally, to the environment as a whole, since wireless signals travel multiple paths and thus multiple reflections to reach the receiver. For example, signals in a room can be reflected by furniture, walls, ceilings, floors, cabinets, and other inanimate objects. Furthermore, the presence of other signal sources in the environment (e.g., another Wi-Fi access point operating on the same channel) can introduce additional interference into the collected data. These factors can make the CSI signals extremely noisy.To reduce the influence of unwanted noise and improve the detection of human movement, the 360 ​​method can apply a series of preprocessing steps to clean up the signal data. For example, 202503111.

[0379] 69 / 91

[0380] Raw CSI data includes null and pilot subcarriers, which are part of the Orthogonal Frequency Division Multiplexing (OFDM) stack to reduce interference for users working on multiple frequency channels. These subcarriers are removed to reduce dimensionality and redundancy, as they also carry no useful information. CSI segments with a high proportion of human motion can then be selected using the annotated labels during data collection.

[0381] One embodiment, discussed below with reference to FIG. 13A, performs preprocessing by using the CSI amplitude as an identification feature. Another embodiment can utilize preprocessing that uses CSI after phase unwrapping and phase cleanup, and is further discussed below with reference to FIG. 13B. Embodiments that utilize Doppler frequency shift (DFS) maps from raw, complex-valued CSI are also possible and provided for. In DFS implementations, amplitude and phase values ​​can be cleansed according to the aforementioned algorithms, followed by performing a short-time Fourier transform on smaller overlapping CSI windows. The result of this workflow can yield DFS maps that exhibit high sensitivity to human motion. One embodiment uses statistical features as input for a machine learning algorithm.In the statistical feature implementation, the preprocessed CSI amplitude or phase can be used to extract specific time and frequency domain information such as mean, variance, fast Fourier transforms (FFT), etc. It should be noted that preprocessing using any combination of features from these implementations is also possible and intended.

[0382] The annotation process (Block 366) can be used for person identification. In various implementations, CSI samples corresponding to a person can be given an anonymous label, e.g., "Person A," to protect privacy. The labels can remain the same for each person regardless of changes in the environment and collection time to maintain consistency for the PID task. Depending on the number of people in the data collection and annotation phases, the type of classifier training can be either binary or multi-class.

[0383] 70 / 91

[0384] Change classification. In this phase, each time segment (e.g., 2 seconds of data) can be annotated with a class, e.g., Person A. In an embodiment configured for home use, the total number of classes can equal the total number of residents in a house, or it could be one more (e.g., an unknown visitor). These annotations can be collected by a user as an initial setup phase or collected in the background based on user interaction with a product, e.g., a coffee maker or a household refrigerator. For example, the coffee maker or refrigerator could simply monitor the residents' behavior for seven days, and when someone dispenses a coffee or opens a door of the refrigerator, it takes the previous sequences of Wi-Fi data and labels them with a user profile, e.g., Person A, and uses this data for future reference.

[0385] Classifier training can be performed in various embodiments, the features of which can be combined. One embodiment can use a 2D convolutional neural network (CNN) based on standard ResNet18 residual connections and squeeze excitation (SE) blocks. Such a network can have a small number of parameters (e.g., 680k) and can take a spatiotemporal CSI amplitude as input, with the input being used to learn to predict the correct label corresponding to each element in the provided datasets. In one embodiment, the proposed neural network can include an input convolutional layer with a kernel size of 7x7. In another embodiment, a different filter size can be used instead of a 7x7 filter, e.g., a 51x51 filter.The large input convolutions in the input space can lead to the ingestion of more CSI information in the time and frequency domains, resulting in enhanced feature representation learning in the downstream convolution blocks. Moving to the deeper layers of the network, the residual identity blocks can preserve more information in subsequent deeper layers, while the SE blocks can improve the network's representation performance by enabling dynamic channel-wise feature recalibration. Implementations can also use a ResNet18 model as a baseline to compare the performance of a network that may have a large number of parameters. The training phase can be completed automatically when the models are set to 202503111.

[0386] 71 / 91

[0387] This achieves high classification accuracy and consequently low training object loss on the validation sets. In addition to amplitude, phase information can also be provided as input for the neural network. Several separate approaches to utilizing phase information can be implemented. For example, amplitude and phase can be chained together as a long sequence and learned jointly by a single network (early fusion). In another example, two separate neural networks can be trained to learn amplitude and phase features separately and chain the feature maps together before distilling the knowledge at classification layers (late fusion).

[0388] In some embodiments of the classifier training performed according to Block 368, statistical features can be used for classification. In these embodiments, a comprehensive set of features can be extracted from Wi-Fi CSI data for use in machine learning models. Both time-domain and frequency-domain features can be extracted from the CSI data, which can be represented as a three-dimensional array with dimensions corresponding to samples, subcarriers, and journals.

[0389] In various implementations, classifier training can include performing principal component analysis (PCA): PCA is applied to transform the data, reduce its dimensionality, and retain essential information. This step involves reshaping each sample to align time steps and channels, and then applying PCA, resulting in a set of principal components that capture the most significant variations in the signal.

[0390] In some implementations, classifier training may include determining the RSSI rate of change by calculating the rate of change of the received signal strength indicator (RSSI) over a specified interval. This can provide insight into how the signal strength changes over time, which can contribute to understanding the dynamics of the wireless channel.

[0391] Implementations of classifier training can also utilize time-domain features. Various statistical measures can be calculated via subcarriers.

[0392] 72 / 91

[0393] including mean, median, variance, standard deviation, skewness, kurtosis, root mean square (RMS), and zero-crossing rate. These features can help characterize the distribution and variability of signals in the time domain.

[0394] Implementations can also, or alternatively, use frequency domain features. A fast Fourier transform (FFT) can be applied to extract features such as spectral centroid, spectral bandwidth, spectral flatness, and peak frequency. These features capture essential properties of the signal's frequency content, such as its dominant frequency and dispersion.

[0395] Time-frequency plots can be used in some implementations. For example, short-time Fourier transform (STFT) coefficients can be calculated and analyzed, providing a combined view of how the frequency content of the signal evolves over time. Correlation analysis can also be used in some implementations. Subcarrier correlation can be calculated for each time point, providing insights into the relationships and dependencies between different parts of the spectrum.

[0396] The various methods used in classifier training can consolidate the diverse features discussed above into a two-dimensional feature matrix suitable for machine learning models. This transformation can involve reshaping and bin averaging, which in turn can reduce the number of time steps required for the classification task. The feature matrix can then be flattened and passed to classifiers such as SVM, XGBoost, MLP, Random Forest, Decision Tree, etc.

[0397] Implementations of classifier training, in which similar statistical features, like those discussed above, can also be used with a sequence model for the classification task to capture the temporal pattern of human movement, exist. Different sequence models can be used in different implementations.

[0398] In one embodiment, the sequence model can include a bidirectional gated recurrent unit (BiGRU) with an attention mechanism. The BiGRU layer 202503111

[0399] 73 / 91

[0400] It can process input sequences in both forward and backward directions and capture dependencies across time steps. The attentional mechanism can assign weights to different parts of the GRU output and focus on more relevant sequence segments for the task at hand. The network can also include a dropout layer for regularization, which in turn can reduce the probability of overfitting. Finally, a fully connected layer can map the attention-weighted features to the desired output classes. This combination of bidirectional GRU and attention can be effective for tasks where understanding the context and meaning of different parts of a sequence is a component. The statistical features can be fed into the BiGRU units in various embodiments from different journals as inputs.In some embodiments, pre-processed CSI amplitude and / or CSI phase information from various journals can be fed into the BiGRU units as inputs.

[0401] Implementations are also possible and planned in which bidirectional Long Short Term Memory (LSTM) with an attention mechanism is used in the classification process as an alternative to the GRU discussed above. Recurrent Neural Network (RNN) architectures can be used in various implementations.

[0402] In some embodiments, transformers are used to model the sequence. Several variations of the transformer can be used, including (a) transformers with static positional coding, (b) transformers with learnable positional coding, and (c) transformers with static positional coding for the time domain and learnable positional coding for the feature domain, and vice versa.

[0403] After classifier training, the trained models can be used for inference evaluation (Block 370) by keeping the parameters frozen. Inference can be performed either in real time or offline. Offline inference can follow the same or similar steps as those discussed above for, e.g., CSI amplitude processing and / or CSI phase processing.

[0404] 74 / 91

[0405] The raw CSI samples are provided as input for the trained model. A method for performing online inference is discussed below with reference to FIG. 13D.

[0406] Although not shown in Fig. 12C, embodiments that include an actuation phase are also possible and provided for. In the actuation phase, a system that includes the receiver and executes the ML model can personalize the user experience of a device or other piece of equipment. The actuation can involve the system performing or controlling various actions or functions (e.g., dispensing coffee according to personal preferences, activating predefined external lighting) with minimal (if any) user input.

[0407] FIG. 13A shows a workflow diagram and signal characteristics for the CSI amplitude preprocessing of one embodiment of the disclosed method. As shown in FIG. 13A, zero subcarriers can be removed from received packets, followed by noise removal using, for example, a Butterworth bandpass filter and Hanning smoothing. Regardless of the type of filter used, the filter operation can remove high and very low frequency noise components corresponding to reflections from inanimate objects (e.g., furniture, cabinets, etc.). Other types of stationary and non-stationary signal processing filters, such as low-pass and wavelet-based methods, can also be implemented as another embodiment.

[0408] DC offset components can also be removed, followed by data normalization. To further reduce dimensionality, lossless frequency-based subsampling can be performed. Alternatively, PCA-based dimensionality reduction can be used instead of subsampling.

[0409] Procedure 400 comprises the extraction of the CSI amplitude (Block 402, graphically represented in 420), which involves determining the raw amplitude of wireless signals received in a packet. The procedure then continues with the removal of null subcarriers (Block 404, graphically represented in 404), since the null subcarriers carry no information and are mainly used to form the 202503111

[0410] 75 / 91

[0411] The signal spectrum is used. The procedure then performs sample duration selection and truncation (Block 407), followed by filter operations that include the removal of high-frequency components (Block 408) and the removal of DC components (Block 410). The filter operations can be visualized as bandpass filtering in 424, smoothed CSI amplitude in 426, and DC-smoothed CSI amplitude in 428. The remaining data are then normalized (Block 412, graphically represented in 430). Finally, the data are subsampled to reduce dimensionality by performing lossless, frequency-based subsampling in the time domain, as shown in 432, with these data being fed into the model for classifier training (Block 424).

[0412] FIG. 13B shows a workflow diagram and signal characteristics for the preprocessing of the phases of channel state information (CSI) for an embodiment of the disclosed method. In this example, the phase of the wireless signals of a packet is used to extract the data for classifier training.

[0413] Procedure 431 comprises the extraction of the CSI phase (block 434) from raw CSI phase data, graphically represented in 450. The null subcarriers are removed, and the phase data is unwrapped (blocks 436 and 438, respectively, graphically represented in 452), followed by phase correction and smoothing (block 440, graphically represented in 454). The CSI phase data is then smoothed (graphically represented in 456), and DC components are subsequently removed (block 442, graphically represented in 458). Data normalization is then performed (block 444, graphically represented in 460). Subsampling is then carried out (462), and the subsampled data is fed into the model for classifier training (block 446).

[0414] FIG. 13C shows a neural network and a workflow diagram illustrating its operation for one embodiment of the disclosed method. In the embodiment shown, the neural network 480 comprises an input convolution layer 471, which in this particular example has a kernel size of 7x7, although embodiments with filters of different sizes are possible and provided for. Generally speaking, the input convolutions can be large in various embodiments to accommodate the input of more CSI data in the time and / or

[0415] 76 / 91

[0416] to enable frequency domains, which can lead to better representations in subsequent convolution blocks of the neural network 480.

[0417] The data output by the input convolution layer 471 can be normalized by a normalization layer 472. The normalized data output by the normalization layer is then fed to a maxpool layer 472, where it can be downsampled to reduce its spatial dimensions, which in turn can make the representation smaller and more manageable. The downsampled data can then be applied to a dropout layer 474 to prevent overfitting in the downstream layers. The operations performed in the dropout layer 474 can include the random ignoring of specific neurons and their respective connections. Neurons can be discarded at a specified rate, and this rate can be tunable in various embodiments.

[0418] The data from dropout block 474 can then be applied to a ResNet layer 475, where additional downsampling and convolution are performed. In this particular embodiment, the data output of this layer comprises 64 channels (and thus 64 feature maps) with a 3x3 kernel for the convolution operation. The data from ResNet layer 475 is then applied to a squeeze-excite layer 476 to further refine the classification by performing channel-wise feature recalibration. The data output from layer 476 then undergoes another dropout operation in dropout layer 477.

[0419] After the dropout operation of layer 477, the resulting data are applied to another ResNet layer 478, where additional downsampling and convolution are performed using a convolution block with a 3x3 matrix. This downsampled data is then fed to another squeeze-excite layer 479, followed by another dropout operation in dropout layer 480. Afterward, the data from dropout layer 480 are applied to an average pooling layer 471 to further reduce the spatial dimensions of the data. The reduced data are then fed to a linear softmax layer 481 for final classification. In the linear portion of layer 481, the output is provided as a combination of inputs, in which a weighted sum of inputs is calculated.

[0420] 77 / 91

[0421] and a bias term is calculated. The softmax function serves as an activation function for a multi-class classification operation, which, in the context of this further disclosure, may involve the identification of one of several different persons. Since the output of the softmax function can be interpreted as the probability that the input data belongs to a particular class, the output, in the context of the disclosure, can provide a probability that the input data corresponds to a specific person.

[0422] FIG. 13D is a workflow diagram illustrating an embodiment of a method for performing inference according to the further disclosure. In particular, the method 489, as depicted in the illustrated embodiment, can represent a real-time inference methodology using input data after training a machine learning model, and thus the process for identifying a specific person. In this example, the CSI traces from real-time packets can enter the model as a continuous data stream. Window-based slicing can be performed to obtain smaller portions of the time-series data and to carry out the subsequent preprocessing steps, which are similar to the training data preparation. After signal cleaning, the signal can be provided as input to the classifier to output a class prediction for each moving window.

[0423] Method 489 comprises real-time CSI windowing (Block 484) and CSI amplitude extraction (Block 486). It should be noted that while Method 489 is discussed here using CSI amplitude information, CSI phase information or other information may be used in other embodiments. Zero subcarrier removal (Block 488), high-frequency noise removal (Block 490), DC component removal (Block 492), and data normalization (Block 494) are also performed in the same or a similar manner as during the training process. The resulting data can then be provided for inference (Block 496) to perform the determination or identification of a person as discussed above.

[0424] 78 / 91

[0425] FIG. 14 shows a schematic diagram of an interaction between a computer-controlled machine 500 and a control system 502. The computer-controlled machine 500 comprises an actuator 504 and a sensor 506. The actuator 504 may comprise one or more actuators, and the sensor 506 may comprise one or more sensors. The sensor 506 is configured to detect a state of the computer-controlled machine 500. The sensor 506 may be configured to encode the detected state into sensor signals 508 and to transmit sensor signals 508 to the control system 502. Non-limiting examples of the sensor 506 include wireless receivers, video, radar, LiDAR, ultrasonic, and motion sensors, as described above with reference to FIGS. 10-11. In one embodiment, the sensor 506 is a wireless sensor configured to detect an environment near the computer-controlled machine 500.Designs in which a combination of different sensors is possible and provided for are also possible.

[0426] The 506 sensor can also be configured in various embodiments as a wireless signal receiver, designed to receive wireless signals from a transmitter (e.g., Wi-Fi). The computer-controlled machine can then use the received wireless signals for personal identification in various embodiments, based on the motion detection of a nearby person and their characteristics, which are used to train a machine learning model.

[0427] The control system 502 is configured to receive sensor signals 508 from the computer-controlled machine 500. As explained below, the control system 502 can also be configured to calculate actuator control commands 510 based on the sensor signals and to transmit actuator control commands 510 to the actuator 504 of the computer-controlled machine 500.

[0428] As shown in FIG. 14, the control system 502 includes a receiver unit 512. The receiver unit 512 can be configured to receive sensor signals 508 from the sensor 506 and convert sensor signals 508 into input signals x. In an alternative embodiment, sensor signals 508 are received directly as input signals x without a receiver unit 512. Each input signal x can be a part of each sensor signal 508. The receiver unit 512 can be configured to

[0429] 79 / 91

[0430] The sensor signal 508 is processed to generate any input signal x. The input signal x can contain data corresponding to an image recorded by the sensor 506.

[0431] The control system 502 includes a classifier 514. The classifier 514 can be configured to classify input signals x into one or more labels using a machine learning (ML) algorithm, such as a neural network described above. The classifier 514 is configured to be parameterized by parameters, such as those described above (e.g., parameter 0). Parameter 0 can be stored in and provided by non-volatile memory 516. The classifier 514 is configured to determine output signals y from input signals x. Each output signal y contains information that assigns one or more labels to each input signal x. The classifier 514 can transmit output signals y to a conversion unit 518. The conversion unit 518 is configured to convert output signals y into actuator control commands 510.The control system 502 is configured to transmit actuator control commands 510 to the actuator 504, which is configured to actuate the computer-controlled machine 500 in response to the actuator control commands 510. In another embodiment, the actuator 504 is configured to actuate the computer-controlled machine 500 directly based on the output signals y.

[0432] Upon receiving the actuator control commands 510, the actuator 504 is configured to perform an action corresponding to the associated actuator control command 510. The actuator 504 may include control logic configured to convert actuator control commands 510 into a second actuator control command used to control the actuator 504. In one or more embodiments, actuator control commands 510 may be used to control a display instead of, or in addition to, an actuator.

[0433] In another embodiment, the control system 502 includes the sensor 506 instead of, or in addition to, the computer-controlled machine 500, which includes the sensor 506. The control system 502 can also include the actuator 504 instead of, or in addition to, the computer-controlled machine 500, which includes the actuator 504.

[0434] 80 / 91

[0435] As shown in FIG. 14, the control system 502 also includes a processor 520 and a memory 522. The processor 520 can comprise one or more processors. The memory 522 can comprise one or more memory devices. The classifier 514 (e.g., machine learning algorithms, such as those described above with respect to the pretrained classifier 306) of one or more embodiments can be implemented by the control system 502, which comprises a non-volatile memory 516, a processor 520, and a memory 522.

[0436] The non-volatile memory 516 can include one or more persistent data storage devices such as a hard disk, optical drive, tape drive, non-volatile solid-state device, cloud storage, or any other device capable of persistently storing information. The processor 520 can include one or more devices selected from high-performance computing (HPC) systems, including high-performance cores, microprocessors, microcontrollers, digital signal processors, microcomputers, central processing units, field-programmable gate arrays, programmable logic devices, state machines, logic circuits, analog circuits, digital circuits, or any other device that manipulates signals (analog or digital) based on computer-executable instructions stored in memory 522.The memory 522 can include a single storage device or a number of storage devices, including but not limited to random access memory (RAM), volatile memory, non-volatile memory, static random access memory (SRAM), dynamic random access memory (DRAM), flash memory, cache memory, or any other device capable of storing information.

[0437] The processor 520 can be configured to read into the memory 522 and execute computer-executable instructions stored in the non-volatile memory 516, which embody one or more machine learning algorithms and / or methodologies of one or more implementations. The non-volatile memory 516 can include one or more operating systems and applications. The non-volatile memory 516 can store compiled and / or interpreted computer programs created using a variety of programming languages ​​and / or technologies.

[0438] 81 / 91

[0439] were, including, without limitation, and either alone or in combination, Java, C, C++, C#, Objective C, Fortran, Pascal, Java Script, Python, Perl and PL / SQL.

[0440] When executed by the processor 520, the computer-executable instructions in the non-volatile memory 516 can cause the control system 502 to implement one or more of the machine learning algorithms and / or methodologies disclosed herein. The non-volatile memory 516 can also contain machine learning data (including data parameters) that support the functions, features, and processes of one or more embodiments described herein.

[0441] The program code embodying the algorithms and / or methodologies described herein may be distributed individually or collectively as a program product in a variety of different forms. The program code may be distributed using a computer-readable storage medium containing computer-readable program instructions to instruct a processor to execute aspects of one or more embodiments. Computer-readable storage media, which are inherently non-transient, may include volatile and non-volatile, as well as removable and non-removable tangible media, implemented in any method or technology for storing information such as computer-readable instructions, data structures, program modules, or other data.Computer-readable storage media may also include RAM, ROM, erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other solid-state storage technology, portable compact disc read-only storage (CD-ROM) or other optical storage, magnetic cartridges, magnetic tape, magnetic storage media or other magnetic storage devices, or any other medium that can be used to store the desired information and that can be read by a computer. Computer-readable program instructions may be downloaded from a computer-readable storage medium to a computer, another type of programmable data processing device, or other device, or transferred over a network to an external computer or external storage device.

[0442] 82 / 91

[0443] Computer-readable program instructions stored on a computer-readable medium can be used to instruct a computer, other types of programmable data processing equipment, or other devices to operate in a specific manner, such that the instructions stored on the computer-readable medium generate a manufactured item containing instructions that implement the functions, actions, and / or operations specified in the flowcharts or diagrams. In certain alternative embodiments, the functions, actions, and / or operations specified in the flowcharts and diagrams can be reordered, processed serially, and / or processed concurrently, which is consistent with one or more embodiments. Furthermore, all flowcharts and / or diagrams can contain more or fewer nodes or blocks than those depicted, which is consistent with one or more embodiments.

[0444] The processes, procedures or algorithms can be embodied wholly or partially using suitable hardware components such as application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), state machines, controllers or other hardware components or devices, or a combination of hardware, software and firmware components.

[0445] FIG. 15 illustrates an application 600 of an embodiment of the method for determining the presence of a user (corresponding to person 40) of a household refrigeration unit 604 (corresponding to the household refrigeration unit 1). The user 602 has purchased groceries at a nearby supermarket and wants to store them in his household refrigeration unit 604. Therefore, he opens the door 606 (corresponding to door 5) of the household refrigeration unit 604. The household refrigeration unit 604 includes a wireless receiver 608 (corresponding to wireless interface 21) that collects channel state information from received packets transmitted by a wireless transmitter (not shown, corresponding to external device 30). The channel state information is preprocessed in the manner described herein. In particular, segments of the channel state information are selected.Using one or more parameters of the selected segments of the channel state information, the presence of user 602 can be identified based on additional packets received by the wireless receiver 202503111.

[0446] 83 / 91

[0447] 608 is received. The intuition behind this is that user 602 has an influence on the communication channel, e.g., due to their body tissue. Therefore, an analysis of the collected channel state information can be used to detect the presence of user 602 near refrigerator 604.

[0448] In response to the detection of user 602's presence near the household refrigeration appliance 604, an operation or function of the appliance 604 can be controlled. Normally, if the refrigerator door is open, a conventional refrigerator will, after a short time (e.g., after 15 or 90 seconds), assume that the user has forgotten to close the door and will activate a door alarm. The door alarm may be a beeping sound to indicate that the temperature inside the refrigerator is rising—in other words, that cooling energy is being lost. Many users find this annoying because they are well aware that the refrigerator door is open but cannot avoid it, as they want to store the groceries they have purchased in the refrigerator.In the illustrated use case 600, the household refrigerator 604 therefore refrains from beeping because it detects that the user 602 is standing in front of the open refrigerator 604 and is therefore well aware that the door 606 is open. In this way, the user 602 is not disturbed, while the refrigerator can still warn the user that the door has been left open if no one is standing in front of the refrigerator.

[0449] While exemplary embodiments are described above, it is not intended that these embodiments describe all possible forms encompassed by the claims of the present invention. The words used in the description are descriptive words and not limiting words. It is understood that various modifications can be made without departing from the spirit and scope of the disclosure. As previously described, the features of different embodiments can be combined to form further embodiments of the invention that may not be explicitly described or illustrated. While various embodiments may have been described as advantageous or preferred over other embodiments or implementations of the prior art with respect to one or more desired features, those skilled in the art recognize that one or more features or

[0450] 84 / 91

[0451] Characteristics may be compromised to achieve desired overall system attributes, which depend on the specific application and implementation. These attributes may include, but are not limited to, cost, strength, durability, life-cycle costs, marketability, appearance, packaging, size, ease of maintenance, weight, manufacturability, ease of assembly, etc. If any embodiments are described as less desirable than other embodiments or implementations of the prior art with respect to one or more features, these embodiments are not outside the scope of the disclosure and may be desirable for certain applications.

[0452] 85 / 91

[0453] REFERENCE MARK LIST

[0454] 1 household refrigerator

[0455] 3 Interior

[0456] 5 5 Door

[0457] 7 Lighting equipment

[0458] 9 Input device

[0459] 9a Operating device

[0460] 11 Door opening device

[0461] 20 Control device

[0462] 21 Wireless interface

[0463] 23 Processing facility

[0464] 25 Signal sequence

[0465] 26 Door alarm

[0466] 15 27 Notification

[0467] 30 External Institution

[0468] 31 Wireless interface

[0469] 40 people

[0470] 50 External user device

[0471] 20 51 Wireless interface

[0472] 60 device; household appliance

[0473] 80 System

[0474] 100 System

[0475] 105 household

[0476] 25,110 First household appliance

[0477] 115 Second household appliance

[0478] 120 Control device

[0479] 125 Processing unit

[0480] 130 Wireless Interface

[0481] 30 135 External Office

[0482] 140 Wireless Interface

[0483] 200 procedures

[0484] 205 Create channel

[0485] z

[0486] Q202503111

[0487] 86 / 91

[0488] 210 determine status information

[0489] 215 changes?

[0490] 220 Compare with another household appliance 225 Identify person

[0491] 230 household appliance control

Claims

202503111 87 / 91 PATENT CLAIMS 1. Method for controlling a household refrigerating appliance (1) comprising a refrigerated interior (3) for storing food, a door (5) for closing the interior (3), a control device (20) and a wireless interface (21), the method comprising the following steps: - Establishing a communication channel via the wireless interface (21) to an external device (30) in the area of ​​the household refrigeration appliance (1); - Collecting channel state information of the communication channel at the wireless interface (21) transmitted by the external device (30); - Monitoring a communication parameter read from the channel state information; - Determining a person (40) within the area of ​​the household refrigeration appliance (1) based on monitoring, wherein the person (40) can be determined based on a change in the communication parameter; and - Controlling a function of the household refrigeration appliance (1) by means of the control device (20) depending on a result of the determination of the person (40).

2. Method according to claim 1, wherein a presence, an absence, a distance, an identity and / or a gesture of the person (40) is determined and output as a result of the determination of the person (40).

3. Method according to claim 1 or 2, wherein controlling a function of the household refrigeration appliance (1) comprises controlling different functions of the household refrigeration appliance (1), each of which is selected depending on the result of the determination of the person (40).

4. The method of claim 3, wherein the control of the various functions can each be activated and / or deactivated independently of one another by means of an input device (9; 50). 88 / 91 5. Method according to one of the preceding claims, wherein the controlled function is a suppression of a door alarm (26) when the door (5) is open, as long as the presence of the person (40) in front of the household refrigeration appliance (1) is determined, and in particular an activation of the door alarm (26) when the door (5) is open, upon determination of an absence of the person (40) in front of the household refrigeration appliance (1) for a predefined period.

6. Method according to claim 5, wherein during the suppression of the door alarm (26) a predefined, in particular acoustic, signal sequence (25) is output, which is selected depending on a number of the determined presences of the person (40) and on a time component.

7. Method according to one of the preceding claims, wherein during an energy-saving mode of the household refrigerating appliance (1) the steps for determining the person (40) are carried out repeatedly after a predefined time sequence.

8. Method according to one of the preceding claims, wherein the controlled function is a suppression and / or activation of an operating action of the household refrigerating appliance (1), in particular a noisy operating action, for a predetermined period.

9. Method according to one of the preceding claims, wherein the controlled function is an activation of a lighting device (7), an operating device (9a), an energy-saving mode, a door opening device (11) perceptible from the outside of the household refrigerating appliance, and / or a control of a function of another device (60), in particular a household appliance (60), in the vicinity of the household refrigerating appliance (1).

10. A method according to any of the preceding claims, wherein controlling the function comprises sending a notification (27) to an external user device (50), the notification (27) containing information about the result of the determination of the person (40) within a predefined time window. 89 / 91 11. Method according to one of the preceding claims, wherein, to establish the communication channel, the wireless interface (21) sends a request signal, in particular a ping, to the external device (30), and the channel status information is collected from subsequently received packets.

12. Method according to any of the preceding claims, wherein the determination of the person (40) is carried out by means of a machine learning algorithm which is configured to determine the person (40) based on the change of the communication parameter.

13. Control device (20) for a domestic refrigeration appliance (1), wherein the control device (20) comprises a wireless interface (21) and a processing unit (23); wherein the processing unit (23) is configured to establish a communication channel via the wireless interface (21) to an external device (30) in the vicinity of the domestic refrigeration appliance (1); to collect channel status information of the communication channel at the wireless interface (21) transmitted by the external device (30); to monitor a communication parameter read out via the channel status information; and, based on the monitoring, to identify a person (40) in the vicinity of the domestic refrigeration appliance (1), wherein the person (40) is identifiable based on a change in the communication parameter;and wherein the control device (20) is configured to control a function of the household refrigeration appliance (1) depending on a result of the determination of the person (40).

14. Household refrigeration appliance (1) comprising a control device (20) according to claim 13.

15. System (80) comprising a household refrigeration appliance (1) according to claim 14 and a further household appliance (1; 60) with a wireless interface (21) and a processing unit (23); wherein the processing unit (23) of the further household appliance (1; 60) is configured to establish a communication channel via the wireless interface (21) to an external device (30) in the vicinity of the household appliance (1; 60); to monitor a communication parameter read via the channel status information; and to transmit a result of the monitoring to the household appliance (1; 60) and / or the household refrigeration appliance (1). 90 / 91 transmit; and wherein the processing unit (23) of the household refrigerating appliance (1) and / or the household appliance (1; 60) is equipped to identify the person (40) on the basis of multiple monitoring operations.

Images