Systems and methods for automatically adjusting condition of an article of furniture

Abstract

An aspect of the present disclosure provides a system comprising an article of furniture, at least one sensor that is part of the article of furniture or adjacent to the article of furniture, and a controller. The controller can be configured to direct the at least one sensor to detect and generate sensing data associated with a user of the article of furniture. The controller can be configured to adjust a condition of the article of furniture based at least in part on the sensing data. The adjustment by the controller can be performed automatically while the user is using the article of furniture, e.g., in absence of any direct instruction or control from the user.

Description

SYSTEMS AND METHODS FOR AUTOMATICALLY ADJUSTING CONDITION OF AN ARTICLE OF FURNITURECROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Patent Application No. 63 / 548,682, filed February 1, 2024, U.S. Provisional Patent Application No. 63 / 639,836, filed April 29, 2024, and U.S. Provisional Patent Application No. 63 / 644,444, filed May 8, 2024, each of which is entirely incorporated herein by reference.BACKGROUND

[0002] A condition (e.g., temperature, posture, vibration, etc.) of an article of furniture (e.g., a bed) can be modified during a person’s use of the article furniture. In some cases, the condition can be modified to improve a quality of a person’s activity on the furniture (e.g., sleeping on the bed). For example, fluid (e.g., liquid or gas) can be directed to flow within the article of furniture. Temperature of the fluid can be modulated to effect regulation of temperature of the article of furniture, to effect improvement of the person’s quality of sleep.SUMMARY

[0003] In one aspect, the present disclosure provides a computer-implemented method comprising: (a) providing, from a database, a predetermined adjustment profile for a condition of an article of furniture of a user, the predetermined adjustment profile comprising a target condition and a target duration for the target condition, wherein the predetermined adjustment profile is executable by a regulating unit that is operatively coupled to the article of furniture and configured to regulate the condition of the article of furniture during a sleep session of the user; (b) receiving a user input comprising at least one user event; (c) automatically modifying the target duration for the target condition based at least in part on the at least one user event, to generate a modified adjustment profile of the condition of the article of furniture; and (d) storing, in the database, the modified adjustment profile for use by the regulating unit.

[0004] In one aspect, the present disclosure provides a computer-implemented method comprising: (a) receiving a user sensing data associated with a user, wherein the user sensing data is of a first data type and is generated by a user sensor while the user is using an article of furniture; (b) applying the user sensing data as an input to a machine learning model that is trained to determine a target condition of the subject, wherein the machine learning model is trained based on a ground truth data comprising a training sensing data of a second data type, thefirst data type and the second data type being different data types; (c) determining, using the machine learning model, the target condition of the subject from the user sensing data, wherein the determining comprises comparing a feature derived from the user sensing data and a feature derived from the training sensing data; and (d) generating an instruction for the article of furniture to change a condition of the article of furniture based on the determined target condition of the subject in (c).

[0005] Another aspect of the present disclosure provides a non-transitory computer readable medium comprising machine executable code that, upon execution by one or more computer processors, implements any of the methods above or elsewhere herein.

[0006] Another aspect of the present disclosure provides a system comprising one or more computer processors and computer memory coupled thereto. The computer memory comprises machine executable code that, upon execution by the one or more computer processors, implements any of the methods above or elsewhere herein.

[0007] In some aspects, the present disclosure provides a computer-implemented method for controlling a condition of an article of furniture, the method comprising: (a) determining, by a computer algorithm, a target condition of a user based on a user sensing data generated by a user sensor while the user is using the article of furniture, wherein the computer algorithm comprises a machine learning model trained to determine the target condition of the subject based on a training data, wherein the training data and the user sensing data are different data types; and (b) generating an instruction for the article of furniture to change the condition of the article of furniture based on the determined target condition of the user.

[0008] In some aspects, the present disclosure provides a computer-implemented method for controlling a condition of an article of furniture, the method comprising: (a) determining a sleep disorder of a user of the article of furniture when a threshold indicator of a user sensing data is identified, wherein the user sensing data is generated by a user sensor while the user is using the article of furniture, and wherein the threshold indicator is based on at least one user factor; and (b) generating an instruction for the article of furniture to change the condition of the article of furniture based on the determined sleep disorder of the user.

[0009] In some aspects, the present disclosure provides a computer-implemented method for controlling a condition of an article of furniture of a user, the method comprising: (a) analyzing a quality of a target sleep phase of the user from a prior sleep session of the user; and (b) automatically modifying a predetermined adjustment profile of the condition of the article of furniture based on the analyzed quality of the target sleep phase, to generate a modified adjustment profile of the condition of the article of furniture, wherein the automaticallymodifying comprises changing a target value of the condition that is associated with a target sleep phase of the user, and wherein the modified adjustment profile is executable by the article of furniture to regulate the condition of the article of furniture.

[0010] In some aspects, the present disclosure provides a computer-implemented method for controlling a condition of an article of furniture of a user, the method comprising: (a) receiving a user input comprising at least one user event; and (b) automatically modifying a predetermined adjustment profile of the condition of the article of furniture based on the at least one user event, to generate a modified adjustment profile of the condition of the article of furniture, wherein the automatically modifying comprises changing a target duration for subjecting at least a portion of the article of furniture to a target value of the condition, and wherein the modified adjustment profile is executable by the article of furniture to regulate the condition of the article of furniture.

[0011] In some aspects, the present disclosure provides a computer computer-implemented method for controlling a condition of an article of furniture, the method comprising: changing a condition of the article of furniture as an alarm to wake up a user of the article of furniture, wherein the changing is based on (i) a target sleep duration of the user and (ii) a time when the user falls asleep on the article of furniture or when the user begins the current use of the article of furniture.

[0012] Additional aspects and advantages of the present disclosure will become readily apparent to those skilled in this art from the following detailed description, wherein only illustrative embodiments of the present disclosure are shown and described. As will be realized, the present disclosure is capable of other and different embodiments, and its several details are capable of modifications in various obvious respects, all without departing from the disclosure. Accordingly, the drawings and description are to be regarded as illustrative in nature, and not as restrictive.INCORPORATION BY REFERENCE

[0013] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference. To the extent publications and patents or patent applications incorporated by reference contradict the disclosure contained in the specification, the specification is intended to supersede and / or take precedence over any such contradictory material.BRIEF DESCRIPTION OF THE DRAWINGS

[0014] The novel features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are utilized, and the accompanying drawings (also “Figure” and “FIG.” herein), of which:

[0015] FIG. 1 shows an example graphical user interface (GUI) for provide a sleep score to a user.

[0016] FIG. 2 shows a computer system that is programmed or otherwise configured to implement methods provided herein.

[0017] FIG. 3 shows an example of bounding boxes that are generated by a computer vision model.

[0018] FIG. 4 shows an example of the computer vision model adapted for detection of snores and noises.

[0019] FIG. 5 shows a depiction of an output with information on a user’s snoring, following the snoring detection methods described herein.DETAILED DESCRIPTION

[0020] While various embodiments of the invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions may occur to those skilled in the art without departing from the invention. It should be understood that various alternatives to the embodiments of the invention described herein may be employed.

[0021] Whenever the term “at least,” “greater than,” or “greater than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “at least,” “greater than” or “greater than or equal to” applies to each of the numerical values in that series of numerical values. For example, greater than or equal to 1, 2, or 3 is equivalent to greater than or equal to 1, greater than or equal to 2, or greater than or equal to 3.

[0022] Whenever the term “at most,” “up to,” “no more than,” “less than,” or “less than or equal to” precedes the first numerical value in a series of two or more numerical values, the term “no more than,” “less than,” or “less than or equal to” applies to each of the numerical values in that series of numerical values. For example, less than or equal to 3, 2, or 1 is equivalent to less than or equal to 3, less than or equal to 2, or less than or equal to 1.

[0023] The terms “furniture,” “article of furniture,” or “piece of furniture,” as used interchangeably herein, generally refer to a bed, a pillow, crib, bassinet, chair, seat, loveseat,sofa, couch, head rest, stool, ottoman, bench, or any panel intended to be covered with a fabric. The article of furniture can be intended for use in a home, an office, a medical facility (e.g., a hospital), or on a vehicle of transportation such as a car, truck, boat, bus, train or the like. The article of furniture can be intended for use for at least one person (and / or at least one animal, such as a pet). The article of furniture can be intended for use for at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more persons. The article of furniture can be intended for use for at most about 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 person. In an example, the article of furniture may be a bed, and the bed may comprise a plurality of sizes comprising single, single extra-long, double, queen, king, super king, etc. In another example, the article of furniture may be an infant warmer (i.e., a babytherm) to provide heat at one or more temperatures to an infant.

[0024] The terms “bed” or “bed device,” as used interchangeably herein, may be an article of furniture used for sleep or rest. The bed may comprise a mattress, a mattress pad, a pillow, and / or a covering thereof (e.g., a blanket). One or more users may sleep or rest on and / or adjacent to a surface of the bed device. The surface may be a top surface of the bed device. The top surface of the bed device may be flat or textured. The bed device may be a mattress. The bed device may be a mattress pad that covers at least a portion of a surface of a mattress or at least a surface of the mattress. The bed device may be a pillow. Alternatively or in addition to, the user(s) may sleep under a surface of the bed device. The surface may be one or more surfaces of a covering, such as, for example, a blanket. The blanket may be disposed on top of at least a part of the user(s). The bed device may be the blanket.

[0025] A temperature of the article of furniture (e.g., the bed device, such as the mattress, the mattress pad, pillow, or the blanket) may be controlled (e.g., increasing, decreasing, or maintaining the temperature of the bed). A temperature of at least a portion of the article of furniture may be controlled. The temperature of the article of furniture may be adjustable or maintained prior to, during, or subsequent to a use (e.g., sleeping or resting for a period of time) by the user(s). In an example, the bed may be pre-warmed (e.g., automatically or per user preference) prior to the use by the user(s). In some cases, temperatures or two or more portions of the article of furniture (e.g., the bed) may be controlled separately or in sync.

[0026] A regulating unit (e.g., a temperature regulating unit) can be coupled (e.g., directly or indirectly) to at least a portion of the article of furniture, to control the temperature of the at least the portion of the article of furniture. In some embodiments, the regulating unit can comprise coils that can be heated and / or cooled. In some embodiments, a variety of heating or cooling mechanisms other than the coils can also be utilized. For example, forced directional gas (e.g., air) heating and / or cooling, liquid (e.g., water) heating and / or cooling, thermoelectric heatingand / or cooling, modifications thereof, or combinations thereof can be used for controlling the temperature of the article of furniture, such as the bed device (e.g., mattress, pillow, or covering thereof).

[0027] The article of furniture (e.g., the bed) may use one or more sensors and / or one or more computer systems to detect sensing data (e.g., one or more biological signals) associated with the user. For example, the sensing data can be utilized to estimate or determine a condition of state of the user prior to, during, or subsequent to using the article of furniture (e.g., determine sleep phase, sleep pattern, disease, disorder, snoring, etc. of the user). The sensor(s) may or may not be a part of the article of furniture. The sensor(s) may be part of the article of furniture. The sensor(s) may be a part of a space (e.g., room) surrounding the article of furniture. The sensor(s) may be worn by the user(s). Non-limiting examples of a sensor can include a capacitance sensor, a temperature sensor, a pressure sensor, a piezoelectric sensor, sound sensor (e.g., a microphone), accelerometer, liquid pressure sensor, etc. The sensing data can be utilized (e.g., analyzed), at least in part, to determine how to regulate temperature of the article of furniture prior to, during, and subsequent to the user’s use of the article of furniture. In some cases, the sensor(s) may be used to detect a property (e.g., temperature, movement, etc.) of the article of furniture or such property of an environment surrounding the article of furniture. The sensor(s) of the article of furniture (e.g., coupled to, integrated into, embedded into, etc.) as provided herein can detect the sensing data when in contact (e.g., direct or indirect contact) with the subject or when not in contact with the subject.

[0028] The article of furniture can comprise from about one zone to about 20 zones. In some cases, the article of furniture can comprise at least or up to about one zone, at least or up to about two zones, at least or up to about three zones, at least or up to about four zones, at least or up to about five zones, at least or up to about six zones, at least or up to about seven zones, at least or up to about eight zones, at least or up to about nine zones, at least or up to about ten zones, at least or up to about 11 zones, at least or up to about 12 zones, at least or up to about 13 zones, at least or up to about 14 zones, at least or up to about 15 zones, at least or up to about 16 zones, at least or up to about 17 zones, at least or up to about 18 zones, at least or up to about 19 zones, or at least or up to about 20 zones. In some cases, each zone of the one or more zones can comprise a temperature control unit. In some cases, a temperature of each zone of the one or more zones can be adjusted independently.

[0029] The term “sleep phase,” as used herein, can refer to a light sleep, deep sleep, or rapid eye movement (“REM”) sleep. There can be two major stages of sleep: a non-REM sleep and a REM sleep. A person can experience a non-REM sleep first, followed by a shorter period ofREM sleep. In some cases, the person can experience a continued cycle of the non-REM sleep and the REM sleep. There may be three stages of non-REM sleep. Each stage can last from 5 to 15 minutes. The person can go through all three stages before reaching REM sleep. In stage one, the person's eyes may be closed, but the person may be easily woken up. This stage may last for 5 to 10 minutes. This stage may be considered as a light sleep. In stage two, the person may be in light sleep. The person's heart rate may slow down and the person's body temperature may drop. The person's body may be getting ready for deep sleep. This stage may also be considered as a light sleep. Stage three may be a deep sleep stage. The person may be harder to rouse during this stage, and if the person was woken up, the person would feel disoriented for a few minutes. During the deep stage of the non-REM sleep, the body may repair and regrow tissues, build bone and muscle, and strengthen the immune system. The REM sleep can happen 90 minutes after a person falls asleep. In some cases, the person may have dreams during the REM sleep. An initial period of the REM sleep may typically last 10 minutes. Any latter period of the REM sleep may get longer, and the final period of the REM sleep may last up to about an hour. The person's heart rate and respiration may quicken during the REM sleep (e.g., during the final period of the REM sleep). The person may have intense dreams during the REM sleep, since the brain is more active. The REM sleep may affect learning of certain mental skills.

[0030] A “sleep pattern”, as used herein, can indicate a recurrence or change in (i) one or more biological signals and / or (i) one or more sleep phases of the user of the bed. The sleep pattern may be described over a period of time (e.g., 0.5 hours, 1 hour, 2 hours, 3 hours, 4 hours, 5 hours, 6 hours, etc.), along with a count of the biological signal(s) or the sleep phase(s). The sleep pattern may comprise a preferred setting of the biological signal(s) or sleep phase(s) of the user. The preferred setting of the biological signal(s) may comprise a type of the biological signal(s), along with a preferred value or range of values of the biological signal(s) (e.g., a preferred body temperature or range of body temperature of the user). The preferred setting of the sleep phase(s) may comprise a type of the sleep phase(s), along with a preferred value or range of values of the sleep phase(s).

[0031] In some embodiments, the sensor(s) of the article of furniture can be disposed within a portion of the article of furniture that corresponds to a target bodily portion of the user, such as head, arms, legs, torso, upper body, lower body, etc.

[0032] A disorder of a user can be a sleep disorder. Non-limiting examples of the sleep disorder may include dyssomnias, such as insomnia, primary hypersomnia (e.g., narcolepsy, idiopathic hypersomnia, recurrent hypersomnia, posttraumatic hypersomnia, menstrual-related hypersomnia), sleep disordered breathing (e.g., sleep apnea, snoring, upper airway resistancesyndrome), circadian rhythm sleep disorders (e.g., delayed sleep phase disorder, advanced sleep phase disorder, non-24-hour sleep-wake disorder), parasomnias (e.g., bedwetting, bruxism, catathrenia, exploding head syndrome, sleep terror, rapid eye movement (REM) sleep behavior disorder, sleep talking, jet lag, restless legs syndrome, sleep deficiency (e.g., over the course of days, weeks, months, etc.), etc.

[0033] In some embodiments, the article of furniture can be a bed (e.g., a mattress or a mattress cover), and the temperature of the bed can be controlled (e.g., based at least in part on the sensing data) to assist the user(s) to fall asleep, assist the user to wake up from sleeping, promote enhanced sleep quality, treat or ameliorate the disorder of the user while sleeping, etc.

[0034] The terms “biological signal” and “bio signal” can be used interchangeably. Examples of the biological signal can include a heart signal (e.g., heart rate variability (HRV), heart rate, time elapsed between two successive R-wave peaks (R-R interval), or sound), a respiration (breathing) signal (e.g., respiration rate or sound), a motion, a temperature, a movement, perspiration, sound, neural activity, blood oxygenation (e.g., as measured from directly or indirectly contacting the skin of the user, etc. The article of furniture (e.g., the bed) may be capable of detecting one or more biological signals of the user(s). For example, the sensing data obtained by the sensor can comprise ballistocardiography (BCG) data or alike, or electrocardiogram (ECG) data or alike. The article of furniture may be capable of adjusting a property of the article of furniture (e.g., temperature or movement of the article of furniture, such as vibration, geometric configuration, etc.) to control (e.g., increase, decrease, or maintain) the biological signal(s) of the user(s) of the article of furniture.

[0035] The term “module” refers broadly to software, hardware, or firmware components (or any combination thereof). Modules are typically functional components that can generate useful data or another output using specified input(s). A module may or may not be self-contained. An application program (also called an “application”) may include one or more modules, or a module may include one or more application programs.

[0036] The term “on top of’ can mean that the two objects, where the first object is “on top of’ the second object, can be rotated so that the first object is above the second object relative to the ground. The two objects can be in direct or indirect contact, or may not be in contact at all.

[0037] The term “real time” or “real-time,” as used interchangeably herein, generally refers to an event (e.g., an operation, a process, a method, a technique, a computation, a calculation, an analysis, an optimization, etc.) that is performed using recently obtained (e.g., collected or received) data. Examples of the event may include, but are not limited to, analysis of sensing data (e.g., one or more biological signals), adjusting a condition of an article of furniture, etc. In somecases, a real-time event may be performed almost immediately or within a short enough time span, such as within at least 0.0001 milliseconds, 0.0005 milliseconds, 0.001 milliseconds, 0.005 milliseconds, 0.01 milliseconds, 0.05 milliseconds, 0.1 milliseconds, 0.5 milliseconds, 1 millisecond, 5 milliseconds, 0.01 seconds, 0.05 seconds, 0.1 seconds, 0.5 seconds, 1 second, or more. In some cases, a real time event may be performed almost immediately or within a short enough time span, such as within at most 1 second, 0.5 seconds, 0.1 seconds, 0.05 seconds, 0.01 seconds, 5 milliseconds, 1 millisecond, 0.5 milliseconds, 0.1 milliseconds, 0.05 milliseconds, 0.01 milliseconds, 0.005 milliseconds, 0.001 milliseconds, 0.0005 milliseconds, 0.0001 milliseconds, or less.

[0038] Systems and methods for adjusting an article of furniture

[0039] The present disclosure provides systems and methods for adjusting (e.g., automatically adjusting) a condition of an article of furniture.

[0040] In some embodiments, the system as provided herein can comprise at least one sensor operatively coupled to the article of furniture. The at least one sensor can be attached to the article of furniture, can be part of (e.g., disposed and hidden in an internal portion of the article of furniture), or disposed near or adjacent to the article of furniture. The at least one sensor can be configured to detect sensing data associated with a user of the article of furniture. The sensing data can be associated with or indicative of a biological signal of the user. The sensing data can comprise a single biological signal data or a plurality of biological signal data. The sensing data can comprise a single type of biological signal (e.g., sound, vibration, temperature, etc.) or a plurality of different types of biological signal (e.g., sound and vibration, sound and temperature, vibration and temperature, etc.).

[0041] In some embodiments, the system can comprise a controller (e.g., a computer processor) configured to adjust the condition of the article of furniture based at least in part on the sensing data of the at least one sensor. The controller can be configured to generate a decision to adjust (e.g., generate a control signal for adjusting) the condition of the article of furniture substantially in real-time or shortly after detection or generation of the sensing data by the at least one sensor. In some cases, the duration or time difference (e.g., a short span of time) between when such decision is made by the controller and when the sensing data is detected or generated by the at least one sensor can be at least or at most about 1 second, at least or at most about 2 seconds, at least or at most about 5 seconds, at least or at most about 10 seconds, at least or at most about 20 seconds, at least or at most about 30 seconds, at least or at most about 1 minute, at least or at most about 2 minutes, at least or at most about 5 minutes, at least or at mostabout 10 minutes, at least or at most about 15 minutes, at least or at most about 20 minutes, at least or at most about 25 minutes, at least or at most about 30 minutes, at least or at most about 40 minutes, at least or at most about 50 minutes, at least or at most about 60 minutes, at least or at most about 1.5 hours, or at least or at most about 2 hours prior to when the decision is generated. Alternatively, the decision to adjust the condition of the article of furniture by the controller and the detection of the sensing data by the at least one sensor may not and need not occur in real-time or within a short span of time between one another as described herein.

[0042] In some embodiments, the controller can be configured to adjust the condition of the article of furniture based on (i) the sensing data (e.g., a current sensing data detected and generated substantially in real-time or within the short span of time) and (ii) a control data associated with a human condition (e.g., sleep condition, such as snoring or apnea). The control data can be a ground truth data associated with the human condition. In some cases, the ground truth data can be a positive ground truth data, such that a similarity (e.g., in profile) between the user’s sensing data and the positive ground truth data can be indicative of a presence of the human condition in the user. In some cases, the ground truth data can be a negative ground truth data, such that a difference (e.g., in profile) between the user’s sensing data and the positive ground truth data can be indicative of a presence of the human condition in the user.

[0043] In some embodiments, upon determining that the user may experience a target human condition (e.g., a target sleep condition, such as snoring, sleep apnea, etc.), the controller can make the decision to adjust the condition of the article of furniture, e.g., to treat or ameliorate such undesired condition of the user.

[0044] In some embodiments, the human condition can be a desirable condition. In some embodiments, the human condition can be an undesirable condition. In some cases, the undesirable condition can comprise a sleep disorder as provided herein.

[0045] In some embodiments, the target human condition can be a sleep disordered breathing (e.g., snoring, sleep apnea, hypopnea, etc.), and vibrating the bed (e.g., temporarily, in a specific vibration pattern, etc.) in real-time or adjusting an angle of the article of furniture in real-time (e.g., temporarily or permanently) can trigger the user to move and change to a position that can result in reduced degree of the sleep disordered breathing.

[0046] In some embodiments, the ground truth data can comprise at least a portion of sensing data collected during the current usage of the article of furniture. For example, sensing data collected earlier (e.g., at least about 10 minutes, at least about 20 minutes, at least about 30 minutes, at least about 1 hours, at least about 1.5 hours, at least about 2 hours, at least about 3 hours, etc., or during the beginning of the user’s usage such as sleep) than the currently detectedsensing data can be utilized as the control data.

[0047] In some embodiments, the ground truth data can comprise historical sensing data collected from the user, e.g., a historical sensing data of the user from at least one prior and separate usage of the article of furniture or a comparable equivalent thereof. The historical sensing data can be collected during one or more previous days by the at least one sensor operatively coupled to the article of furniture (e.g., a sensor in the article of furniture) or a different sensor associated with the user (e.g., a wearable sensor such as a wearable watch such as Apple Watch, Galaxy Watch, Pixel Watch, Fitbit, etc.).

[0048] In some embodiments, the historical sensing data can be a portion (e.g., a sub-section) of historical sensing data associated with the user. The controller can be configured to analyze (e.g., automatically analyze) the historical sensing data and identify and / or extract (e.g., automatically identify and / or extract) the portion of the historical sensing data that can be usable as the ground truth data. In some embodiments, the user can provide (e.g., label) an input to the controller via a user device (e.g., via a graphical user interface (GUI) of a user application on the user device, such as a mobile phone) as to when (e.g., date and / or time of sleep) the user suspects the user has experienced the human condition. In some cases, the article of furniture can be a bed device, and the user can indicate (e.g., via the GUI or any other user device in the bed device or operatively coupled thereto) that they have experienced the human condition (e.g., as to when they have snored or experienced sleep apnea). The user can provide such indication while on the bed, e.g., in the middle of the night or upon waking up in the morning, or after getting out of the bed. In some embodiments, the controller can be configured to display the user’s historical sensing data (e.g., heart signal data such as HRV or breathing signal data) via the GUI, and the user can provide (e.g., label) which portion of the historical sensing data corresponds to the human condition that the user may have experienced. Based on the user’s input, the controller can be configured to identify and / or extract the portion of the historical sensing data that can be usable as the ground truth data. In some embodiments, any information or indication about the user’s human condition (e.g., snoring, sleep apnea) can be provided by the user or an additional user. The additional user can be a different user of the same article of furniture (e.g., a partner, a spouse, a family member, etc.). Alternatively, the additional user may not need to be a user of the same article of furniture, but is capable of recognizing that the user has experienced the human condition while the user is using the article of furniture. Yet in another alternative, the additional user may be a third party who has access to the user’s historical sensing data, wherein the third party can manually identify and label one or more portions of the historical sensing data as an instance of a target human condition.

[0049] In some embodiments, the historical sensing data can be a computed data (e.g., an average value or profile, a median value or profile) generated from a plurality of data collected an extended period of time, such as over at least or up to about 1 hour, at least or at most about 2 hours, at least or at most about 5 hours, at least or at most about 10 hours, at least or at most about 15 hours, at least or at most about 20 hours, at least or at most about 25 hours, at least or at most about 30 hours, at least or at most about 40 hours, at least or at most about 50 hours, or at least or at most about 100 hours. In some embodiments, the historical sensing data can be a computed data (e.g., an average value or profile, a median value or profile) generated from a plurality of data collected over a plurality of days (e.g., consecutive or non-consecutive days), such as over at least or at most about 2 days, at least or at most about 3 days, at least or at most about 4 days, at least or at most about 5 days, at least or at most about 6 days, at least or at most about 7 days, at least or at most about 8 days, at least or at most about 9 days, at least or at most about 10 days, at least or at most about 15 days, or at least or at most about 20 days.

[0050] In some embodiments, the ground truth data can comprise a reference data that is not associated with the user. In some cases, the reference data can be associate with (e.g., collected from) another individual or another cohort of individuals. The reference data can be collected while using a comparable article of furniture as that of the user, or can be collected without using such article of furniture. The reference data can be collected by a user device (e.g., a smart watch as provided herein) or a medical device. For example, the reference data can be clinical data collected from a cohort of individuals. In some cases, the reference data can be an artificial data that is not collected from any specific individual. For example, the reference data can comprise predicted or hypothetical data of one or more biological signals (e.g., to be utilized as pseudoground truth data).

[0051] In some embodiments, the controller can be configured to adjust the condition of the article of furniture based on the sensing data (e.g., a current sensing data detected and generated substantially in real-time or within the short span of time), in absence of any control data associated with the human condition.

[0052] In some embodiments, adjustment of the condition of the article of furniture by the controller (e.g., based on analysis of the sensing data, with or without use of the control data such as ground truth data) can be based on a rule-based algorithm, e.g., based on a set of predetermined characteristics of the sensing data (e.g., predetermined patterns or ranges) to determine presence / absence of the human condition and make a decision on adjusting the condition of the article of furniture.

[0053] In some embodiments, the controller can utilize at least one classifier (e.g., at least orat most about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more classifiers) for making the decision as provided herein, e.g., comparing the sensing data and the control data. As provided herein, using the control data or making a decision based at least in part on the control data can comprise using the at least one classifier. A classifier can be configured to receive an input comprising at least the sensing data, analyze the input, and provide an output comprising, but not limited to, (i) a decision to adjust the condition of the article of furniture, (ii) a target condition of the target of furniture (e.g., target temperature, target vibration state, etc.), (iii) a target time to implement the adjustment of the condition of the article of furniture, and / or (iv) a target duration of the implemented change of the condition of the article of furniture.

[0054] In some embodiments, the classier as provided herein can be trained via supervised learning. For example, the control data (e.g., pressure data such as audio data, piezoelectric sensor data, etc.) can be labeled (e.g., via human analysis) as ground truth data. In some embodiments, the classifier as provided herein can be trained via unsupervised learning.

[0055] In some embodiments, the sensing data and the control data (e.g., the ground truth data) can be same types of data (e.g., same pressure data, such as same data collected from one or more piezoelectric sensors).

[0056] In some embodiments, the sensing data and the control data (e.g., the ground truth data) can be different types of data, such as different types of biological signals obtained from different types of sensors. For example, the sensing data can be a first type of biological signal selected from the group comprising heart signal, motion, temperature, movement, perspiration, sound, and / or neural activity, and the control data can be a second and different type of biological signal selected from the same group comprising heart signal, motion, temperature, movement, perspiration, sound, and / or neural activity. Accordingly, the classifier can be configured to (e.g., can be trained to) analyze the sensing data of the first type of biological signal and convert (or transform) at least a portion of the sensing data into a converted sensing data of the second type of biological signal. Subsequently, the controller (e.g., via the classifier or a different classifier) can compare the converted sensing data and the control data, which are now of the same type of data, and provide any one of the output (i) through (iv) as aforementioned. Alternatively, the classifier can be configured to convert (or transform) at least a portion of the sensing data into a converted sensing data of the third data type (or a common data type) that is different from (i) the data type of the sensing data and (ii) the data type of the control data.

[0057] In some embodiments, training the classifier as provided herein can comprise applying the control data (or ground truth data) to at least one machine learning algorithm comprising a first plurality of dimensions (e.g., a machine learning architecture such asconvolutional neural network). In some cases, the classifier can be configured to convert any input data (e.g., the sensing data of the first type) into an additional data type comprising (or consisting of) a smaller number of dimensions than the first plurality of dimensions (e.g., filtering out one or more nodes associated with noise or background data). In some cases, the machine learning architecture may not be a deep neural network.

[0058] In some embodiments, the classifier as provided herein can be trained to identify (e.g., extract, isolate, etc.) one or more features that can be commonly identifiable in both the control data and the sensing data. The one or more common features can comprise at least 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more common feature(s). The one or more common features can comprise at most 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 common feature(s). In some embodiments, the control data and / or the sensing data can comprise or can be indicative of a vibration data (e.g., audio data and / or pressure data as provided herein can be indicative of a vibration data), and non-limiting examples of the vibration data can include intensity and / or frequency of the vibration data.

[0059] In some embodiments, the control data can be associated with (or can be) an audio data (e.g., as measured by a microphone), and the sensing data can be associated with (or can be) a non-audio data, such as, for example, pressure, acceleration, temperature, strain, and / or force data (e.g., as measured by a piezoelectric sensor or a different pressure sensor of the article of furniture). One or more classifiers can be utilized to identify and transform at least a portion of the non-audio sensing data (e.g., pressure / compression data) into predicted audio sensing data, such that the predicted audio sensing data can be compared to the control audio data, e.g., to determine whether the user is experiencing a target human condition (e.g., a sleep disorder such as snoring) and make a decision as to if, when, and / or how to adjust the condition of the article of furniture (e.g., vibrate at least a portion of the article of furniture, adjust an shape or angle of the article of furniture, adjust temperature, etc.).

[0060] In some embodiments, the control data can be associated with (or can be) a respiratory signal data (e.g., respiratory data from a respiratory belt, heart rate and / or oxygen level data from a pulse oximeter), and the sensing data can be associated with (or can be) a different data type, such as, for example, pressure, acceleration, temperature, strain, and / or force data (e.g., as measured by a piezoelectric sensor or a different pressure sensor of the article of furniture). One or more classifiers can be utilized to identify and transform at least a portion of the sensing data (e.g., pressure / compression data) into predicted respiratory signal sensing data, such that the predicted respiratory signal sensing data can be compared to the control respiratory signal data, e.g., to determine whether the user is experiencing a target human condition (e.g., a sleep disorder such as sleep apnea or hypopnea) and make a decision as to if, when, and / or howto adjust the condition of the article of furniture (e.g., vibrate at least a portion of the article of furniture, adjust an shape or angle of the article of furniture, adjust temperature, etc.).

[0061] In some embodiments, the control data can be obtained via one or mor polysomnogram (PSG) devices configure to utilize, for example, electroencephalogram, electrooculogram, electromyogram, electrocardiogram (e.g., to measure heart rate), pulse oximetry (e.g., percentage of oxygen in one’s blood or “SpO2”), and / or airflow and respiratory effort (e.g., to measure respiratory rate), to biological signals associated with sleep disorders such as sleep apnea or hypopnea. The control data from each control subject can be analyzed (e.g., by a trained technician) to identify or grade sleep disorder events. Subsequently, a classifier (e.g., a machine learning algorithm) can be trained on the control data and / or the analysis thereof for identifying or grading the sleep disorder events, such that the classifier can be configured to (1) receive biological signal data that is of a different data type (e.g., different sensor data type) than any of the control data and (2) determine what the sleep disorder event the received biological signal data would correspond to. For example, when a user sleeps on the article of furniture as provided herein and the user’s biological signal data from a pressure sensor such as a piezoelectric sensor is obtained and fed in to the classifier, the classifier can (1) transform at least a portion of the pressure sensor data into a predicted data of the same type as the control data utilized for training the classifier and (2) analyze the transformed data to predict (e.g., with a probability score) whether the user is experiencing the sleep disorder. In another example, the classifier can be trained to transform both the control data and the pressure sensor data (e.g., as measured by the article of furniture) into a different data type (e.g., a computer generated data type), and compare the control data and the pressure sensor data in the language of the different data type to predict whether the user is experiencing the sleep disorder.

[0062] In some embodiments, upon adjusting the condition of the article of furniture based at least in part on identifying a target human condition of the user, the controller can be configured to monitor any change in the target human condition of the user (e.g., via continuing to monitor subsequent sensing data associated with the user and comparing it to the control data). The controller can be configured to change in real-time the adjusted condition of the article of furniture (e.g., reduce vibration, reduce tilt angle, etc.) based on the detected change in the target human condition subsequent to the initial adjustment of the condition. For example, vibration strength can be reduced or changed (e.g., to a different vibration pattern) when the controller determines that the degree of the target human condition (e.g., sleep disordered breathing) has been diminished, or can be turned off when the controller determines that the target human condition is no longer detectable.

[0063] In some embodiments, the controller can be configured to (e.g., via use of at least one classifier) determine (i) a number of users (e.g., one or two users) currently present on or adjacent to the article of furniture (e.g., on top of a bed device), (ii) an approximate or substantially precise location of each user with respect to the article of furniture (e.g., on the left side of the bed device, on the right side of the bed device, etc.), and / or (iii) identity of each user where the identity is based on information about the user previously stored in a database.

[0064] In some embodiments, a sensing data from a sensor can be a collection of data (e.g., a collection of biological signals) measured from two or more users simultaneously. For example, data from a sensor of the article of furniture (e.g., a pressure sensor such as a piezoelectric sensor) can have signals (e.g., movement signals or heart signals) at least partially overlapping one another. In such case, the controller can be configured to (e.g., via use of at least one classifier) identify and isolate two or more separate sensing data associated with the two or more users. For example, a first user may be closer to the sensor than the second user, and the relative intensity of biological signals of the two users as detected by the sensor can be different (e.g., the detected biological signal from the first user may exhibit a stronger signal intensity than that from the second user). The controller (e.g., via the classifier) can be configured to identify and isolate two separate biological signal sensing data based at least in part on the different intensities. The system can comprise a plurality of sensors at different locations relative to the article of furniture (e.g., a first sensor closer to the first user and another sensor closer to the second user), and the controller (e.g., via the classifier) can utilize (e.g., analyze and compare) sensing data from the plurality of sensors to identify and isolate a plurality of separate biological signal sensing data respectively for each of the plurality of users. In some cases, a classifier (e.g., machine learning algorithm) is trained on a plurality of different sensing data collected by one or more sensors of the article of furniture, such as pressure sensor data (e.g., piezoelectric sensor data), capacitive sensor data, and / or temperature sensor (e.g., thermistor sensor) data. The classifier can be trained on large datasets (e.g., from a large group of users with single and multiple occupancy examples, such as double occupancy examples) with truth labels, to be able to classify presence of one or more users (or absence thereof) on the article of furniture and location of the one or more users on the article of furniture (e.g., classify whether a user is on a left side of the article of furniture or a right side of the article of furniture).

[0065] In some embodiments, the controller can be configured to store and retrieve an adjustment profile of the condition of the article of furniture for the user. For example, the article of furniture can be a bed device, and the adjustment profile can be a schedule of predetermined adjustments of the condition (e.g., temperature) of the bed device throughout a use (e.g., a singleuse such as for sleeping at night, etc.). The predetermined adjustments of the condition can comprise varying degree of the condition (e.g., varying target temperatures) and / or varying duration of the adjusted condition (e.g., duration of a change in temperature of the article of furniture). In some cases, the controller can receive an input from the user (e.g., the GUI) about a user event, and the controller can be configured to retrieve an adjustment profile associated with the user, and modify the adjustment profile based on the user event to generate a modified adjustment profile that is more suitable for such user event (e.g., via one or more classifiers as provided herein). Alternatively or in addition to, the controller can be in digital communication with a database associated with the user, and the controller can automatically retrieve one or more user events from the database. Such database can be associated with a digital calendar of the user. The modified adjustment profile can be stored and labeled with the corresponding user event, such that the controller can retrieve it for future uses. In some embodiments, once the modified adjustment profile is utilized to modify the condition of the article of furniture, the controller can be configured to monitor quality of use of the article of furniture (e.g., sleep quality of the user), and further adjust the modified adjustment profile for continued learning and optimization of the adjustment profile for the specific user event.

[0066] In some embodiments, the adjustment profile as provided herein can be modified (e.g., automatically modified) based on one or more user events as provided herein.

[0067] In some embodiments, the adjustment profile can comprise one or more adjustments of a condition of the article of furniture and, optionally, a duration or time of onset for each adjustment of the one or more adjustment(s), such as at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 25, 30, 35, 40, 45, 50, or more adjustments, or at most about 50, 45, 40, 35, 30, 25, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 adjustment(s).

[0068] In some embodiments, the adjustment profile can comprise one or more adjustments of the condition of the article of furniture corresponding to (or associated with) one or more target sleep phases (e.g., REM sleep, deep sleep, light sleep, etc.). In some cases, the adjustment profile can comprise a target time of onset of the target condition for a respective target sleep phase, e.g., a target time of onset that is predicted based on one or more prior sleep session of the user and / or of a cohort of individuals (with or without the user). Alternatively or in addition to, the adjustment profile can comprise a target duration of the target condition for the respective target sleep phase, e.g., based on one or more prior sleep session of the user and / or of a cohort of individuals (with or without the user). Alternatively or in addition to, the onset of the execution (or implementation) of the target condition for the respective target sleep phase from theadjustment profile can be based on detecting (e.g., in real-time or substantially in real-time) such target sleep phase of the user while the user is sleeping, e.g., based on one or more biological signals of the user detected while the user is sleeping. For example, a predetermined target temperature of the article of furniture for REM sleep (or deep sleep) from the adjustment profile can be implemented (e.g., by a temperature regulating unit coupled to the article of furniture) upon detecting (e.g., automatically detecting) that the user is or has entered into REM sleep (e.g., based on analysis of one or more biological signals detected by a user sensor associated with the article of furniture).

[0069] In some embodiments, the condition of the article of furniture can be a temperature of the article of furniture, and modifying the adjustment profile can comprise modifying (e.g., increasing or decreasing) one or more target temperatures (e.g., a target temperature corresponding to a specific sleep phase) by at least about 0.1 degrees Celsius (°C), 0.2°C, 0.3 °C, 0.4°C, 0.5°C, 0.6°C, 0.7°C, 0.8°C, 0.9°C, 1°C, 2°C, 3°C, 4°C, 5°C, 6°C, 7°C, 8°C, 9°C, 10°C, 11°C, 12°C, 13°C, 14°C, 15°C, 20°C, 25°C, 30°C, 35°C, 40°C, 45°C, 50°C, or more; or by at most about 50°C, 49°C, 48°C, 47°C, 46°C, 45°C, 40°C, 35°C, 30°C, 25°C, 20°C, 15°C, 14°C, 13°C, 12°C, 11°C, 10°C, 9°C, 8°C, 7°C, 6°C, 5°C, 4°C, 3°C, 2°C, 1°C, 0.9°C, 0.8°C, 0.7°C, 0.6°C, 0.5°C, 0.4°C, 0.3°C, 0.2°C, 0.1°C, or less. Alternatively or in addition to, the modifying of the adjustment profile can comprise modifying (e.g., increasing or decreasing) the one or more target temperatures (e.g., the one or more existing target temperatures of the adjustment profile) by at least about 0.1%, 0.2%, 0.3%, 0.4%, 0.5%, 0.6%, 0.7%, 0.8%, 0.9%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%, 11%, 12%, 13%, 14%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 60%, 70%, 80%, 90%, 100%, 150%, 200%, 250%, 300%, 400%, 500%, or more; or by at most about 500%, 400%, 300%, 250%, 200%, 150%, 100%, 90%, 80%, 70%, 60%, 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 14%, 13%, 12%, 11%, 10%, 9%, 8%, 7%, 6%, 5%, 4%, 3%, 2%, 1%, 0.9%, 0.8%, 0.7%, 0.6%, 0.5%, 0.4%, 0.3%, 0.2%, 0.1%, or less.

[0070] In some embodiments, modifying of the adjustment profile can comprise modifying (e.g., lengthening or shortening) the respective duration of the adjustment of the condition of the article of furniture. The respective duration of the adjustment can be modified (e.g., lengthened or shortened) by at least about 1%, 2%, 3%, 4%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 60%, 70%, 80%, 90%, 100%, 150%, 200%, or more, or by at most about 200%, 150%, 100%, 90%, 80%, 70%, 60%, 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, 5%, 4%, 3%, 2%, 1%, or less.

[0071] In some embodiments, modifying of the adjustment profile can comprise modifying (e.g., increasing or decreasing) a total number of the plurality of adjustments within theadjustment profile. The total number of the plurality of adjustments within the adjustment profile can be changed (e.g., increased or decreased) by at least about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more adjustment(s), or at most about 10, 9, 8, 7, 6, 5, 4, 3, 2, or 1 adjustment(s).

[0072] In some embodiments, modifying of the adjustment profile can comprise modifying (e.g., increasing or decreasing) a target level of the condition of the article of furniture (e.g., modifying a target temperature of a zone of a bed device). The target level of the condition of the article of furniture can be modified (e.g., lengthened or shortened) by at least about 1%, 2%, 3%, 4%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 60%, 70%, 80%, 90%, 100%, 150%, 200%, or more, or by at most about 200%, 150%, 100%, 90%, 80%, 70%, 60%, 50%, 45%, 40%, 35%, 30%, 25%, 20%, 15%, 10%, 5%, 4%, 3%, 2%, 1%, or less.

[0073] Non-limiting examples of the user event can include whether the use of the article of furniture is for a night sleep, whether the use of the article of furniture is for napping, whether the use of the article of furniture is for overcoming jetlag, a particular calendar event (e.g., an appointment or a meeting to attend to, an alarm for waking up, etc.) scheduled to occur after the use of the article of furniture, or any daily event that has occurred on that day prior to use of the article of furniture, etc. Non-limiting examples of the daily event can include whether the user has exercised (e.g., exercise pattern of the user prior to the use of the article of furniture), food or drink consumption (e.g., meals, coffee, dessert, fruits, vegetables, etc.), etc. In some embodiments, the user can provide (e.g., type in) the user event or the daily event via the GUI associated with the controller. In some embodiments, the GUI can present a list of preset user events or daily events (e.g., notations or tabs) for the user to select from. In such case, the GUI can also permit the user to update the list to include any new user event or daily event. In some embodiments, the controller (e.g., via the classifier) can analyze the selected user events and / or daily events along with the sensing data and / or adjusted condition of the article of furniture, e.g., to look for correlations between such factors and generate a new adjustment profile of the condition of the article of furniture for the user that is better optimized for better usage (e.g., better sleep).

[0074] In some embodiments, non-limiting examples of the exercise pattern can comprise duration and / or frequency of walking, running, swimming, basketball, baseball, hockey, tennis, gymnastics, standing for duration of time, etc. In some embodiments, non-limiting examples the food consumption data can comprise types of foods consumed by the user (e.g., elementary foods, pre-packaged meals, home-cooked meals, fruits, vegetables, etc.), amounts of foods consumed by the user, frequency of food consumption by the user, and / or time of the food consumption within the day.

[0075] In some embodiments, the controller can be configured to modify a condition of the article of furniture in accordance with the user event, regardless of whether or not the condition of the article of furniture is not being controlled based on a pre-generated adjustment profile. For example, even when the condition of the article of furniture is not being controlled based on any pre-generated adjustment profile, the controller can be configured to modify a condition of the article of furniture in accordance with the user event.

[0076] In some embodiments, the controller can be configured to monitor the duration of current usage of the article of furniture by the user. In some cases, the controller can be configured to determine how long the user has been on the article of furniture (e.g., bed device) during the current use. For example, by using the sensing data, the controller can determine whether the user is sleeping on the bed device (e.g., by determining one or more sleep phases of the user), and the controller can further determine how long the user has been sleeping on the bed device.

[0077] In some embodiments, the controller can be configured to determine (e.g., identify) and monitor a sleep phase or sleep pattern of the user while the user is using the article of furniture. The controller can be further configured to adjust the condition of the article of furniture based on the sleep phase or sleep pattern. In some cases, the controller can be configured to adjust the condition of the article of furniture upon determining that, for example, (i) the user has experienced a target sleep phase, (ii) the user has experienced the target sleep phase for a target duration of time, and / or (iii) a current sleep phase of the user.

[0078] In some embodiments, the article of furniture can be a bed device, and the controller can be configured to adjust the condition of the bed device based on two or more members selected from: duration of current usage of the bed device, duration of one or more sleep phases of the user (e.g., a total duration of sleep, a duration of a particular sleep such as lights sleep or REM sleep, etc.) during the current usage of the bed device, and one or more user events as provided herein. In some embodiments, the adjustment of condition can be utilized as an adaptive alarm to wake up the user based on the two or more members.

[0079] In some embodiments, the controller can be configured to activate the thermal alarm (e.g., determine a target temperature of the bed device and a target time to adjust the temperature of the bed device substantially to the target temperature) based at least in part on (i) a time of the first appointment in the user’s calendar for the day, (ii) a duration of the user’s current sleep on the bed device, (iii) the user’s current sleep phase, (iv) the current time, and / or (v) a wake-up time that is pre-determined or pre-set by the user. Accordingly, the user can be woken up via the thermal alarm at a time that is different (e.g., earlier) than the user’s predetermined wake-uptime, at a target time that is sufficiently before the time of the first appointment, at the target time that is after the user has been sleeping for at least a threshold amount of time (e.g., at least about 3, 4, 5, 6, 7, 8, or more hours), and while the user is in a target sleep phase (e.g., light sleep). In some cases, taking into account both the duration of the user’s current sleep and the target sleep phase can ensure that the user is not woken up at any occurrence of the target sleep phase, but only during the occurrence of the target sleep that is after the user has slept for a certain amount of time.

[0080] In some embodiments, operation of the adaptive alarm can be by temperature adjustment, vibration, sound, visual cue, audiovisual cue, or any combination thereof. The adaptive alarm of the article of furniture (e.g., bed device) may be a vibration adaptive alarm. The adaptive alarm described herein can be based on one or more user factors and / or one or more sleep factors. In some embodiments, one or more adaptive alarms may be used for one or more users. For example, a first adaptive alarm may be assigned to a first user and may be based on one or more first user factors and / or one or more first sleep factors. A second adaptive alarm may be assigned to a second user and may be based on one or more second user factors and / or one or more second sleep factors. As such, an article of furniture (e.g., bed device) described herein can comprise at least about 1, 2, 3, 4, 5, or greater than about 5 alarms (e.g., adaptive alarms). In some embodiments, an article of furniture (e.g., bed device) described herein can comprise no alarm (e.g., adaptive alarm).

[0081] The user factor can comprise a sleep duration (e.g., a desired sleep duration) for a use of the article of furniture. In some embodiments, the user factor can comprise a sleep duration (e.g., a desired sleep duration) for a night (e.g., a sleep session) of the article of furniture. In some embodiments, the user factor can comprise a sleep duration (e.g., a desired sleep duration or target sleep duration) for a use and / or night (e.g., sleep session) of the article of furniture. The user factor may be provided by the user. The user factor can be provided by the user to the article of furniture or a controller coupled thereof. In some embodiments, the user factor can be provided by the user prior to the use of the article of furniture (e.g., bed device). In some embodiments, the user factor can be provided by the user during the use of the article of furniture (e.g., bed device). In some embodiments, the user factor can be provided by the user via graphical user interface (GUI).

[0082] In some embodiments, the user factor can be provided by the user at least about 1 second, at least about 5 seconds, at least about 10 seconds, at least about 15 seconds, at least about 30 seconds, at least about 1 minute, at least about 5 minutes, at least about 15 minutes, at least about 30 minutes, at least about 45 minutes, at least about 1 hour, at least about 2 hours, atleast about 3 hours, at least about 4 hours, at least about 5 hours, at least about 6 hours, at least about 12 hours, or greater than about 12 hours prior to the use of the article of furniture (e.g., bed device). In some embodiments, the user factor can be provided by the user at most about 12 hours, at most about 6 hours, at most about 5 hours, at most about 4 hours, at most about 3 hours, at most about 2 hours, at most about 1 hour, at least about 45 minutes, at least about 30 minutes, at least about 15 minutes, at least about 5 minutes, at least about 1 minute, at least about 30 seconds, at least about 15 seconds, at least about 10 seconds, at least about 5 seconds, at least about 1 second, or less than about 1 second prior to the use of the article of furniture (e.g., bed device).

[0083] The sleep factor can comprise a duration of time. For example, the sleep factor can comprise a duration of time that the user transitions from being awake to being asleep. A time when the user falls asleep can comprise the user falling asleep on the article of furniture (e.g., bed device). As another example, the sleep factor can comprise a duration of time that the user begins a use of the article of furniture (e.g., bed device). In some embodiments, a time when the user falls asleep can be the same as a time that the user begins a use of the article of furniture (e.g., bed device). In some embodiments, a time when the user falls asleep can be different than a time that the user begins a use of the article of furniture (e.g., bed device).

[0084] A controller can set a “time zero” for the duration of time (e.g., a duration of time for an alarm described herein). The controller can set the “time zero” for a user’s sleep based on a sleep factor. The controller can set the “time zero” for a user’s sleep based on a user factor. In some embodiments, the controller can set the “time zero” for a user’s sleep based on a time when the user falls asleep. A controller can track and / or determine when a user reaches a sleep duration (e.g., a desired sleep duration or target sleep duration). A controller can track and / or determine when a user is expected to reach a sleep duration (e.g., a desired sleep duration or target sleep duration). For example, a controller can track and / or determine when a user is at least about 1 second, at least about 5 seconds, at least about 10 seconds, at least about 15 seconds, at least about 30 seconds, at least about 1 minute, at least about 5 minutes, at least about 10 minutes, at least about 15 minutes, at least about 30 minutes, or at least about 1 hour within reaching a sleep duration (e.g., a desired sleep duration or target sleep duration). The controller can track and / or determine a time when a user reaches or is expected to reach a sleep duration (e.g., a desired sleep duration or target sleep duration) during a use (e.g., current use) of the article of furniture described herein. In some embodiments, the controller may pause tracking the duration of time if or when the user wakes up prior to the user reaching the sleep duration (e.g., a desired sleep duration or target sleep duration). In some embodiments, the controller may notpause tracking the duration of time if or when the user wakes up prior to the user reaching the sleep duration (e.g., a desired sleep duration or target sleep duration).

[0085] A controller can turn on an alarm (e.g., adaptive alarm) at the time when the user reaches or is expected to reach the sleep duration (e.g., a desired sleep duration). Turning on the alarm may comprise inducing a change in at least a portion of the article of furniture. Turning on the alarm may comprise a temperature adjustment, vibration, audio stimulus, visual stimulus, audiovisual stimulus, or any combination thereof. In some embodiments, a controller can turn on an alarm (e.g., adaptive alarm) prior to the time when the user reaches or is expected to reach the sleep duration (e.g., a desired sleep duration). In some embodiments, a controller can turn on an alarm (e.g., adaptive alarm) after the time when the user reaches or is expected to reach the sleep duration (e.g., a desired sleep duration or target sleep duration). The controller may turn on the alarm (e.g., adaptive alarm) at most about 1 hour, at most about 30 minutes, at most about 20 minutes, at most about 15 minutes, at most about 10 minutes, at most about 5 minutes, at most about 4 minutes, at most about 3 minutes, at most about 2 minutes, at most about 1 minute, at most about 45 seconds, at most about 30 seconds, at most about 20 seconds, at most about 10 seconds, at most about 5 seconds, at most about 4 seconds, at most about 3 seconds, at most about 2 seconds, at most about 1 second, or less than about 1 second after the time when the user reaches or is expected to reach the sleep duration (e.g., a desired sleep duration or target sleep duration).

[0086] As an example, the controller can (i) set “time zero” for the user’s sleep based on the sleep factor (e.g., a time at which a user falls asleep), (ii) track and / or determine when the user reaches or is expected to reach a desired sleep duration (e.g., from the set “time zero”) for the current use of the article of furniture, and (iii) turn on the adaptive alarm at, upon, or shortly thereafter the time that the user reaches or is expected to reach the desired sleep duration. Such time that the adaptive alarm is turned on may or may not be the same as the user’s predetermined (or preset) alarm time. In some embodiments, a user may preset a duration of time (e.g., a desired sleep duration or target sleep duration) of 8 hours prior to a sleep session or use of the article of furniture. This duration of time may be a “time zero”. The “time zero” may start when the user falls asleep (e.g., falls asleep on the article of furniture). The controller may track the duration of time (e.g., the 8 hours) as the user uses the article of furniture (e.g., as the user is asleep). The controller may turn on the alarm (e.g., adaptive alarm) once the 8 hours has surpassed. Alternatively, the controller may turn on the alarm between 1 second to 30 minutes prior to the 8 hours being surpassed. The controller may turn on the alarm between 1 second to 30 minutes after the 8 hours being surpassed.

[0087] The alarm described herein may be turned off. In some embodiments, the alarm can be turned off automatically. For example, the alarm can be turned off automatically when the user ceases using the article of furniture (e.g., bed device). As another example, the alarm can be turned off automatically after a duration of time (e.g., at least about or at most about 1 second, 5 seconds, 10 seconds, 15 seconds, 30 seconds, 1 minute, 5 minutes, 15 minutes, 30 minutes, 45 minutes, or 1 hour). The duration of time may be measured from a time point when the user may be awake. The alarm may be turned off automatically by reverting at least a portion of a change in the article of furniture that was induced when turning on (e.g. initiating) the alarm. In some embodiments, the alarm can be turned off after the user’s sleep duration (e.g., desired sleep duration or target sleep duration). The user’s sleep duration can be set by the graphical user interface. In some embodiments, the alarm may be turned off by the user’s input. For example, the alarm may be turned off by the user via the graphical user interface.

[0088] In some embodiments, the controller can be operatively coupled to a plurality of articles of furniture. The controller can be linked to each article of furniture by, for example, using a GUI of a user device to scan a code (e.g., a machine readable code, such as a barcode) of the article of furniture, or by determining an identify of the user based on sensing data while the user is using the article of furniture. The controller can be configured to adjust conditions of the plurality of articles of furniture. The controller can be configured to receive sensing data from the plurality of articles of furniture, and display at least a portion of the sensing data (e.g., a summary chart, table, graph, text, etc.) or analysis thereof (e.g., sleep score) to a GUI of one or more users associated with the plurality of articles of furniture.

[0089] In some embodiments, a machine readable code as provided herein can be specifically assigned for each article of furniture. The machine readable code can be a linear code or a multidimensional code (e.g., a two-dimensional code). Non-limiting examples of the machine readable code can include Australia Post barcode, Codabar, Code 25 (interleaved or noninterleaved), Code 11, Code 32 (or Farmacode), Code 39, Code 49, Code 93, Code 128, Digital indeX (DX), European Article Numbers (EAN), Facing Identification Mark, Intelligent Mail barcode, Interleaved 2 of 5 (ITF), Modified Plessey, Pharmacode, Postal Alpha Numeric Encoding Technique (PLANET), PostBar, Postal Numeric Encoding Technique, ETniversal Product Code (e.g., EIPC-A and UPC-E), etc. Examples of the matrix MRC 1022 include Aztec, ColorCode, Color Construct Code, CrontoSign, CyberCode, d- touch, DataGlyphs, Data Matrix, Datastrip Code, Digimarc Barcode, DotCode, DWCode, EZcode, High Capacity Color Barcode, Han Xin Barcode, HueCode, InterCode, MaxiCode, Mobile Multi-Colored Composite (MMCC), NexCode, PDF417, Qode, Quick Response (QR) code, ShotCode, Snapcode, SPARQCode,VOICEYE, etc.

[0090] In some embodiments, the plurality of articles of furniture can be for a single user, and the user can select a specific article of furniture from the plurality on the GUI, to indicate which article of furniture will be used that day. The plurality of articles can be positioned at different rooms and / or at different locations, and the controller can be configured to implement an adjustment profile of the condition of an article of furniture of the plurality of articles of furniture to one or more other articles of furniture of the plurality. For example, the adjustment profile (e.g., specifically generated for the user) can be stored in a database (e.g., in a user profile in a cloud database), such that the adjustment profile can be implemented to any desired article of furniture of the plurality.

[0091] In some embodiments, the plurality of articles of furniture can be for a designated cohort of users (e.g., a self-designated cohort, such as partners or family members). A first member of the cohort can grant a second member of the cohort access to (i) database comprising sensing data associated with the first member and / or (ii) control of condition of the article of furniture of the first member, e.g., via the GUI associated with the first member. Accordingly, the second member can, e.g., remotely via the GUI associated with the second member, (i) review the sensing data (e.g., current or historical) of the first member and / or (ii) control the condition of the first member’s article of furniture. Accordingly, the second member can have access to a “centralized vision” or “centralized access” to sensing data from the cohort of users and / or operations of the plurality of articles of furniture. For example, the first member can be a child, and he second member can be a parent. In another example, the first member can be an individual in need of medical monitoring or assistance (e.g., a patient, an elderly individual, etc.), and the second member can be a guardian or a caretaker.

[0092] In some embodiments, as provided herein, the controller can be configured to turn on an alarm based at least in part on the sensing data, e.g., to alert the user or wake up the user from sleeping. The alarm can be thermal alarm (e.g., adjustment of temperature of the article of furniture), vibration alarm (e.g., adjustment of vibration of one or more motors of the article of furniture), sound alarm (e.g., activation of sound), etc., or any combination thereof. For example, the alarm can be a combination of thermal alarm and vibration alarm. In some embodiments, the controller can be configured to turn off the alarm based at least in part on the sensing data. For example, the time is before the user’s predetermined alarm time, and the controller can determine that the user is not using the article of furniture (e.g., the user is not on the bed device), and the controller can make a decision to cancel the upcoming alarm (e.g., vibration alarm) based on the determination of the user’s absence. In another example, the time is soon after the user’spredetermined alarm time and the alarm is activated and still on. Upon determining that the user is not using the article of furniture (e.g., the user has woken up early), the controller can make a decision to turn off the alarm accordingly.

[0093] In some embodiments, the operation of the article of furniture for a user as provided herein can comprise controlling a condition associated with the article of furniture, e.g., based on a predetermined adjustment profile of such condition. The predetermined adjustment profile can comprise a target value of the condition associated with a target sleep phase (or a sleep phase of interest) of the user. For example, a computer processor operatively coupled to the article of furniture can operate and execute the predetermined adjustment profile during the user’s sleep session, and upon detecting that the user is in the target sleep phase (or about to enter the target sleep phase), the computer processor can adjust the condition associated with the article of furniture to about the target value of the condition associated with the target sleep phase. The predetermined adjustment profile may or may not change (e.g., automatically updated by a computer processor) throughout use of the predetermined adjustment profile during a sleep session. For example, the predetermined adjustment profile can be changed or updated, and the operation of the article of furniture as provided herein can comprise automatically modifying one or more features of the predetermined adjustment profile, e.g., the target value of the condition associated with the target sleep phase, a duration of maintaining the condition associated with the article of furniture at the target value, a rate of change of a prior value to the target value, etc. In some cases, the condition can be a condition of the article of furniture, such as temperature of the article of furniture (e.g., temperature of a mattress, pillow, or a cover thereof). In some cases, the condition can be a condition of an environment of the article of furniture, such as temperature of a room containing the article of furniture, humidity, oxygen level, light level, noise level, etc.

[0094] In some embodiments, the operation of the article of furniture for the user as provided herein can be controlled based at least in part on analysis of quality of the target sleep phase of the user from a prior sleep session (e.g., based on biological signal data associated with the user that is collected during the prior sleep session by one or more sensors of the article of furniture). In some embodiments, the analysis of the quality of the target sleep phase can comprise determining whether the quality of the target sleep phase of the prior sleep session is below, comparable to, and / or above a predetermined threshold value. The predetermined threshold value can be a preferred amount or duration of a particular sleep phase over the course of a fraction of or the entirety of a sleep session. The predetermined threshold value can be an average value generated from historical data of the user and / or a cohort of additional users (e.g., a cohort of at least or at most about 10, 20, 50, 100, 200, 500, 1,000, 2000, 5,000, 10,000, 20,000,50,000, 100,000, or more users.

[0095] In some embodiments, the operation of the article of furniture can be controlled differently depending on whether the quality of the target sleep phase of the prior sleep session is below, comparable to, and / or above the predetermined threshold value. For example, anything below the predetermined threshold value can be considered a poor (or sub-optimal) sleep phase quality, and anything at or above the predetermined threshold value can be considered a normal sleep phase quality. In some embodiments, the operation of the article of furniture can be controlled when the quality of the target sleep phase of the prior sleep session is below the predetermined threshold value. For example, the target value of the condition associated with the target sleep phase for the current or subsequent sleep session can be modified when the quality of the target sleep phase of the prior sleep session is below the predetermined threshold value. In some embodiments, the operation of the article of furniture may or may not be controlled when the quality of the target sleep phase of the prior sleep session is comparable to and / or above the predetermined threshold value. For example, the target value of the condition associated with the target sleep phase for the current or subsequent sleep session may or may not be modified when the quality of the target sleep phase of the prior sleep session is comparable to and / or above the predetermined threshold value. In some embodiments, the target value of the condition associated with the target sleep phase for the current or subsequent sleep session may be changed when (i) the quality of the target sleep phase of the prior sleep session is below the predetermined threshold value and (ii) the quality of the target sleep phase of the prior sleep session is comparable to and / or above the predetermined threshold value. For example, the target value may be increased when (i) is met and decreased when (ii) is met, or vice versa. In another example, the target value may be increased to a greater degree when (i) is met as compared to when (ii) is met. In another example, the target value may be decreased to a greater degree when (i) is mt as compared to when (ii) is met.

[0096] A value scale may be based on a non-artificial and / or non-temperature scale. The scale may be user-friendly and easy for a user of the article of furniture to interpret. The scale may be a linear scale. The scale may be a non-linear scale. For example, +1 on the scale may correspond to +1 °C, +2 on the scale may correspond to +2 °C, and +3 on the scale may correspond to +3 °C. The farther from 0 on the scale may correspond to a greater temperature change. For example, +1 on the scale may correspond to +1 °C, +5 on the scale may correspond to +7.5 °C, and / or +10 on the scale may correspond to +13 °C.

[0097] In some embodiments, the quality of the target sleep phase of the prior sleep session can be analyzed (or determined) prior to or during the current or subsequent sleep session. Insome cases, after the user’s prior sleep session but before the user begins the subsequent sleep session, the quality of the target sleep phase of the prior sleep session can be analyzed and the predetermined adjustment profile for the subsequent sleep session can be modified in accordance with the embodiments as provided herein. In some cases, during the user’s current sleep session (e.g., when the user is experiencing the target sleep phase during the current sleep session), the quality of the target sleep phase of the prior sleep session can be analyzed (e.g., substantially in real-time) and the predetermined adjustment profile for the subsequent sleep session can be modified in accordance with the embodiments as provided herein. For example, the user’s current sleep session may be initiated based on the same predetermined adjustment profile as that from the prior sleep session, and when the user is detected to have entered a target sleep phase, the quality of the target sleep phase of the prior sleep session can be analyzed to decide whether or not to change the remaining portion of the predetermined adjustment profile for the current sleep session (e.g., to decide whether or not to adjust a condition associated with the article of furniture substantially in real-time).

[0098] In some embodiments, when the user is having or is suspected of having the same target sleep phase during a current or subsequent sleep session of the user, the condition associated with the article of furniture can be modified, as compared to that from when the user had the same target sleep phase during the prior sleep session, e.g., to enhance quality of the target sleep phase of the user during the current or subsequent sleep session. Such modification of the condition associated with the article of furniture can be performed (e.g., independently) for each type of target sleep phase. The target sleep phase can comprise REM sleep, deep sleep, light sleep, and / or a combination thereof. The prior sleep session can comprise at least or at most about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 20, 25, 30, 40, 50, 60, 70, 80, 90, or 100 days. For example, the prior sleep can comprise (or can consist of) an immediately preceding sleep session of the user (e.g., while using the article of furniture), such as, for example, the day immediately before current or subsequent use of the article of furniture.

[0099] In some embodiments, the analysis of the quality of the target sleep phase (e.g., that from a prior sleep session) as provided herein can be performed for a fraction of or all of the duration of a single sleep session. In some cases, the duration of sleep as provided herein can be substantially the same or shorter than the amount of time the use was using the article of furniture during the single sleep session (e.g., the prior single sleep session). For example, the duration of sleep can exclude any time the user is using the article of furniture but not asleep, e.g., the time before sleep onset, as determined by analyzing the biological signal data of the user measured during such time. The analysis of the quality of the target sleep phase can be performed for (i) atleast or at most about 1, 2, 3, 4, or 5 of first hours of the single sleep session; (ii) at least or at most about 1, 2, 3, 4, or 5 of middle hours of the single sleep session; or (iii) at least or at most about 1, 2, 3, 4, or 5 of last hours of the single sleep session. The analysis of the quality of the target sleep phase can be performed for at least or at most about 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% of the duration of the single sleep session. The analysis of the quality of the target sleep phase can be performed for (i) at least or at most the first 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, or 80% of the duration of the single sleep session; (ii) at least or at most the middle 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, or 80% of the duration of the single sleep session; or (iii) at least or at most the last 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, or 80% of the duration of the single sleep session.

[0100] In some embodiments, the user can experience a plurality of occurrences of the same type of target sleep phase during a single sleep session (e.g., during a prior single sleep session). For example, a sleep cycle can be defined to comprise different sleep stages (e.g., light sleep, deep sleep, and REM sleep), and the user can experience a plurality of such sleep cycles over the course of a single sleep session.

[0101] In some embodiments, the analysis of the quality of the target sleep phase as provided herein can be performed for a fraction of or all of the plurality of occurrences of the target sleep phase from a single sleep session (e.g., a prior single sleep session), such as at least or at most about 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 occurrences of the target sleep phase from a single sleep session. The analysis of the quality of the target sleep phase can be performed for (i) at least or at most about 1, 2, 3, 4, or 5 of first occurrences of the target sleep phase from the single sleep session; (ii) at least or at most about 1, 2, 3, 4, or 5 of middle occurrences of the target sleep phase from the single sleep session; or (iii) at least or at most about 1, 2, 3, 4, or 5 of last occurrences of the target sleep phase from the single sleep session. The analysis of the quality of the target sleep phase can be performed for at least or at most about 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, or 100% of the plurality of occurrences of the target sleep phase from the single sleep session.

[0102] In some embodiments, upon analyzing the quality of the target sleep phase from a prior sleep session, the computer processor can change the condition associated with the article of furniture for the same target sleep phase of the current or subsequent sleep session. The condition can be changed for a fraction of or all of the plurality of occurrences of the same target sleep phase during the current or subsequent sleep session. In some cases, within the current or subsequent sleep session, the condition associated with the article of furniture for an occurrence (first occurrence) of the target sleep phase and that for an additional occurrence (secondoccurrence) of the target sleep phase can be modified to the same target value or different target values. In some cases, the condition associated with the article of furniture for one or more occurrences of the target sleep phase during a portion (first portion) and that for one or more occurrences of the target sleep phase during an additional portion (second portion) can be modified to the same target value or different target values. For example, the first portion can be an early phase of sleep and the second portion can be a late phase of sleep, and a modified target value (e.g., target temperature value of the article of furniture) associated with one or more occurrences of the target sleep phase of the early phase can be different that (e.g., higher than or lower than) a modified target value associated with one or more occurrences of the target sleep phase of the early phase by at least or at most about 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9% 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 120%, 150%, 200%, or more; or at least or at most about 0.1°C, 0.2°C, 0.3°C, 0.4°C, 0.5°C, 0.6°C, 0.7°C, 0.8°C, 0.9°C, 1°C, 1.5°C, 2°C, 3°C, 4°C, 5°C, 6°C, 7°C, 8°C, 9°C, 10°C, 11°C, 12°C, 13°C, 14°C, 15°C, 20°C, or more.

[0103] In some embodiments, the user can be determined to have experienced lower quality of first sleep phase type (e.g., one member from REM sleep, deep sleep, and light sleep) and also lower quality of second sleep phase type (e.g., the other member from REM sleep, deep sleep, and light sleep) during a prior sleep session. Following, when modifying the condition associated with the article of furniture for a current or subsequent sleep session based on such determination, the condition for the first sleep phase type and the condition for the second sleep phase time can be modified similarly or differently. In some cases, a target value of the condition (e.g., target value of the temperature of the article of furniture) for the first sleep phase type can be increased, while the condition for the second sleep phase type can be decreased. In some cases, the target value of the condition for the first sleep phase type can be changed by a first degree or amount, and the target value of the condition for the second sleep phase type can be changed by a second degree or amount that is different than the first degree or amount by at least or at most about 0.5%, 1%, 2%, 3%, 4%, 5%, 6%, 7%, 8%, 9% 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 120%, 150%, 200%, or more. For example, the first sleep phase type can be REM sleep, and the second sleep phase type can be deep sleep and / or light sleep, or vice versa.

[0104] In some embodiments, upon analyzing the quality of the target sleep phase from a prior sleep session, the computer processor can change the condition associated with the article of furniture for the same target sleep phase of the current or subsequent sleep session based on a combination of a plurality of factors comprising (A) the analyzed quality of the target sleep phasefrom the prior sleep session and (B) one or more additional factors as provided herein. The one or more additional factors can comprise at least or at most about 1, 2,3, 4, 5, 6, 7, 8, 9, 10, or more factors, such as any additional member of the user factor, the sleep factor, and / or the environment factor as provided herein. In some cases, the combination of the plurality of factors can comprise a plurality of factors selected from the group consisting of the analysis of the quality of the target sleep phase, age of the user (e.g., whether the user is less than, equal to, or greater than a threshold age), gender of the user (e.g., male or female), geolocation of the user, and health condition of the user. In some cases, the threshold age can be at least or at most about 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, or 90. The threshold age can be about 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80, 85, or 90.

[0105] In some embodiments, the computer system can utilize a decision tree model comprising (i) one of the plurality of factors as a root node, (ii) each of the other(s) of the plurality of factors as an internal node that is branching from the root node or from another internal node, and (iii) a decision to change the condition associated with the article of furniture (e.g., modify a target value of the condition associated with the target sleep phase of the predetermined adjustment profile for the current or subsequent sleep session) as a leaf node. Two or more different leaf nodes (e.g., corresponding to different combinations of the plurality of factors) can comprise different decisions as to the change of the condition of the article of furniture. For example, two or more different leaf nodes can comprise different modifications of the target value of the condition associated with the target sleep phase of the predetermined adjustment profile for the current or subsequent sleep session. Thus, the decision tree model can be utilized to update the predetermined adjustment profile that was utilize during a prior sleep session, to enhance sleep quality of the user (and thus health condition of the user) when the user uses the updated predetermined adjustment profile during a current or subsequent sleep session.

[0106] In some embodiments, the decision tree model can be a classification tree model or a regression tree model. The decision tree model may utilize one or more machine learning algorithms as provided herein.

[0107] In some aspects, provided herein are methods for controlling a condition of an article of furniture. The method can comprise determining a sleep disorder of the user. The sleep disorder can comprise disordered breathing (e.g., apnea and / or snoring). The sleep disorder may be a sleep disorder described herein. The sleep disorder of a user may be determined when a threshold indicator of a data is identified. In some embodiments, the data can be a user sensing data. The user sensing data can be generated be a user sensor. In some embodiments, the user sensing data can be a sensor described herein (e.g., a temperature sensor, a piezoelectric sensor,or any combination thereof). The user sensing data can be generated while a user is using the article of furniture. In some embodiments, the threshold indicator can be based on at least one user factor, wherein the user factor is a user factor described herein. In some embodiments, the user factor can be biological sex, age, weight, health condition, exercise information, diet information (or food consumption information), or any combination thereof. The method can comprise generating an instruction for the article of furniture. The instruction may comprise instruction to change the condition of the article of furniture based on a determined sleep disorder. For example, the condition may be a temperature of the article of furniture (e.g., bed device). As another example, the condition may be an angle of a section of the article of furniture (e.g., bed device).

[0108] Determining the sleep disorder may be performed by a machine learning model. The model can be trained by any training methods described herein. In some embodiments, the model can be trained to identify the threshold indicator from the user sensing data. In some embodiments, determining the sleep disorder may be performed without a machine learning model (e.g., in the absence of the model). Determining the sleep disorder may be based on a time-frequency analysis. The time-frequency analysis can comprise Fourier transformation, Laplace transformation, Wavelet transformation, or any combination thereof. In some embodiments, the time-frequency analysis can comprise Fourier transformation, wherein the Fourier transformation comprise one or more members selected from the group consisting of a Discrete Fourier Transformation (DFT), a Fast Fourier Transformation (FFT), a Principal Component Analysis (PCA) transformation, an Inverse Discrete Fourier Transformation, and an Inverse Fast Fourier Transformation.

[0109] In some embodiments, determining the sleep disorder and generating the instruction for the article of furniture (e.g., bed device) may be performed in real-time. In some embodiments, generating the instruction for the article of furniture (e.g., bed device) may be performed at most about 1 hour, at most about 50 minutes, at most about 40 minutes, at most about 30 minutes, at most about 25 minutes, at most about 20 minutes, at most about 15 minutes, at most about 10 minutes, at most about 9 minutes, at most about 8 minutes, at most about 7 minutes, at most about 6 minutes, at most about 5 minutes, at most about 4 minutes, at most about 3 minutes, at most about 2 minutes, at most about 1 minute, at most about 50 seconds, at most about 40 seconds, at most about 30 seconds, at most about 20 seconds, at most about 10 seconds, at most about 5 seconds, at most about 1 second, or less than about 1 second after determining the sleep disorder. In some embodiments, determining the sleep disorder and generating the instruction for the article of furniture (e.g., bed device) may be performed during asingle use of the user. In some embodiments, determining the sleep disorder and generating the instruction for the article of furniture (e.g., bed device) may be performed over multiple uses of the article of furniture.

[0110] In some embodiments, the methods described herein comprise generating an instruction for the article of furniture to change the condition of the article of furniture. The condition can be a movement of at least a portion of the article of furniture (e.g., bed device). The portion of the article of furniture may be a portion by the head of a user, a portion by the foot of a user, a portion by the chest of a user, a portion by the arm of a user, a portion by the leg of a user, or any combination thereof. The article of furniture can comprise at least one actuator configured to induce the movement. The instruction can comprise instructions for one or more motors to induce the movement. The movement may comprise a vibration. The movement may comprise an angular and / or height adjustment of the portion of the article of furniture. In some embodiments, the movement can comprise both a vibration and an angular and / or height adjustment of the portion of the article of furniture.[OHl] In some embodiments, the methods described herein may comprise determining a change in a target condition of the user. A machine learning model may be used to determine the change in a target condition of the user. In some embodiments, a machine learning model may not be used to determine the change in a target condition of the user. The change in the target condition of the user may be based on an additional user sensing data. The addition user sensing data may be generated by an additional user sensor (e.g., an additional temperature sensor and / or an additional piezoelectric sensor). The methods can comprise generating an additional instruction for the article of furniture to change the condition of the article of furniture based on the change in the target condition of the user.

[0112] In some aspects, the present disclosure provides a system for controlling a condition of an article of furniture. The system can comprise a computer processor and / or a computer memory coupled thereto. The computer memory can comprises a machine executable code. The code may be executed by the processor to implement the methods described herein.

[0113] One or more factors for controlling operation of the article of furniture

[0114] In some embodiments, operation of the article of furniture as provided herein can be controlled based on a user factor, a sleep factor, an environment factor, a combination thereof, and / or an analysis thereof. The user factor, the sleep factor, and / or the environment factor can be collected and stored in a database prior to and / or during the use of the article of furniture by a user.

[0115] In some embodiments, the user factor can comprise one or more members comprising biological sex of the user, age of the user, weight of the user, health condition of the user, exercise information of the user, diet information (or food consumption information) of the user, and / or a combination thereof. In some case, a plurality of members of the user factor and / or a plurality of sub-types of each member of the user factor can be utilized (e.g., analyzed) by the algorithm. The algorithm can be trained to give equal weight or different weights of significance to the plurality of members and / or the plurality of sub-types (e.g., at least or at most about 1%, 2%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, or 99% of difference in significance between two different members and / or between the plurality of sub-types).

[0116] The operation of the article of furniture can be controlled based on the user factor, and optionally the sleep factor and / or the environment factor. The operation of the article of furniture can be controlled based on the sleep factor, and optionally the user factor and / or the environment factor. The operation of the article of furniture can be controlled based on the environment factor, and optionally the user factor and / or the sleep factor.

[0117] In some embodiments, the user factor, the sleep factor, and / or the environment factor can be provided by the user via a portal such as a graphical user interface (GUI) of a user application on a user device (e.g., a computer, a mobile phone, a smart watch, smart glasses, etc.) that is operatively coupled to a computer processor as provided herein.

[0118] In some embodiments, the user factor, the sleep factor, and / or the environment factor (i) may not be personally provided by the user via the portal and (ii) can be collected via one or more sensors operatively coupled to the article of furniture or the environment of the article of furniture. As provided herein, such sensor(s) can include, for example, a capacitance sensor, a temperature sensor, a pressure sensor, a piezoelectric sensor, sound sensor (e.g., a microphone), accelerometer, liquid pressure sensor, etc. The one or more sensors can be provided as part of the article of furniture or as a separate device different from the article of furniture. For example, the article of furniture may be a bed device, and the one or more sensors may be provided as part of (i) the headboard, (ii) sideboard or side wall, (iii) platform for holding a mattress, (iv) a canopy-like structure for defining a small environment (e.g., enclosure) that is surrounding the bed device and is smaller than a size of the room, or a combination thereof. Alternatively or in addition to, any of the sensor as provided herein can be coupled to (or can be a part of) a user device that is worn by the user, such as a smart watch, a smart ring, smart glasses, etc.

[0119] In some embodiments, any information or data associated with the user factor, the sleep factor, and / or the environment factor can be stored in and retrievable from a database (e.g., a cloud database) in digital communication with the computer processor as provided herein.

[0120] In some embodiments, the computer processor can utilize one or more Al models (e.g., machine learning models as provided herein) to control (e.g., automatically control) operation of the article of furniture or a make a decision to control operation of the article of furniture as provide herein. The one or more algorithms can be utilized, for example, to control operation (e.g., generate operation parameters, such as output level or pattern, operation time, etc.) of a temperature regulator of the article of furniture or an environment regulator (e.g., which regulates one or more environmental conditions, such as ambient temperature) as provided herein.

[0121] In some embodiments, such factor (e.g., the user factor, the sleep factor, and / or the environment factor) can be (i) a current factor as collected, measured, and / or analyzed during a current use of the article of furniture, (ii) a past factor collected, measured, and / or analyzed from one or more past uses of the article of furniture by the user, (iii) a desired factor for the current use of the article of furniture, e.g., as provided by the user via the GUI, and / or (iv) a target factor for the current use of the article of furniture, e.g., based on analysis of the current factor and / or the past factor as provided herein. In some past uses, the one or more past uses can comprise at least or at most about 1 past use, 2 past uses, 3 past uses, 4 past uses, 5 past uses, 6 past uses, 7 past uses, 8 past uses, 9 past uses, 10 past uses, 15 past uses, 20 past uses, 30 past uses, 40 past uses, 50 past uses, 60 past uses, 70 past uses, 80 past uses, 90 past uses, or 100 past uses. In some cases, a plurality of past uses (e.g., 3 past uses) may or may not be consecutive uses. In some past uses, the one or more past uses can comprise at least or at most about 1 past day, 2 past days, 3 past days, 4 past days, 5 past days, 6 past days, 7 past days, 8 past days, 9 past days, 10 past days, 15 past days, 20 past days, 30 past days, 40 past days, 50 past days, 60 past days, 70 past days, 80 past days, 90 past days, 4 past months, 5 past months, 6 past months, 7 past months, 8 past months, 9 past months, 10 past months, 11 past months, or 12 past months. In some cases, the past factor can be raw data or processed data thereof (e.g., averaging, conditioning averaging, weighted averaging, arithmetic mean, geometric mean, harmonic mean, etc.). For example, the past factor can be the user’s average sleep time or sleep latency over the past three days or uses of the bed device.

[0122] In some embodiments, the user factor, the sleep factor, and / or the environment factor can be based on factual information associated with the user, e.g., based on information or data that is (i) provided by the user (e.g., via GUI) and / or (ii) measured by the one or more sensors as provided herein, wherein such information or data is obtained during a current use or prior use of the article of furniture. For example, the user factor can comprise a desired sleep time (or duration) for the current use of the bed device, as provided by the user. In another example, thesleep factor can be based on a current biological signal level (e.g., heart rate) or sleep phase (e.g., REM sleep or deep sleep) of the user that is measured / ascertained while the user is sleeping. In some embodiments, the user factor, the sleep factor, and / or the environment factor can be based on a target value thereof, e.g., generated by the computer processor. For example, the user factor can comprise a sleep time (or duration) of the user from one or more prior uses of the bed device. In another example, the sleep factor can be based on a target biological signal level (e.g., target heart rate that is generated based on past historical data) or target sleep phase (e.g., REM sleep or deep sleep) of the user.

[0123] In some embodiments, the user factor can comprise one or more members comprising biological sex of the user, age of the user, health condition of the user, exercise information of the user, diet information (or food consumption information) of the user, and / or a combination thereof. In some case, a plurality of members of the user factor and / or a plurality of sub-types of each member of the user factor can be utilized (e.g., analyzed) by the algorithm. The algorithm can be trained to give equal weight or different weights of significance to the plurality of members and / or the plurality of sub-types (e.g., at least or at most about 1%, 2%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, or 99% of difference in significance between two different members and / or between the plurality of sub-types).

[0124] In some embodiments, a health condition of the user can be a health status of the user can be a generic classification such as, for example, “normal” (or “healthy”) or “abnormal” (or “unhealthy”). In some embodiments, a health condition of the user can be a physical health condition (e.g., cold, fever, fatigue, menopause, disease type, etc.) or a mental health condition (e.g., stress, anxiety, etc.).

[0125] In some embodiments, the health condition of the user can be objective. For example, the health condition can be provided in the user’s own verbal or written descriptions, which can be analyzed via one or more computer algorithms such as, for example, natural language processing (NLP) algorithm and / or convolution neural network (CNN) algorithm for image processing. In some embodiments, the health condition of the user can be based on a conclusion of a healthcare provider (e.g., a doctor) that is provided to the computer processor. In some embodiments, the health condition of the user can be determined based a sensing data (e.g., biological signals) measured by one or more sensors as provided herein, e.g., while the user is using the article of furniture during the current use or past use of the article of furniture.

[0126] In some embodiments, the exercise information and / or diet information of the user can be objective. For example, the exercise information and / or diet information can be provided in the user’s own verbal or written descriptions, which can be analyzed via one or more computeralgorithms (e.g., NLP, CNN, etc.) as provided herein. In some embodiments, the exercise information and / or diet information can be retrieved from a database associated with a user device as provided herein, e.g., a smart watch comprising one or more sensors to track movement or exercise of the user during or prior to the user’s use of the article of furniture. Non-limiting examples of exercise information can include duration and / or frequency of walking, running, swimming, basketball, baseball, hockey, tennis, gymnastics, standing for duration. Non-limiting examples of diet information can include types of foods consumed by the user (e.g., elementary foods, pre-packaged meals, home-cooked meals, fruits, vegetables, etc.), amounts of foods consumed by the user, frequency of food consumption by the user, and / or time of the food consumption within the day.

[0127] In some embodiments, the sleep factor can comprise one or more members comprising a sleep routine, one or more biological signals (e.g., biometrics) of the user, a sleep phase information of the user, the user’s preferred conditions of the bed device during sleep, and / or a combination thereof. In some case, a plurality of members of the sleep factor and / or a plurality of sub-types of each member of the sleep factor can be utilized (e.g., analyzed) by the algorithm. The algorithm can be trained to give equal weight or different weights of significance to the plurality of members and / or the plurality of sub-types (e.g., at least or at most about 1%, 2%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, or 99% of difference in significance between two different members and / or between the plurality of sub-types).

[0128] In some embodiments, the sleep factor can comprise a sleep routine. The sleep routine can comprise one or more members comprising a bed time, sleep time, sleep latency, wake up time, and a combination thereof. In some cases, the bed time can be a time that the user begins using the bed device, such as a time that the user gets on the bed. In some cases, the sleep time can be a time that the user transitions from being awake to being asleep, e.g., when the user enters a light sleep on the bed device for the first time on the particular day. The time of the entry into the light sleep can be determined, e.g., based on analysis of sensing data from the one or more sensors (e.g., heart rate data from a pressure sensor, such as a piezoelectric sensor). In some cases, the sleep latency can be a duration of time between the bed time and the sleep time. In some cases, the wake up time can be when the user transitions from being asleep (e.g., in one or more sleep phases) to being awake. Alternatively, the wake up time can be a predetermined alarm time (e.g., as determined by the user or by the computer processor).

[0129] In some embodiments, the sleep factor can comprise one or more biological signals of the user, as provided herein, or analysis thereof. In some cases, the one or more biological signals can comprise sensing data collected during or over a current use of the article of furniture(e.g., bed device). For example, the sensing data can be measured or collected substantially in real-time. Alternatively or in addition to, the sensing data can be collected during the current use and over at least or at most about 5 minutes, 10 minutes, 15 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes, 60 minutes, 70 minutes, 80 minutes, 90 minutes, 100 minutes, 110 minutes, 120 minutes, 3 hours, 4 hours, 5 hours, 6 hours, 7 hours, or 8 hours prior to use of the sensing data (e.g., processing or analysis of the sensing data) by the computer processor. In some cases, the sleep factor can comprise analysis of the one or more biological signals. The sleep factor can be based on whether the currents sensing data of the biological signal of the user is within or outside of a predetermined range (e.g., average or normal range) of past sensing data of the biological signal of the user. For example, the biological signal can be a heart rate (e.g., resting heart rate, HRV), and the sleep factor can be whether or not the user’s current heart rate is within a predetermined range of the heart rate. In another example, the biological signal can be breathing rate, and the sleep factor can be whether or not the user’s current breathing rate is within a predetermined range of the breathing rate (e.g., between about 12 and about 20 breaths per minute).

[0130] In some embodiments, the sleep factor can comprise one or more sleep phase information of the user, as provided herein, or analysis thereof. In some cases, the sleep factor can comprise one or more information about the sleep phase, such as a time of onset of the sleep phase, duration of the sleep phase, frequency of the sleep phase (e.g., over the current sleep, or a daily average over a plurality of prior sleeps), etc. In some cases, the computer processor can determine whether or not the time of onset of the sleep phase was earlier or later than a reference time by at least or at most about 5 minutes, 10 minutes, 15 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes, 60 minutes, 70 minutes, 80 minutes, 90 minutes, 100 minutes, 110 minutes, or 120 minutes. Non-limiting examples of the reference time can include predetermined time as determined by the computer processor; one or more sleep routine info such as bed time, sleep time, etc.; onset of a previous sleep phase, etc.

[0131] In some cases, the sleep phase information can comprise the user’s current sleep phase, e.g., as determined substantially in real time. In some cases, the sleep phase information can comprise a past sleep phase. The past sleep phase can be from the user’s current sleep night and / or from the user’s past sleep night(s). For example, the sleep factor can be what type of sleep phase the user was in prior to (e.g., immediately prior to) the current sleep phase.

[0132] In some cases, the sleep phase information can comprise an amount or total duration of a particular sleep phase (e.g., deep sleep, REM sleep, light sleep, or any combination thereof) that the user needs or is suspected of needing during the current sleep / night. The amount or totalduration of the particular sleep phase can be provided by the user, or can be determined by the computer processor, by one or more Al models (e.g., machine learning models) as provided herein, based on the user’s past sleep phase information data from past sleeps or past uses of the bed device. For example, the amount or total duration of the particular sleep phase can be based on the amount or total duration of the same particular sleep phase that the user experienced during the prior night’s sleep (or sleep from a plurality of prior nights). The amount or total duration of the particular sleep phase that the user needs (e.g., as determined by the computer processor) can be at least or at most about 1%, 5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 110%, 120%, 150%, 200%, 250%, 300%, 400%, or 500% more than that from the prior night’s sleep (or an average or median value from a plurality of prior sleeps); or can be at least or at most about 1%, 5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 110%, 120%, 150%, 200%, 250%, 300%, 400%, or 500% less than that from the prior night’s sleep (or an average or median value from a plurality of prior sleeps).

[0133] In some cases, the sleep phase information can comprise a pattern (or trend) of sleep phases of the user from the current night’s sleep (e.g., thus far) and / or from one or more prior nights’ sleep. A subject can experience at least one cycle of a plurality of different sleep phases, such as experiencing, sequentially, light sleep, deep sleep, and REM sleep. A subject can experience a plurality of cycles of different sleep phases, each cycle comprising two or more members of light sleep, deep sleep, and / or REM sleep. In some cases, the pattern of sleep phases can comprise a proportion of a total duration of a particular sleep phase (e.g., deep sleep or REM sleep) relative to the duration of the entire sleep, and / or relative to a total duration of a different sleep phase (e.g., total duration of deep sleep vs total duration of REM sleep). For example, the computer processor can determine that the proportion of the total duration of the particular sleep phase needs to increase (or decrease) by at least or at most about 1%, 5%, 10%, 15%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 100%, 110%, 120%, 150%, 200%, 250%, 300%, 400%, or 500% for the current night as compared to one or more past nights. In some cases, the pattern of sleep phases can comprise the number of occurrences of a particular sleep phase over the plurality of cycles per night. For example, the computer processor can determine that the number of occurrences of the particular sleep phase needs to increase (or decrease) by at least or at most about 1 time, 2 times, 3 times, 4 times, or 5 times for the current night as compared to one or more past nights. In some cases, the pattern of sleep phases can comprise the number of cycles per night. For example, the computer processor can determine that the number of cycles needs to increase (or decrease) by at least or at most about 1 cycle, 2 cycles, 3 cycles, 4 cycles, or 5 cycles for the current night as compared to one or more past nights. In some cases, the sleep phaseinformation can comprise a sleep pattern of the user, as provided herein.

[0134] In some embodiments, the sleep factor can comprise one or more target conditions of the bed device of the user, e.g., at a specific time during sleep. The target condition(s) can be predetermined by the user (e.g., via the GUI). The target condition(s) can be predetermined by the computer processor, e.g., a predetermined schedule of target condition(s) for the night. Nonlimiting examples of a target condition can include a target temperature of the bed device (e.g., as regulated by a temperature controller coupled to or in fluid communication with the bed device, such as via one or more fluid channels within the bed device), a target temperature of the environment of the bed device (e.g., as regulated by a HVAC system), a configuration or angle of the bed device (e.g., incline or decline of an adjustable bed device), etc.

[0135] In some embodiments, the environment factor can comprise one or more members comprising temperature (e.g., may or may not be the same as the temperature of the bed device itself), air quality, humidity, pressure, oxygen level, nitrogen level, noise, light (e.g., brightness, darkness, specific light wavelengths), a combination thereof, and / or a change thereof. In some cases, the change of the environment factor can be determined by comparing at least two environment factor data that are apart (e.g., in terms of data collection time) by at least or at most about 1 minute, 2 minutes, 3 minutes, 4 minutes, 5 minutes, 6 minutes, 7 minutes, 8 minutes, 9 minutes, 10 minutes, 15 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes, 60 minutes, 70 minutes, 80 minutes, 90 minutes, 100 minutes, 120 minutes, 2.5 hours, 3 hours, 4 hours, 5 hours, 6 hours, 7 hours, 8 hours, 12 hours, 16 hours, 24 hours, 36 hours, 48 hours, 3 days, 4 days, 5 days, 6 days, or 7 days. In some cases, the environment factor can be a target environment factor, e.g., as predetermined by the user (e.g., via the GUI) or predetermined by the computer processor. In some cases, the environment factor can be a degree of difference (or similarity) between a current measurement of the environment factor (e.g., as measured substantially in real time) and a target value of the environment factor for that time, e.g., whether the current measurement of the environment factor is greater than (or less than) that of the target value by at least or at most about 1%, 2%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, or 99%. In some cases, the system can comprise or can be operatively coupled to (i) one or more environment sensors configured to detect such environment factor and / or (ii) a database (e.g., a cloud database) comprising information associated with such environment factor.

[0136] Additional details

[0137] In some embodiments, data provided by the user or collected by the processor (e.g., from a database or from one or more sensors as provided herein) can be processed by thecomputer processor, e.g., to perform (e.g., automatically perform via computer processor) any one or the methods as provided herein, such as adjusting a condition (e.g., temperature, sound, etc.) of the article of furniture or the surrounding environment thereof.

[0138] Non-limiting examples of operations of data processing can include filtering, linear filtering, nonlinear filtering, folding, grouping, energy computation, lowpass filtering, bandpass filtering, highpass filtering, median filtering, rank filtering, quartile filtering, percentile filtering, mode filtering, finite impulse response (FIR) filtering, infinite impulse response (HR) filtering, moving average (MA) filtering, autoregressive (AR) filtering, autoregressive moving averaging (ARMA) filtering, selective filtering, adaptive filtering, interpolation, decimation, subsampling, upsampling, resampling, time correction, time base correction, phase correction, magnitude correction, phase cleaning, magnitude cleaning, matched filtering, enhancement, restoration, denoising, smoothing, signal conditioning, enhancement, restoration, spectral analysis, linear transform, nonlinear transform, inverse transform, frequency transform, inverse frequency transform, Fourier transform (FT), discrete time FT (DTFT), discrete FT (DFT), fast FT (FFT), wavelet transform, Laplace transform, Hilbert transform, Hadamard transform, trigonometric transform, sine transform, cosine transform, discrete cosine transform (DCT), power-of-2 transform, sparse transform, graph-based transform, graph signal processing, fast transform, a transform combined with zero padding, cyclic padding, padding, zero padding, feature extraction, decomposition, projection, orthogonal projection, non-orthogonal projection, over-complete projection, eigen-decomposition, singular value decomposition (SVD), principle component analysis (PCA), independent component analysis (ICA), grouping, sorting, thresholding, soft thresholding, hard thresholding, clipping, soft clipping, first derivative, second order derivative, high order derivative, convolution, multiplication, division, addition, subtraction, integration, maximization, minimization, least mean square error, recursive least square, constrained least square, batch least square, least absolute error, least mean square deviation, least absolute deviation, local maximization, local minimization, optimization of a cost function, neural network, recognition, labeling, training, clustering, machine learning, supervised learning, unsupervised learning, semi-supervised learning, self-supervised, comparison with another TSCI, similarity score computation, quantization, vector quantization, matching pursuit, compression, encryption, coding, storing, transmitting, normalization, temporal normalization, frequency domain normalization, classification, clustering, labeling, tagging, learning, detection, estimation, learning network, mapping, remapping, expansion, storing, retrieving, transmitting, receiving, representing, merging, combining, splitting, tracking, monitoring, matched filtering, Kalman filtering, particle filter, interpolation, intrapolation, extrapolation, histogram estimation,importance sampling, Monte Carlo sampling, compressive sensing, representing, merging, combining, splitting, scrambling, error protection, forward error correction, doing nothing, time varying processing, conditioning averaging, weighted averaging, arithmetic mean, geometric mean, harmonic mean, averaging over selected frequency, averaging over antenna links, logical operation, permutation, combination, sorting, AND, OR, XOR, union, intersection, vector addition, vector subtraction, vector multiplication, and vector division.

[0139] In some cases, one or more of the operations of data processing as provided herein can be performed via one or more machine learning models (e.g., one or more trained classifiers). In some cases, one or more of the operations of data processing as provided herein can be performed by a computer processor without utilizing a machine learning models.

[0140] In some cases, more than one model may be used in parallel such as in an ensemble model. In some cases, more than one model may be used with some models being pretrained. In some cases, more than one model may take in more than one type of input data (ex. Image data, audio data, tabular data, text data). In some cases, models that take in more than one type of data may be multi-modal artificial intelligence models (ex. Large language models, diffusion models).

[0141] In some embodiments, the system as provided herein can comprise at least one sensor operatively coupled to the article of furniture. The at least one sensor can be attached to the article of furniture, can be part of (e.g., disposed and hidden in an internal portion of the article of furniture), or disposed near or adjacent to the article of furniture. The at least one sensor can be configured to detect sensing data associated with a user of the article of furniture. The sensing data can be associated with or indicative of a biological signal of the user. The sensing data can comprise a single biological signal data or a plurality of biological signal data. The sensing data can comprise a single type of biological signal (e.g., sound, vibration, temperature, etc.) or a plurality of different types of biological signal (e.g., sound and vibration, sound and temperature, vibration and temperature, etc.). Alternatively or in addition to, any of the sensor as provided herein can be coupled to (or can be a part of) a user device as provided herein.

[0142] In some embodiments, the system can comprise a controller (e.g., a computer processor) configured to adjust the condition / operation of the article of furniture or other devices (e.g., one or more environment regulators, one or more speakers, one or more coverings, one or more optical sources, etc.) based at least in part on data, such as, for example, user factor, the sleep factor, the environment factor, etc. The controller can be configured to generate a decision to adjust (e.g., generate a control signal for adjusting) the condition / operation of the article of furniture or other devices substantially in real-time or shortly after detection or generation of the sensing data by the at least one sensor. In some cases, the duration or time difference (e.g., ashort span of time) between when such decision is made by the controller and when the sensing data is detected or generated by the at least one sensor can be at least or at most about 1 second, at least or at most about 2 seconds, at least or at most about 5 seconds, at least or at most about 10 seconds, at least or at most about 20 seconds, at least or at most about 30 seconds, at least or at most about 1 minute, at least or at most about 2 minutes, at least or at most about 5 minutes, at least or at most about 10 minutes, at least or at most about 15 minutes, at least or at most about 20 minutes, at least or at most about 25 minutes, at least or at most about 30 minutes, at least or at most about 40 minutes, at least or at most about 50 minutes, at least or at most about 60 minutes, at least or at most about 1.5 hours, or at least or at most about 2 hours prior to when the decision is generated. Alternatively, the decision to adjust the condition / operation of the article of furniture or other devices by the controller and the detection of the sensing data by the at least one sensor may not and need not occur in real-time or within a short span of time between one another as described herein.

[0143] In some embodiments, upon determining that the user may experience a target human condition (e.g., a disease such as a heart condition, or a target sleep condition such as snoring, sleep apnea, etc.), the controller can make the decision to adjust the condition / operation of the article of furniture or other devices, e.g., to treat or ameliorate such undesired condition of the user.

[0144] In some embodiments, the human condition can be a desirable condition. In some embodiments, the human condition can be an undesirable condition. In some cases, the undesirable condition can comprise a sleep disorder as provided herein.

[0145] In some embodiments, the controller can utilize at least one classifier (e.g., at least or at most about 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, or more classifiers) for making the decision as provided herein, e.g., analyzing the user factor, the sleep factor, and / or the environment factor, etc. A classifier can be configured to receive an input comprising at least the user factor, the sleep factor, and / or the environment factor, and provide an output comprising, but not limited to, an instruction for controlling operation of one or more environment regulators, one or more speakers, one or more coverings, one or more optical sources, etc.

[0146] In some embodiments, the controller can be configured to (e.g., via use of at least one classifier and based on the user signal as detected) determine (i) a number of users (e.g., one or two users) currently present on or adjacent to the article of furniture (e.g., on top of a bed device),(ii) an approximate or substantially precise location of each user with respect to the article of furniture (e.g., on the left side of the bed device, on the right side of the bed device, etc.), and / or(iii) identity of each user where the identity is based on information about the user previouslystored in a database.

[0147] In some embodiments, an article of furniture can be associated with a user profile of a user, which user profile is stored in a database (e.g., a cloud database). The user profile can be accessible (e.g., readable, editable, etc.) by the controller as provided herein. The user profile can be accessible (e.g., readable, editable, etc.) by the user via a graphical user interface (GUI) of a user application installed on a user device. The user profile can comprise one or more adjustment profiles of the condition of the article of furniture or the environment thereof for the user, as described herein. The user profile can comprise any additional information about the user including, but not limited to, biological sex, age, height, weight, medical history, family history, identity of family members, genetic information, blood information, information about additional article(s) of furniture of the same user or other users (e.g., other members of the cohort as provided herein), etc. The additional information can be provided by the user. The additional information can be retrieved from another database associated with the user (e.g., a database associated with the user’s smart watch, with the user’s genetic analysis such as 23andMe, etc.).

[0148] In some embodiments, any action taken by the controller to operate the article or furniture (e.g., adjustment setting of a condition of an article of furniture) and / or any information associated with operation of the article of furniture (e.g., a sensing data, an adjustment profile of a condition of an article of furniture, etc.) can be stored in the database associated with the user. Data associated with such action or information can be stored (or updated) per each use of the article of furniture. Data associated with such action or information can be stored along with (e.g., labeled with) a unique identifier of the particular article of furniture (e.g., information associated with the machine readable code of the article of furniture) and / or date and time of recordation of the data, e.g., to track the source of the data. Data of different types (e.g., different sensing data, different conditions and adjustments thereof) can be charted when stored (e.g., as a table or map), such that the data of different types can either be retrieved, reviewed, analyzed, and / or updated individually or as a group.

[0149] In some embodiments, the controller can be configured to compute a score indicative of a quality of the user’s usage of the article of furniture, based at least in part on analyzing the user factor, the sleep factor, and / or the environment factor. For example, after usage, the user can provide feedback or input via GUI associated with the controller about how the user perceives the usage was like, and the controller can utilize such information to compute the score. In some embodiments, the article of furniture can be a bed device, and the controller can be configured to compute a sleep score indicative of a sleep quality of the user while sleeping on the bed device. The sleep score can be determined based on the sensing data (e.g., occurrence orduration of one or more particular sleep phases, order of a plurality of sleep phases, heart signal such as HRV or heart rate of the user while sleeping, movement of the user while sleeping, etc.). For example, the sleep score can be computed based on the relative proportion of different sleep phases (e.g., percentage of deep sleep, percentage of REM sleep, percentage of light sleep, etc.) during the night, a sleep routine of the user (e.g., whether the user has gone to bed and woken up consistently over the course of days, weeks, or months), and / or a total time slept. Alternatively or in addition to, the sleep score can be determined based on the user’s feedback after the user wakes up. For example, the GUI can provide a scale (e.g., on a scale from 1 to 10) for the user to mark to indicate the user’s perception of the sleep quality, and the controller can either accept the feedback as truth for the sleep score or utilize the feedback to generate a new sleep score.

[0150] In some embodiments, the sleep score as computed by the controller can be relative to a threshold (or benchmark) sleep score. The threshold sleep score can be based on a compiled sleep score and / or sensing data from a population of users (e.g., at least about 10, 50 ,100, 500, 1,000, 5,000, 10,000, 20,000, 50,000, or more users) collected over the course of days, weeks, months, or years (e.g., at least or at most about 1 year, at least or at most about 2 years, at least or at most about 3 years, at least or at most about 4 years, at least or at most about 5 years, etc.). Alternatively, in some embodiments, the sleep score as computed by the controller can be more individualized, and the threshold sleep score can be based on a compiled sleep score and / or sensing data from the specific user collected, for example, collected longitudinally, over the course of days, weeks, months, or years (e.g., at least or at most about 1 year, at least or at most about 2 years, at least or at most about 3 years, at least or at most about 4 years, at least or at most about 5 years, etc.).

[0151] In some embodiments, the controller can be configured to analyze the score indicative of the quality of the user’s usage of the article of furniture (e.g., sleep score), and determine how to adjust the condition of the article of furniture during a subsequent use of the article of furniture by the user. In some embodiments, if the sleep score is below a threshold (e.g., as determined by the user or automatically determined by the controller), the controller can be configured to modify (e.g., automatically modify) the previously used adjustment profile of a condition of the bed device, in order to enhance the sleep score during the user’s subsequent use of the article of furniture or equivalent thereof. In some embodiments, adjustment profile of the user can be modified (e.g., a previously used adjustment profile of the user can be modified for a subsequent use of the article of furniture by the user) based on (i) the sleep score and / or (ii) one or more components or factors utilized (or analyzed) to generate the sleep score (e.g., quality or duration of one or more target sleep phases such as REM sleep, deep sleep, or light sleep).

[0152] In some embodiments, a score (e.g., sleep score) as provided herein can be a rating. The rating can be a numerical rating, alphabetical rating, alphanumerical rating, percentage, graphical system, etc. In some cases, the ratings can be provided based on a scale, wherein a higher rating can be associated with a better user experience (e.g., a better sleep quality) as compared to a lower rating. In some cases, the rating can be based on a 0 percent (%) to 100 % scale, a higher percentage value representing a better user experience (e.g., a better sleep quality). The rating can be at least about 0 %, 1 %, 2 %, 3 %, 4 %, 5 %, 6 %, 7 %, 8 %, 9 %, 10 %, 20 %, 30 %, 40 %, 50 %, 60 %, 70 %, 80 %, 90 %, 99 %, or more. The rating can be at most about 100 %, 90 %, 80 %, 70 %, 60 %, 50 %, 40 %, 30 %, 20 %, 10 %, 9 %, 8 %, 7%, 6 %, 5 %, 4 %, 3 %, 2 %, 1 %, or less. In some cases, the rating can be based on another numerical scale, such as, for example, a 0 to 10 scale (e.g., 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 in an order of positive effects). In some cases, the rating can be based on a graphical scale, such as, for example, a 1 to 5 stars scale (e.g., 1, 2, 3, 4, and 5 starts in an order of positive effects). In some cases, the rating can be an alphabetical rating system, such as, for example, one or more of D-, D, D+, C-, C, C+, B-, B, B+, A-, A, and A+.

[0153] In some embodiments, the calculated sleep score and / or the raw data of the user of the article of furniture (e.g., that is used for determining the calculated sleep score) can be compared to a benchmark data (e.g., calculated sleep score, raw data thereof, analysis thereof such as an average value, etc.) of multiple users of comparable article of furniture. The multiple users can comprise at least about 10, at least about 50, at least about 100, at least about 500, at least about 1,000, at least about 5,000, at least about 10,000, or more users. The benchmark data can be collected over a duration of time, such as at least about 1 year, at least about 2 years, at least about 3 years, at least about 4 years, at least about 5 years, or more. The benchmark data can comprise data collected within the past 1 year, 2 years, 3 years, 4 years, 5 years, or more. For example, the comparison can show that the user of the article of furniture has a higher or lower sleep score compared to other users of the comparable article of furniture. In some embodiments, the sleep score of the user of the article of furniture can be compared to a benchmark data (e.g., the average sleep score) of the user from prior uses of the article of furniture. For example, the comparison can show that the user of the article of furniture has a higher or lower sleep score compared to prior uses of the article of furniture by the user. Alternatively or in addition to, the benchmark data can be a personalized benchmark data of the user (e.g., not indicative of any other individuals). The personalized benchmark data can comprise a target goal or value of the user, or a value (e.g., average value) from one or more prior nights. Alternatively or in addition to, the benchmark data can be a threshold value (e.g., a global threshold value) applicable to aplurality of users.

[0154] In an example, sleep score and / or raw data indicative of the sleep routine can be compared against a personalized benchmark of the user. In another example, sleep score and / or raw data indicative of the quality of sleep can be compared against a population benchmark. In another example, sleep score and / or raw data indicative of the total sleep time can be compared against a threshold value (e.g., about 8 hours as a target threshold value of total sleep time per night).

[0155] FIG. 1 schematically illustrates an example graphical user interface (GUI) displaying information about the user’s sleep score (“Sleep Fitness Score”) for a selected day or date (Friday or “F”). The sleep score can be provided in one or more ratings such as numerical score (e.g., 88 out of 100) or textual descriptions (e.g., Good). The GUI can provide one or more factors utilized for determining the sleep score, such as the sleep quality (“Quality”), the sleep routine of the user (“Routine), and the total time slept (“Time slept”). Each of the factors can be provided with their own analysis or scoring (e.g., as determined by the classifier), in order for the user to see, for example, possible reasons for their sleep score being higher or lower than expected. In some cases, the GUI can allow the user to select at least one of the one or more factors on the GUI, to trigger the GUI to display more detailed information about the selected factor. For example, as demonstrated in FIG. 1, the sleep quality factor can be selected, to prompt the GUI to display overall progress of the user’s sleep stages during the selected night (e.g., color schematics to denote different sleep stages, x-axis to denote time, and y-axis to denote a biological signal of the user or that of the bed or environment detected during sleep).

[0156] In some embodiments, the systems and methods of the present disclosure can be utilized for evaluating quality of the user’s sleep while sleeping on the article of furniture such as a bed device, e.g., determining a sleep metric such as a sleep score. The article of furniture can detect (e.g., via one or more sensors of the article of furniture one or more biological signals of the user as provided herein while they are sleeping. For example, the one or more biological signals can comprise heart signal (e.g., heart rate, heart variability rate, etc.), respiratory signal (e.g., respiratory rate), or both. The one or more biological signals can comprise a plurality of different biological signal types that can be obtained from a same sensor (e.g., same piezoelectric sensor) or different sensors. The controller (e.g., computer processor) operatively coupled to the article of furniture can determine (e.g., compute) different sleep stages (e.g., light sleep, REM sleep, and / or deep sleep), their order with respect to one another, and duration(s) thereof throughout the night. Identification of the sleep stages and durations thereof can be performed via classifiers such as machine learning algorithms that have been trained on ground truth datawith respect to each of the sleep stages and the associated biological signal(s). Subsequently, the identified sleep stages, their order, and / or durations thereof can be analyzed (e.g., via classifiers such as machine learning algorithms) to determine if the user is or has experienced “low quality” sleep over night. In some cases, one or more factors such as (i) presence or absence of any particular sleep stage (e.g., deep sleep, REM sleep) and / or (ii) duration of at least one particular sleep stage relative to the duration of the entire sleep (e.g., % REM sleep and / or % deep sleep relative to total sleep duration) can be utilized to determine the quality of the user’s sleep over night. In some cases, the controller can generate and / or report out a sleep score indicative of the quality of the user’s sleep.

[0157] In some embodiments, the quality of the sleep or the sleep score, as provided herein, can be based at least in part on a threshold value determined by the user or automatically determined by the classifier as provided herein. The threshold value can be at least or at most about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 99%, 100%, 110%, 120%, 130%, 140%, 150%, 160%, 170%, 180%, 190%, or 200% of a target sleep phase relative to the duration of the entire sleep (e.g., % REM sleep and / or % deep sleep relative to total sleep duration) or relative to another sleep stage (e.g., % REM sleep relative to light sleep, % deep sleep relative to light sleep, % REM sleep relative to % deep sleep).

[0158] In some embodiments, the quality of the sleep or the sleep score, as provided herein, can be based at least in part on a threshold duration of a sleep stage (e.g., deep sleep, REM sleep), e.g., whether such measured or determined duration is longer than or shorter than a threshold duration. The threshold duration can be at least or at most about 5 minutes, 10 minutes, 15 minutes, 20 minutes, 30 minutes, 40 minutes, 50 minutes, 60 minutes, 1.5 hours, 2 hours, 2.5 hours, 3 hours, 3.5 hours, 4 hours, 4.5 hours, 5 hours, 5.5 hours, 6 hours, 6.5 hours, 7 hours, or 8 hours.

[0159] In some embodiments, the automatic adjustment of the condition of the article of furniture (e.g., as determined and implemented by the controller) can be programmed to persist for a designated duration of time (e.g., a designated number of uses of the article of furniture, a designated number of days, weeks, or months). The duration of time can be designated by the user. Alternatively or in addition to, the duration of time can be designated by the controller. In some cases, based on the human condition associated with the user detected by the controller, the controller can designate an appropriate duration of time to continue the automated adjustment of the condition of the article of furniture. For example, when the human condition is an acute condition (e.g., fever), the controller can be configured to designate a short duration of time (e.g.,a few days to a week). In another example, when the human condition is not an acute condition (e.g., persistent snoring or sleep apnea), the controller can be configured to designate a long duration of time (e.g., a plurality of months, years, or even indefinite). In some cases, the designated duration of time can be determined to determine a target condition of the article of furniture to be applied to the article of furniture (e.g., based on an averaging effect of the conditions of the article of furniture to the user’s health condition or score such as sleep score).

[0160] In some embodiments, the condition of the article of furniture can comprise for example, temperature of the article of furniture, vibration of the article of furniture (e.g., to provide a massage function to the user, to enhance sleep quality of the user, to wake up the user from sleeping, etc.), shape of the article of furniture (e.g., tile angle of at least a portion of the article of furniture, such as a bed or a chair), etc. In some cases, the article of furniture can comprise an instrument to regulate such condition of the article of furniture. The instrument can comprise at least a portion of a temperature control unit to adjust temperature of the article of furniture. The temperature control unit can comprise an electric blanket or one or more channels to permit flow of temperature-controlled fluid (e.g., liquid or air). In some cases, the instrument can comprise one or more motors to induce such vibration in at least a portion of the article of furniture. In some cases, the instrument can comprise one or more actuators to adjust the shape of the bed.

[0161] In some embodiments, the condition of the article of furniture can be vibration, e.g., via one or more motors as provided herein. The article of furniture can be configured to turn on or off vibration of at least a portion of the article of furniture (e.g., at a location that correspond to a specific bodily location such as head, arms, torso, legs, etc.). The one or more motors can be configured to vibrate in one or more vibration patterns, which pattern can be selected by the controller based on its analysis of the human condition of the user. The vibration patterns can differ from one another by varying durations of vibration, frequencies of vibration, intensities of vibration, etc.

[0162] The condition of the article of furniture can comprise a condition of an environment surrounding or comprising the article of furniture. The condition of the environment can comprise ambient temperature, ambient pressure, light, noise, humidity, oxygen level, scent, etc., and such condition can be regulated by one or more instruments operatively coupled to (e.g., disposed within or adjacent to) the environment, such as a thermostat, a humidifier, an oxygen regulator, a light, a speaker, a humidifier, an electric diffuser, etc.

[0163] In some embodiments, the systems and methods of the present disclosure can be utilized (e.g., via the article of furniture and / or the controller) to control other devices (orinstruments as used interchangeably herein) in the environment or associated with the user. Nonlimiting examples of the other devices can include, but are not limited to, an alarm, a coffee machine, a lock, a user device (e.g., a mobile phone, a personal computer, a smart watch, etc.), a car, an exercise machine, etc.

[0164] Another aspect of the present disclosure provides a system comprising one or more computer processors and computer memory coupled thereto. The computer memory comprises machine executable code that, upon execution by the one or more computer processors, implements any of the methods above or elsewhere herein such as, for example, one or more operations of the sensor unit as provided herein.

[0165] Artificial intelligence., machine learning, and algorithms

[0166] In some embodiments, any decision or action of the controller (or computer processor as used interchangeably herein) as provided herein can be decided (or generated) by analyzing data based on one or more computer implemented models (e.g., Al models such as ML models) or computer algorithms.

[0167] In some embodiments, the term “classifier” and “machine learning model” can be used interchangeably herein. A machine learning model can be trained based by analyzing (e.g., via one or more computer algorithms) one or more training data sets, such that the machine learning model can be used for analyzing (e.g., classifying) new data set. For example, the machine learning model can be an architecture (or an organization) of a plurality of computer algorithms that is trained accordingly.

[0168] In some embodiments, the computer implemented model is a classifier as provided herein can be utilized to process (e.g., analyze) one or more of the user factor, the sleep factor, and / or the environment factor.

[0169] In some embodiments, the classifier can be trained based on past data associated with the subject or a cohort of subjects. The past data can comprise one or more of (i)-(vi) as provided herein. The past data can comprise data collected at least or at most about 1 day, 2 days, 3 days, 4 days, 5 days, 6 days, 7 days, 8 days, 9 days, 10 days, 11 days, 12 days, 13 days, 14 days, 3 weeks, 4 weeks, 5 weeks, 6 weeks, 7 weeks, 8 weeks, 9 weeks, 10 weeks, 11 weeks, 12 weeks, 4 months, 5 months, 6 months, 7 months, 8 months, 9 months, 10 months, 11 months, 12 months, 15 months, 18 months, 21 months, or 24 months prior to use of the classifier to determine any action or operation as provided herein. In some cases, the classifier can be continuously trained with new data comprising one or more of (i)-(vi). The frequency of continuous training (or update) of the classifier can be at least once per day, week (or multiple weeks), month (ormultiple months), year (or multiple years), etc.

[0170] In some embodiments, the classifier can be trained based on data that is not provided by the user or collected by the sensors as provided herein. In some embodiments, the classifier can be trained on a reference data. In some cases, the reference data can comprise clinical data (e.g., thermal imaging data and analysis thereof from experimental studies or clinical studies) collected from a cohort of individuals. In some cases, the reference data can be an artificial data that is not collected from any specific individual. For example the reference data can comprise predicted or hypothetical data of one or more biological signals (e.g., to be utilized as pseudoground truth data). In some cases, the reference data can be utilized as a ground truth data.

[0171] In some embodiments, the classifier as provided herein can be trained by applying computer algorithms (e.g., deep learning algorithms, clustering algorithms, forest based models, regression models, classifier models, etc.) on the control data as disclosed herein as a training dataset. Non-limiting examples of training paradigms that include computer algorithms for training a classifier can include supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, self-learning, feature learning, anomaly detection, association rules, etc. In some cases, the classifier can be trained by using one or more learning models on such training dataset. Non-limiting examples of learning models or model architectures can include artificial neural networks (e.g., convolutional neural networks (CNN), recurrent neural networks (RNN), long short-term models (LSTM), encoder-decoder, support vector machines (SVM), generative adversarial networks (GAN), diffusion models, transformers, graph neural networks (GNN), large-language models, U-net architecture neural network, etc.), backpropagation, boosting, decision trees, support vector machines, regression analysis, Bayesian networks, Markov models, hidden Markov models, simulated annealing, genetic algorithms, kernel estimators, conditional random field, random forest, ensembles of classifiers, minimum complexity machines (MCM), probably approximately correct learning (PACT), etc.

[0172] In some embodiments, the classifier can continue to be trained by applying computer algorithms on the control data as well as any new input data (e.g., the current sensing data detected and generated substantially in real-time or within the short span of time as provided herein) as a training dataset.

[0173] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can be an unsupervised, semi-supervised or selfsupervised model, fully-supervised, etc. Models such as these may not require the output of the model to be used as the output used for downstream applications or modeling while other embodiments may. Non-limiting methods include transformer models, BERT models, ERNIEmodels, autoencoders, variational autoencoders, generative adversarial networks, ladder network, Siamese networks.

[0174] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can be a generative model. Non-limiting examples may include variational autoencoders, encoders, deconvolution encoders, generative adversarial networks, transformer models.

[0175] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can be a multimodal artificial intelligence model. Examples of such models include any model which takes in features of different types or modalities, such as but not limited to; image, video, text, sensor, spatial, tabular, genetic, clinical, temporal, discrete or continuous. Non-limiting examples of such methods include large language models (LLM), diffusion models, generative pretrained transformer models.

[0176] In some embodiments, pretrained models may be used individually or with other models or algorithms for various purposes. Machine learning models may be untrained and later trained to optimize for some task or tasks. In some cases, they may use labeled data. In some cases, they may use unlabeled data. In some cases, they may use partially labeled data. In some cases, they may use data whose labels are generated by a machine learning model that may be the same model being trained or may be a different model and the labels may be generated during training or outside of the training of the model. On the other hand, pretrained models are models wherein the training has been completed. A model may be trained multiple times. In some cases, a model is continually trained and copies of the parameters at different time points used as trained versions of the model. In some cases, a model may be alternated between training and use (inference, classification, regression, etc.).

[0177] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can be trained in a transfer learning method. In such embodiments, the model may first be trained on a plurality of data that is only related to the desired task that the model through the data modality where the plurality of training data may contain information relevant to the desired task of the model in its some form. The computer implemented model may then be used on a different plurality of data that is specific to the desired task. Examples of such training methods may, in a non-limiting way, include training a model for disease detection through infrared data by first training the model on data collected by a sensor in various settings (e.g., outdoors, indoors, in a car, in different climates); the method would then train the model a second time, while retaining the trained parameters of the first training step, on infrared data specific to disease detection. The benefit of such training methodscan be one or both of: (1) they can allow for a model to learn the larger distribution of the modalities being input to the model and (2) they can allow a model to be trained using a smaller set of data specific to the desired task without suffering from undergeneralization.

[0178] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can utilize user input. The user input is not limited to input methods of the device or devices disclosed herein and may include, but are not limited to, smartphones, computers, web-interface, app-enabled devices, user made devices, third party devices, third party apps. In some cases, such device can comprise a graphical user interface (GUI) configured to allow the user to provide the input. Such user input may be the result of a prompt from the input method. User input may come from a source such as, but not limited to, a user of the devices disclosed herein, or may be other users such as a friend, family member, a physician or a professional of any kind. For example, the input can comprise how well the user slept (e.g., or how the user perceives the sleep quality of one or more past nights) or how the user thinks about the user’s health.

[0179] In some embodiments, the computer implemented method (e.g., for determining the physiological condition of the subject) can use data collected about the environment or other contextual information not specific to the user. Examples of such data sources can include, but are not limited to, daylight cycle, time of year, household temperature, humidity, air quality, general stressor (e.g., news stories), noise level, allergen levels, indoor light levels, traffic levels in the household etc.

[0180] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can have access to data provided by other devices such as, but not limited to smart home devices. Examples of such devices may include, but are not limited to, thermostats, humidity sensors, lights, devices connected to the internet of things, devices connected by z-wave protocol, matter protocol, devices connected by zigbee protocol, devices connected physically, devices connected through internet connection, devices on a mesh network, window controllers, shade controllers, blind controllers, doorbells, security cameras, door locks, door sensors, window sensors, plugs. In some cases, the computer implemented model may interface with smart home apps such as Phillips Hue, Smart assistant, Amazon Alexa, Samsung SmartThings, Google Nest, Smart Home Manager.

[0181] In some embodiments, the computer implemented model (e.g., for determining the physiological condition of the subject) can have access to information available on the internet. Examples of such information include, but aren’t limited to, social media data, internet use data, purchasing habits.

[0182] A model described herein may be any model capable of image analysis and / or image classification. A model can be a computer vision model, where a computer vision algorithm can refer to methods allowing computers to analyze and / or interpret data from images or videos. Image analysis can comprise object recognition, image segmentation, motion detection, or any combination thereof. A computer vision model may comprise (i) image acquisition, in which an image can be captured; (ii) preprocessing, in which an image can be cleaned up and / or sharpened by adjusting brightness, contrast, and / or removal of noise; (iii) feature extraction, in which an algorithm can identify features within the image, for example edges, comers, colors, patterns, or any combination thereof; (iv) classification or analysis, in which the algorithm can form decision about what objects are present based on one or more extracted features; or (v) any combination thereof. A model described herein can comprise a vision transformer and / or a convolution imaging model.

[0183] In some embodiments, the model may be a computer vision model. An algorithm can be used to detect a stimulus. The stimulus may be a movement, vibration, auditory stimulus, visual stimulus, or any combination thereof. In some embodiments, the stimulus may be indicative of snoring. Snoring may be detected by a movement of a user, by a vibration of a user, by a sound from a user, or any combination thereof. In some embodiments, snoring may be detected by one or more sensors described herein. For example, snoring of a user (e.g., a user of a bed device) may be detected by one or more piezo sensors described herein. Snoring can be detected in a time domain (e.g., over a duration of time) and / or by a frequency domain (e.g., a vibration and / or auditory frequency). Snoring may be detected over any duration of time while a user is asleep. Snoring may be detected over any sleep phase of the user (e.g., light sleep, deep sleep, REM sleep, or any combination thereof).

[0184] Snoring may be detected using an algorithm described herein. A machine learning algorithm may be trained on a data set. The data set can comprise a number of snoring detections from one or more sensors (e.g., one or more piezo sensors). A spectrogram can be generated from the one or more piezo sensors. In some embodiments, the algorithm may train on real-time piezo spectrograms to identify snoring. The spectrogram may be analyzed by an algorithm (e.g., a machine learning algorithm). In some embodiments, the spectrogram may be converted prior to being analyzed by the algorithm. In some embodiments, the spectrogram may not be converted prior to being analyzed by the algorithm.

[0185] Spectrograms may comprise a frequency axis ranging from about 0 Hz to about 500 Hz. Spectrograms may comprise a frequency axis ranging from about 0 Hz to about 10 Hz, about 0 Hz to about 25 Hz, about 0 Hz to about 50 Hz, about 0 Hz to about 75 Hz, about 0 Hz to about100 Hz, about 0 Hz to about 150 Hz, about 0 Hz to about 200 Hz, about 0 Hz to about 250 Hz, about 0 Hz to about 300 Hz, about 0 Hz to about 400 Hz, about 0 Hz to about 500 Hz, about 10 Hz to about 25 Hz, about 10 Hz to about 50 Hz, about 10 Hz to about 75 Hz, about 10 Hz to about 100 Hz, about 10 Hz to about 150 Hz, about 10 Hz to about 200 Hz, about 10 Hz to about 250 Hz, about 10 Hz to about 300 Hz, about 10 Hz to about 400 Hz, about 10 Hz to about 500 Hz, about 25 Hz to about 50 Hz, about 25 Hz to about 75 Hz, about 25 Hz to about 100 Hz, about 25 Hz to about 150 Hz, about 25 Hz to about 200 Hz, about 25 Hz to about 250 Hz, about 25 Hz to about 300 Hz, about 25 Hz to about 400 Hz, about 25 Hz to about 500 Hz, about 50 Hz to about 75 Hz, about 50 Hz to about 100 Hz, about 50 Hz to about 150 Hz, about 50 Hz to about 200 Hz, about 50 Hz to about 250 Hz, about 50 Hz to about 300 Hz, about 50 Hz to about 400 Hz, about 50 Hz to about 500 Hz, about 75 Hz to about 100 Hz, about 75 Hz to about 150 Hz, about 75 Hz to about 200 Hz, about 75 Hz to about 250 Hz, about 75 Hz to about 300 Hz, about 75 Hz to about 400 Hz, about 75 Hz to about 500 Hz, about 100 Hz to about 150 Hz, about 100 Hz to about 200 Hz, about 100 Hz to about 250 Hz, about 100 Hz to about 300 Hz, about 100 Hz to about 400 Hz, about 100 Hz to about 500 Hz, about 150 Hz to about 200 Hz, about 150 Hz to about 250 Hz, about 150 Hz to about 300 Hz, about 150 Hz to about 400 Hz, about 150 Hz to about 500 Hz, about 200 Hz to about 250 Hz, about 200 Hz to about 300 Hz, about 200 Hz to about 400 Hz, about 200 Hz to about 500 Hz, about 250 Hz to about 300 Hz, about 250 Hz to about 400 Hz, about 250 Hz to about 500 Hz, about 300 Hz to about 400 Hz, about 300 Hz to about 500 Hz, or about 400 Hz to about 500 Hz.

[0186] The spectrogram from the piezo sensor may be converted into an image (e.g., image data). The image may be analyzed by a model (e.g., a computer vision model). The model may identify which side of a bed, mattress, or mattress cover a snore originates from. The model may place a box around a snore in the image converted from the spectrogram. An image can comprise a two-dimensional image (e.g., a two-dimensional object). An image can comprise a three- dimensional image (e.g., a three-dimensional object). In some embodiments, the image can have one or more color channels. The image can have three color channels. For example, the image may have a red channel, a green channel, and a blue channel. Preprocessing from the model can convert data from the one or more piezo sensors (e.g., two piezo sensor channels) into one or more images (e.g., one or more two-dimensional spectrogram representations).

[0187] The training data can comprise a data associated with one or more individuals. The training data associated with at least about 1 individual, at least about 2 individuals, at least about 3 individuals, at least about 4 individuals, at least about 5 individuals, at least about 10 individuals, at least about 15 individuals, at least about 20 individuals, at least about 30indivi duals, at least about 40 individuals, at least about 50 individuals, at least about 100 individuals, at least about 200 individuals, at least about 300 individuals, at least about 400 individuals, at least about 500 individuals, at least about 1000 individuals, at least about 2000 individuals, at least about 3000 individuals, at least about 4000 individuals, at least about 5000 individuals, at least about 10000 individuals, or greater than about 10000 individuals. Trained data may be labeled in multiple ways. For example, trained data may be labeled as a snore. Trained data may be labeled as a noise (e.g., a large spike in the data that may indicate movement of a user). Trained data may be labeled as other, where other may signify an alternative stimulus or condition from a snore or noise. For example, a label of other may signify speech of a user, wheezing and / or sneezing of a user, vibration alarms, or any combination thereof. Training may be on a number of iterations of the image data generated from the piezo sensors (e.g., iterations of spectrograms generated from piezo sensors for snoring detection). For example, the model may be trained on at least about 5, at least about 10, at least about 15, at least about 20, at least about 25, at least about 50, at least about 100, at least about 500, at least about 1,000, at least about 5,000, at least about 10,000, at least about 50,000, at least about 100,000, at least about 500,000, at least about 1 million, at least about 2 million, or greater than about 2 million iterations of spectrograms from piezo sensor data. In some embodiments, the model may be trained on at most about 2 million, at most about 1 million, at most about 500000, at most about 100000, at most about 50000, at most about 10000, at most about 5000, at most about 1000, at most about 500, at most about 100, at most about 50, at most about 25, at most about 20, at most about 15, at most about 10, at most about 5, or less than about 5 iterations of spectrograms from piezo sensor data.

[0188] The model (e.g., the computer vision model) can be trained by transfer learning from checkpoint pretrained for a computer vision task. For example, the model may be trained on the Common Objects in Context (COCO) dataset. The model can be trained to generate and / or apply bounding boxes with classes of labels. The label classes may indicate a side of a snore, noise, stimulus, or any combination thereof. A label from the model may indicate a snore on a left side of a bed device described herein. A label from the model may indicate a snore on a right side of a bed device described herein. A label from the model may indicate a noise on a left side of a bed device described herein. A label from the model may indicate a noise on a right side of a bed device described herein. A label from the model may indicate a vibration and / or auditory stimulus that may not be a snore on a left side of a bed device described herein. A label from the model may indicate a a vibration and / or auditory stimulus that may not be a snore on a right side of a bed device described herein.

[0189] A computer algorithm can be configured to convert the user sensing data to a same data type as the training data. For example, if the user sensing data is electrical signal data and the training data is image data, the algorithm can convert the electrical signal data to image data. As another example, if the user sensing data is pressure data and the training data is image data, the algorithm can convert the pressure data to image data. The converted user sensing data can be referred to as processed user sensing data. The processed user sensing data can be analyzed to determine one or more conditions (e.g., one or more target conditions) of the user. The target condition can be a sleep disorder. The target condition can be a breathing disorder. The target condition can be snoring. The user may have at least about 1, 2, 3, 4, 5, or more target conditions. The target condition can comprise insomnia, hypersomnia, sleep apnea, circadian rhythm sleep disorder, restless legs syndrome, sleep deficiency, or any combination thereof.

[0190] The processed user sensing data can comprise one or more image data. In some embodiments, the processed user sensing data can comprise a single image data. The single image data may comprise a plurality of channels. A channel of the plurality of channels can be indicative of a user sensing data from a user sensor. In some embodiments, one or more channels may be indicative of one or more user sensing data from the same user. In some embodiments, one or more channels may be indicative of one or more user sensing data from two or more users (e.g., two or more different users). In some embodiments, a first channel can be indicative of a first user sensing data from a first user sensor of the article of furniture and a second channel can be indicative of a second user sensing data from a second user sensor of the article of furniture.

[0191] Snoring detection using a method described herein may detect snoring of one or more users with at least about 50% accuracy, at least about 60% accuracy, at least about 70% accuracy, at least about 75% accuracy, at least about 80% accuracy, at least about 85% accuracy, at least about 90% accuracy, at least about 91% accuracy, at least about 92% accuracy, at least about 93% accuracy, at least about 94% accuracy, at least about 95% accuracy, at least about 96% accuracy, at least about 97% accuracy, at least about 98% accuracy, at least about 99% accuracy, or greater than about 99% accuracy. In some embodiments, snoring detection using a method described herein may detect snoring of one or more users with at most about 99% accuracy, at most about 98% accuracy, at most about 97% accuracy, at most about 96% accuracy, at most about 95% accuracy, at most about 94% accuracy, at most about 93% accuracy, at most about 92% accuracy, at most about 91% accuracy, at most about 90% accuracy, at most about 85% accuracy, at most about 80% accuracy, at most about 75% accuracy, at most about 70% accuracy, at most about 60% accuracy, at most about 50% accuracy, or less than about 50% accuracy.

[0192] In some embodiments, snoring detection using a method described herein may detect snoring of one or more users from about 50% accuracy to about 99% accuracy. In some embodiments, snoring detection using a method described herein may detect snoring of one or more users from about 50% accuracy to about 60% accuracy, about 50% accuracy to about 70% accuracy, about 50% accuracy to about 75% accuracy, about 50% accuracy to about 80% accuracy, about 50% accuracy to about 85% accuracy, about 50% accuracy to about 90% accuracy, about 50% accuracy to about 92% accuracy, about 50% accuracy to about 94% accuracy, about 50% accuracy to about 96% accuracy, about 50% accuracy to about 98% accuracy, about 50% accuracy to about 99% accuracy, about 60% accuracy to about 70% accuracy, about 60% accuracy to about 75% accuracy, about 60% accuracy to about 80% accuracy, about 60% accuracy to about 85% accuracy, about 60% accuracy to about 90% accuracy, about 60% accuracy to about 92% accuracy, about 60% accuracy to about 94% accuracy, about 60% accuracy to about 96% accuracy, about 60% accuracy to about 98% accuracy, about 60% accuracy to about 99% accuracy, about 70% accuracy to about 75% accuracy, about 70% accuracy to about 80% accuracy, about 70% accuracy to about 85% accuracy, about 70% accuracy to about 90% accuracy, about 70% accuracy to about 92% accuracy, about 70% accuracy to about 94% accuracy, about 70% accuracy to about 96% accuracy, about 70% accuracy to about 98% accuracy, about 70% accuracy to about 99% accuracy, about 75% accuracy to about 80% accuracy, about 75% accuracy to about 85% accuracy, about 75% accuracy to about 90% accuracy, about 75% accuracy to about 92% accuracy, about 75% accuracy to about 94% accuracy, about 75% accuracy to about 96% accuracy, about 75% accuracy to about 98% accuracy, about 75% accuracy to about 99% accuracy, about 80% accuracy to about 85% accuracy, about 80% accuracy to about 90% accuracy, about 80% accuracy to about 92% accuracy, about 80% accuracy to about 94% accuracy, about 80% accuracy to about 96% accuracy, about 80% accuracy to about 98% accuracy, about 80% accuracy to about 99% accuracy, about 85% accuracy to about 90% accuracy, about 85% accuracy to about 92% accuracy, about 85% accuracy to about 94% accuracy, about 85% accuracy to about 96% accuracy, about 85% accuracy to about 98% accuracy, about 85% accuracy to about 99% accuracy, about 90% accuracy to about 92% accuracy, about 90% accuracy to about 94% accuracy, about 90% accuracy to about 96% accuracy, about 90% accuracy to about 98% accuracy, about 90% accuracy to about 99% accuracy, about 92% accuracy to about 94% accuracy, about 92% accuracy to about 96% accuracy, about 92% accuracy to about 98% accuracy, about 92% accuracy to about 99% accuracy, about 94% accuracy to about 96% accuracy, about 94% accuracy to about 98%accuracy, about 94% accuracy to about 99% accuracy, about 96% accuracy to about 98% accuracy, about 96% accuracy to about 99% accuracy, or about 98% accuracy to about 99% accuracy.

[0193] A model and / or algorithm described herein can analyze one or more data types. In some embodiments, the data type can comprise pressure data, temperature data, electrical signal data, acceleration data, strain data, force data, image data, or any combination thereof. In some embodiments, a model and / or algorithm described herein can analyze a first data type and a second type. A first data type may be the same data type as the second type. A first data type may be a different data type from the second type.

[0194] Analysis of the image data (e.g., extracting one or more morphological features from the image data) can be performed (e.g., automatically) in at most about 1 hour, at most about 50 minutes, at most about 40 minutes, at most about 30 minutes, at most about 25 minutes, at most about 20 minutes, at most about 15 minutes, at most about 10 minutes, at most about 9 minutes, at most about 8 minutes, at most about 7 minutes, at most about 6 minutes, at most about 5 minutes, at most about 4 minutes, at most about 3 minutes, at most about 2 minutes, at most about 1 minute, at most about 50 seconds, at most about 40 seconds, at most about 30 seconds, at most about 20 seconds, at most about 10 seconds, at most about 5 seconds, at most about 1 second, or less than about 1 second. In some cases, the analysis can be performed in real-time.

[0195] Using the models and / or algorithms described herein, one or more users of the bed device may have a reduction in snoring over one or more sleep sessions. In some embodiments, the systems and / or methods described herein may reduce snoring by at least about 5%, at least about 10%, at least about 15%, at least about 20%, at least about 25%, at least about 30%, at least about 35%, at least about 40%, at least about 45%, at least about 50%, at least about 55%, at least about 60%, at least about 70%, at least about 80%, at least about 90%, or greater than about 90% compared to a level of snoring of a user that does not utilize the bed device and / or methods described herein. In some embodiments, the systems and / or methods described herein may reduce snoring by at most about 90%, at most about 80%, at most about 70%, at most about 60%, at most about 55%, at most about 50%, at most about 45%, at most about 40%, at most about 35%, at most about 30%, at most about 25%, at most about 20%, at most about 15%, at most about 10%, at most about 5%, or less than about 5% compared to a level of snoring of a user that does not utilize the bed device and / or methods described herein. In some embodiments, the systems and / or methods described herein may reduce snoring by 100% (e.g., the systems and / or methods described herein may stop the snoring of the user).

[0196] In some embodiments, the systems and / or methods described herein may reducesnoring from about 5 % to about 90 %. In some embodiments, the systems and / or methods described herein may reduce snoring from about 5 % to about 10 %, about 5 % to about 15 %, about 5 % to about 20 %, about 5 % to about 25 %, about 5 % to about 30 %, about 5 % to about 40 %, about 5 % to about 50 %, about 5 % to about 60 %, about 5 % to about 70 %, about 5 % to about 80 %, about 5 % to about S •0 %, about 10 % to about 15 ° / •o, about 10 % to about 20 %, about 10 % to about 25 %, about 10 % to about 30 %, about 10 % to about 40 %, about 10 % to about 50 %, about 10 % to about 60 %, about 10 % to about 70 %, about 10 % to about 80 %, about 10 % to about 90 %, about 15 % to about 20 %, about 15 % to about 25 %, about 15 % to about 30 %, about 15 % to about 40 %, about 15 % to about 50 %, about 15 % to about 60 %, about 15 % to about 70 %, about 15 % to about 80 %, about 15 % to about 90 %, about 20 % to about 25 %, about 20 % to about 30 %, about 20 % to about 40 %, about 20 % to about 50 %, about 20 % to about 60 %, about 20 % to about 70 %, about 20 % to about 80 %, about 20 % to about 90 %, about 25 % to about 30 %, about 25 % to about 40 %, about 25 % to about 50 %, about 25 % to about 60 %, about 25 % to about 70 %, about 25 % to about 80 %, about 25 % to about 90 %, about 30 % to about 40 %, about 30 % to about 50 %, about 30 % to about 60 %, about 30 % to about 70 %, about 30 % to about 80 %, about 30 % to about 90 %, about 40 % to about 50 %, about 40 % to about 60 %, about 40 % to about 70 %, about 40 % to about 80 %, about 40 % to about 90 %, about 50 % to about 60 %, about 50 % to about 70 %, about 50 % to about 80 %, about 50 % to about 90 %, about 60 % to about 70 %, about 60 % to about 80 %, about 60 % to about 90 %, about 70 % to about 80 %, about 70 % to about 90 %, or about 80 % to about 90 %.

[0197] There can be advantages to conversion of the piezo-derived spectrogram into image data. An advantage of transfer learning from a computer vision model can be the abundance of image datasets and / or pre-trained model checkpoints available. By starting with a model (e.g., a computer vision model) that has been pre-trained on large image databases, snoring can be detected with greater accuracy. For example, one can achieve satisfactory performance on the snore detection with significantly fewer labeled snore examples compared to a model trained from scratch using randomly-initialized model weights. There can be advantages for using vibration and / or pressure sensors rather than auditory sensors for snoring detection. For example, by detecting vibrations from the bed device, the methods and / or algorithms described herein can distinguish between one or more users of the bed device. As another example, by detecting vibrations from the bed device, the methods and / or algorithms described herein can distinguish which side of the bed device a snoring user is on. Once determining a side of the bed device, a section of the bed device can be moved to reduce and / or stop the snoring.Computer systems

[0198] The present disclosure provides computer systems that are programmed to implement methods of the disclosure. FIG. 2 shows a computer system 1101 that is programmed or otherwise configured to direct operation of the system of the present disclosure (e.g., the article of furniture, the sensor, the controller, etc.). The computer system 1101 can be an electronic device of a user or a computer system that is remotely located with respect to the electronic device. The electronic device can be a mobile electronic device.

[0199] The computer system 1101 includes a central processing unit (CPU, also “processor” and “computer processor” herein) 1105, which can be a single core or multi core processor, or a plurality of processors for parallel processing. The computer system 1101 also includes memory or memory location 1110 (e.g., random-access memory, read-only memory, flash memory), electronic storage unit 1115 (e.g., hard disk), communication interface 1120 (e.g., network adapter) for communicating with one or more other systems, and peripheral devices 1125, such as cache, other memory, data storage and / or electronic display adapters. The memory 1110, storage unit 1115, interface 1120 and peripheral devices 1125 are in communication with the CPU 1105 through a communication bus (solid lines), such as a motherboard. The storage unit 1115 can be a data storage unit (or data repository) for storing data. The computer system 1101 can be operatively coupled to a computer network (“network”) 1130 with the aid of the communication interface 1120. The network 1130 can be the Internet, an internet and / or extranet, or an intranet and / or extranet that is in communication with the Internet. The network 1130 in some cases is a telecommunication and / or data network. The network 1130 can include one or more computer servers, which can enable distributed computing, such as cloud computing. The network 1130, in some cases with the aid of the computer system 1101, can implement a peer-to-peer network, which may enable devices coupled to the computer system 1101 to behave as a client or a server.

[0200] The CPU 1105 can execute a sequence of machine-readable instructions, which can be embodied in a program or software. The instructions may be stored in a memory location, such as the memory 1110. The instructions can be directed to the CPU 1105, which can subsequently program or otherwise configure the CPU 1105 to implement methods of the present disclosure. Examples of operations performed by the CPU 1105 can include fetch, decode, execute, and writeback.

[0201] The CPU 1105 can be part of a circuit, such as an integrated circuit. One or more other components of the system 1101 can be included in the circuit. In some cases, the circuit isan application specific integrated circuit (ASIC).

[0202] The storage unit 1115 can store files, such as drivers, libraries and saved programs. The storage unit 1115 can store user data, e.g., user preferences and user programs. The computer system 1101 in some cases can include one or more additional data storage units that are external to the computer system 1101, such as located on a remote server that is in communication with the computer system 1101 through an intranet or the Internet.

[0203] The computer system 1101 can communicate with one or more remote computer systems through the network 1130. For instance, the computer system 1101 can communicate with a remote computer system of a user. Examples of remote computer systems include personal computers (e.g., portable PC), slate or tablet PC’s (e.g., Apple® iPad, Samsung® Galaxy Tab), telephones, Smart phones (e.g., Apple® iPhone, Android-enabled device, Blackberry®), or personal digital assistants. The user can access the computer system 1101 via the network 1130.

[0204] Methods as described herein can be implemented by way of machine (e.g., computer processor) executable code stored on an electronic storage location of the computer system 1101, such as, for example, on the memory 1110 or electronic storage unit 1115. The machine executable or machine readable code can be provided in the form of software. During use, the code can be executed by the processor 1105. In some cases, the code can be retrieved from the storage unit 1115 and stored on the memory 1110 for ready access by the processor 1105. In some situations, the electronic storage unit 1115 can be precluded, and machine-executable instructions are stored on memory 1110.

[0205] The code can be pre-compiled and configured for use with a machine having a processer adapted to execute the code, or can be compiled during runtime. The code can be supplied in a programming language that can be selected to enable the code to execute in a precompiled or as-compiled fashion.

[0206] Aspects of the systems and methods provided herein, such as the computer system 1101, can be embodied in programming. Various aspects of the technology may be thought of as “products” or “articles of manufacture” typically in the form of machine (or processor) executable code and / or associated data that is carried on or embodied in a type of machine readable medium. Machine-executable code can be stored on an electronic storage unit, such as memory (e.g., read-only memory, random-access memory, flash memory) or a hard disk. “Storage” type media can include any or all of the tangible memory of the computers, processors or the like, or associated modules thereof, such as various semiconductor memories, tape drives, disk drives and the like, which may provide non-transitory storage at any time for the softwareprogramming. All or portions of the software may at times be communicated through the Internet or various other telecommunication networks. Such communications, for example, may enable loading of the software from one computer or processor into another, for example, from a management server or host computer into the computer platform of an application server. Thus, another type of media that may bear the software elements includes optical, electrical and electromagnetic waves, such as used across physical interfaces between local devices, through wired and optical landline networks and over various air-links. The physical elements that carry such waves, such as wired or wireless links, optical links or the like, also may be considered as media bearing the software. As used herein, unless restricted to non-transitory, tangible “storage” media, terms such as computer or machine “readable medium” refer to any medium that participates in providing instructions to a processor for execution.

[0207] Hence, a machine readable medium, such as computer-executable code, may take many forms, including but not limited to, a tangible storage medium, a carrier wave medium or physical transmission medium. Non-volatile storage media include, for example, optical or magnetic disks, such as any of the storage devices in any computer(s) or the like, such as may be used to implement the databases, etc. shown in the drawings. Volatile storage media include dynamic memory, such as main memory of such a computer platform. Tangible transmission media include coaxial cables; copper wire and fiber optics, including the wires that comprise a bus within a computer system. Carrier-wave transmission media may take the form of electric or electromagnetic signals, or acoustic or light waves such as those generated during radio frequency (RF) and infrared (IR) data communications. Common forms of computer-readable media therefore include for example: a floppy disk, a flexible disk, hard disk, magnetic tape, any other magnetic medium, a CD-ROM, DVD or DVD-ROM, any other optical medium, punch cards paper tape, any other physical storage medium with patterns of holes, a RAM, a ROM, a PROM and EPROM, a FLASH-EPROM, any other memory chip or cartridge, a carrier wave transporting data or instructions, cables or links transporting such a carrier wave, or any other medium from which a computer may read programming code and / or data. Many of these forms of computer readable media may be involved in carrying one or more sequences of one or more instructions to a processor for execution.

[0208] The computer system 1101 can include or be in communication with an electronic display 1135 that comprises a user interface (UI) 1140 for providing. Examples of UI’s include, without limitation, a graphical user interface (GUI) and web-based user interface.

[0209] Methods and systems of the present disclosure can be implemented by way of one or more algorithms. An algorithm can be implemented by way of software upon execution by thecentral processing unit 1105. The algorithm can, for example, assist in comparison of the sensing data and control data as provided herein.EXAMPLES

[0210] Example 1A: Snore detection algorithm

[0211] Systems and methods of the present disclosure can be utilized to detect a user’s snoring during sleep, and induce a change in a condition (e.g., movement such as vibration, a change in shape such as bed tilt angle) of the article of furniture (e.g., bed device such as a mattress, a pillow, or a cover thereof) in real-time to trigger the user’s position change and thereby reduce or stop the user’s snoring.

[0212] In some embodiments, data associated with snoring (e.g., audio data) collected from a study (e.g., a clinical study) can be utilized as ground truth.

[0213] In some embodiments, data associated with snoring, including non-audio data such as pressure-related data (e.g., piezoelectric sensor data) can be collected and utilized as a training data set to train a classifier. The classifier can be trained to receive sensor data from the article of furniture, such as the non-audio data, and convert at least a portion of the sensor data into audio data or audio-like data. The classifier can then compare the converted data to the ground truth data (which is audio-based data) to determine in real-time whether a user of the article of furniture is snoring.

[0214] In some embodiments, upon determination of the snoring, the controller of the article of furniture can induce the change in a condition the article of furniture substantially in real-time to trigger the user’s position change and thereby reduce or stop the user’s snoring.

[0215] In some embodiments, upon induction of the change in the condition of the article of furniture, the controller can continue to monitor the snoring (or change thereof) of the user, to further determine whether to revert back (or undo) at least a portion of the induced change in the condition of the article of furniture (e.g., stop vibration, until the bed, etc.).

[0216] Example IB: Target human condition detection algorithm

[0217] In some embodiments, the classifier as provided herein can be trained based on a control data, such that the classifier can be configured to analyze a sensing data to identify a target human condition (e.g., a sleep disordered breathing such as snoring). Both of the control data and the sensing data can be same type of data (e.g., pressure or vibration data, such as piezoelectric sensor data). Such same type of data can be collected by comparable or same type of sensors. For example, the control data and the sensing data can be measured and collected from users by the sensors comparable articles of furniture as provided herein (e.g., collectedduring one or more sleep sessions of the users). Analysis of such sensing data by the classifier can be utilized to control one or more conditions of the article of furniture (e.g., a temperature of a bed device, a configuration such as incline or decline of a bed device, etc.) and / or an environment the article of furniture.

[0218] In some embodiments, the control data can be collected from a plurality of users. The plurality of users can be a collection or snorers and non-snorers. Alternatively or in addition to, the plurality of users can comprise (or consist of) snorers (e.g., data from non-snorers identified and removed). In some embodiments, the plurality of users can be at least about 100, 200, 300, 400, 500, 600, 700, 800, 900, 1,000, 1,500, 2,000, 2,500, 3,000, 4,000, 5,000, 6,000, 7,000 8,000, 9,000, 10,000, or more users. In some embodiments, the plurality of users can be at most about 10,000, 9,000, 8,000„000, 6,000, 5,000, 4,000, 3,000, 2,500, 2,000, 1,500, 1,000, 900, 800, 700, 600, 500, 400, 300, 200, 100, or less users.

[0219] In some embodiments, the control data (e.g., from the plurality of users, such as a plurality of snorers) can be collected over, in total or combination, at least about 1,000, 2,000, 5,000, 10,000, 20,000, 50,000, 100,000, 200,000, 500,000, 1 million, 2 million, 3 million, 4 million, 5 million, 6 million, 7 million, 8 million, 9 million, 10 million, 15 million, 20 million, 25 million, 30 million, 40 million, 50 million, 100 million, 200 million, 300 million, 400 million, 500 million, or more minutes. In some embodiments, the control data can be collected over, in total or combination, at most about 500 million, 400 million, 300 million, 200 million, 100 million, 50 million, 40 million, 30 million, 25 million, 20 million, 15 million, 10 million, 9 million, 8 million, 7 million, 6 million, 5 million, 4 million, 3 million, 2 million, 1 million, 500,000, 200,000, 100,000, 50,000, 20,000, 10,000, 5,000, 2,000, 1,000 or less minutes. In some cases, the control data can comprise (or consist of) positive data indicative of the target human condition (e.g., positive snoring data).

[0220] In some embodiments, the classifier as provided herein can be utilized to detect the target human condition of the user and adjust a condition of the article of furniture (or its surrounding) substantially in real-time to reduce a degree of the target human condition (e.g., snoring) or a chance of the target human condition re-occurring during the same sleep session. The degree of the target human condition or the change of it re-occurring during the same sleep session can be reduced by at least about 1%, 2%, 5%, 10%, 15%, 20%, 25%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, 95%, 99%, or more. The degree of the target human condition or the change of it re-occurring during the same sleep session can be reduced by most about 99%, 95%, 90%, 80%, 70%, 60%, 50%, 40%, 30%, 25%, 20%, 15%, 10%, 5%, 2%, 1%, or less.

[0221] Example 2: Sleep apnea or hypopnea algorithm

[0222] Systems and methods of the present disclosure can be utilized to detect a user’s sleep apnea or hypopnea during sleep, and induce a change in a condition (e.g., movement such as vibration, a change in shape such as bed tilt angle) of the article of furniture (e.g., bed device such as a mattress, a pillow, or a cover thereof) in real-time to trigger the user’s position change and thereby reduce or stop the user’s sleep apnea or hypopnea.

[0223] Example 3: Adaptive alarm

[0224] Systems and methods of the present disclosure can be utilized to implement an adaptive alarm that is not fixed at a predetermined alarm time of the user.

[0225] In some embodiments, the user can be sleeping on an article of furniture, such as a bed device. The controller of the article of furniture can be configured to adjust temperature and / or vibration of the bed device to wake up the user at a time (e.g., which is automatically determined by the controller) that is (i) before the user’s predetermined wake-up time, (ii) when the user is in light sleep, (iii) after the user has slept for a minimum threshold duration of time (e.g., at least 3, 4, 5, or 6 hours), and (iv) sufficiently before an upcoming user event (e.g., a calendared event) of the user on that day. For example, the user’s predetermined wake-up time (e.g., as provided by the user via GUI operatively coupled to the controller) can be 7 a.m., and the controller can determine a target time to begin adjusting the temperature and / or vibration of the bed device to wake up the user based a plurality of factors (or all factors) from (i)-(iv) as aforementioned. In some cases, the minimum threshold duration of time can comprise the user’s desired (or target) sleep duration.

[0226] In some embodiments, operation of the adaptive alarm (e.g., via temperature adjustment, vibration, or both) for the article of furniture (e.g., bed device) can be based on a user factor and / or a sleep factor. The user factor can comprise a desired (or target) sleep duration for the use or for the night, e.g., which can be provided by the user to the article of furniture or a controller coupled thereof prior to or during the use of the article of furniture (e.g., via GUI). The sleep factor can comprise a time that the user transitions from being awake to being asleep (e.g., when the user falls sleep on the bed device), a time that the user begins the current use of the article of furniture (e.g., when the user gets on the bed device), or a combination thereof. Accordingly, the controller can (i) set “time zero” for the user’s sleep based on the sleep factor (e.g., when the user falls asleep), (ii) track and / or determine when the user reaches or is expected to reach the desired sleep duration (e.g., from the set “time zero”) for the current use of the article of furniture, and (iii) turn on the adaptive alarm at, upon, or shortly thereafter (e.g., no later than1 minute, 2 minutes, 3 minutes, 4 minutes, 5 minutes, 10 minutes, 15 minutes, 20 minutes, 30 minutes, or 1 hour) the time that the user reaches or is expected to reach the desired sleep duration. Such time that the adaptive alarm is turned on may or may not be the same as the user’s predetermined (or preset) alarm time. In some cases, the alarm can be turned off automatically (e.g., revert at least a portion of any change of the article of furniture induced to initiate the alarm) upon determining that the user is no longer using the article of furniture (e.g., not on the bed), the current time is after the user’s predetermined wake-up time (e.g., as set by the user previously via the GUI), or both. Alternatively, turning off the alarm may need the user’s input (e.g., via the GUI).

[0227] Example 4: Identify sleep quality of user based on biological signal

[0228] Systems and methods of the present disclosure can be utilized to identify sleep quality patterns of a user based on biological signals of the user during use of the system.

[0229] In some embodiments, the user can be sleeping on an article of furniture, such as a bed device. The article of furniture can detect one or more biological signals of the user while they are sleeping, such as heart rate and / or heart rate variability. The controller of the article of furniture can compare the current biological signals of the user while sleeping with historical biological signals collected during prior use of the article of furniture. Based on the comparison of the current biological signals and the historical biological signals, the controller can determine the sleep quality of the user. For example, if the current biological signals (e.g., heart rate and / or heart rate variability) differ than the historical biological signals, the controller can send a notification to a user device indicating a lower quality of sleep of the user during the current use of the article of furniture.

[0230] In some embodiments, the user can be sleeping on an article of furniture, such as a bed device. The article of furniture can detect one or more biological signals of the user while they are sleeping, such as heart rate and / or respiratory rate (e.g., derived from sensor data such as piezoelectric sensor data). The controller of the article of furniture can determine (e.g., compute) different sleep stages (e.g., light sleep, REM sleep, and / or deep sleep) and duration(s) thereof throughout the night. Identification of the sleep stages and durations thereof can be performed via classifiers such as machine learning algorithms. Subsequently, the identified sleep stages and durations thereof can be analyzed to determine if the user is or has experienced “low quality” sleep over night. In some cases, one or more factors such as (i) presence or absence of any particular sleep stage and / or (ii) duration of at least one particular sleep stage relative to the duration of the entire sleep (e.g., % REM sleep and / or % deep sleep relative to total sleepduration) can be utilized to determine quality of the user’s sleep over night.

[0231] In some embodiments, analysis of one or more sleep qualities of a prior sleep session can be utilized to determine a target condition of the article of furniture for a subsequent sleep session. In some embodiments, the target condition can be a temperature of the article of furniture, and a target temperature (e.g., associated with a target sleep phase) and / or a respective duration or time of onset for such target temperature can be modified for the subsequent sleep session. For example, an existing temperature adjustment profile for the article of furniture (e.g., as described in Example 3) can be modified accordingly for the subsequent sleep session.

[0232] Example 5: Labelling sleep data of the user collected by the article of furniture

[0233] Systems and methods of the present disclosure can be utilized to label sleep data of the user collected by the article of furniture during use of the system.

[0234] In some embodiments, prior to, during, or following use of the article of furniture, the user can provide (e.g., via graphical user interface (GUI) of a user application on a mobile device) to a controller (which is operatively coupled to the article of furniture) information regarding one or more user events. The one or more user events can be from the past (e.g., event(s) occurred on that day or on one or more previous days) or can be for the future (e.g., event(s) planned to occur in one or more days). The information can comprise the event(s) along with related date, time, duration, and / or location of the event(s). Non-limiting examples of the user event(s) include caffeine consumption and frequency, travel information, exercise information, current health status, food consumption and frequency, etc. The GUI can display one or more digital labels (e.g., tabs, stickers, tags, etc.), each label comprising one of pregenerated user events, such that the user can select desired digital labels for a target date. The GUI can also display an empty digital label, such that the user can generate a new digital label with a new user event that the GUI has not provided.

[0235] In some embodiments, the article of furniture can collect one or more biological signals of the user while they are sleeping. Using the one or more biological signals collected, the controller can compute a sleep score indicative of a sleep quality of the user while sleeping on the article of furniture.

[0236] In some embodiments, the one or more user events added by the user can then be associated with the one or more biological signals and / or sleep score for that use of the article of furniture. Accordingly, the controller can re-analyze the biological signal(s) and / or sleep score in view of the added user event(s), to re-generate the sleep score for the use of the article of furniture that occurred following the added user event(s). Alternatively or in addition to, the usercan manually change the sleep score of one or more previous nights and provide the correlated user event(s) as reasons for the change, and the controller can analyze the changed sleep score and the event(s) (e.g., determine correlations using a machine learning algorithm) for future uses of the article of furniture.

[0237] In some embodiments, the controller of article of furniture can analyze one or more biological signals and / or sleep scores collected over multiple uses of the article of furniture. The controller can look for correlations between lower and / or higher sleep scores and one or more user events. For example, the controller can determine that the user can experience a lower sleep score on nights when the user drank two or more cups of coffee prior to use of the article of furniture. In another example, the controller can determine that the user can experience a higher sleep score on nights when the user exercised prior to use of the article of furniture.

[0238] Example 6: Sleep score calculation based on sleep quality

[0239] Systems and methods of the present disclosure can be utilized to calculate a sleep score based on data (e.g., raw data) associated with a user of the article of furniture during use of the system. Non-limiting example of the data (e.g., raw data) can include sleep phases, sleep quality, sleep routine, duration of time slept, biological signals collected during sleep, environment signals collected during sleep, etc.).

[0240] In some embodiments, the article of furniture can collect one or more biological signals of the user while they are sleeping. The one or more biological signals of the user can be used to calculate a sleep score based on the quality of sleep of the user. For example, the sleep score can be calculated based on the relative proportion of different sleep phases (e.g., percentage of deep sleep, percentage of REM sleep, percentage of light sleep, etc.) during the night, a sleep routine of the user (e.g., whether the user has gone to bed and woken up consistently over the course of days, weeks, or months), and / or a total time slept.

[0241] In some embodiments, the calculated sleep score and / or the data (e.g., raw data) of the user of the article of furniture can be compared to a benchmark data (e.g., calculated sleep score, raw data thereof, analysis thereof such as an average value, etc.) of multiple users of comparable article of furniture. The multiple users can comprise at least about 10, at least about 50, at least about 100, at least about 500, at least about 1,000, at least about 5,000, at least about 10,000, or more users. The benchmark data can be collected over a duration of time, such as at least about 1 year, at least about 2 years, at least about 3 years, at least about 4 years, at least about 5 years, or more. The benchmark data can comprise data collected within the past 1 year, 2 years, 3 years, 4 years, 5 years, or more. For example, the comparison can show that the user of the article offumiture has a higher or lower sleep score compared to other users of the comparable article of furniture. In some embodiments, the sleep score of the user of the article of furniture can be compared to a benchmark data (e.g., the average sleep score) of the user from prior uses of the article of furniture. For example, the comparison can show that the user of the article of furniture has a higher or lower sleep score compared to prior uses of the article of furniture by the user.

[0242] FIG. 1 schematically illustrates an example graphical user interface (GUI) displaying information about the user’s sleep score (“Sleep Fitness Score”) for a selected day or date (Friday or “F”). The sleep score can be provided in one or more ratings such as numerical score (e.g., 88 out of 100) or textual descriptions (e.g., Good). The GUI can provide one or more factors utilized for determining the sleep score, such as the sleep quality (“Quality”), the sleep routine of the user (“Routine”), and the total time slept (“Time slept”). Each of the factors can be provided with their own analysis or scoring (e.g., as determined by the classifier), in order for the user to see, for example, possible reasons for their sleep score being higher or lower than expected. In some cases, the GUI can allow the user to select at least one of the one or more factors on the GUI, to trigger the GUI to display more detailed information about the selected factor. For example, as demonstrated in FIG. 1, the sleep quality factor can be selected, to prompt the GUI to display overall progress of the user’s sleep stages during the selected night (e.g., color schematics to denote different sleep stages, x-axis to denote time, and y-axis to denote a biological signal of the user or that of the bed or environment detected during sleep).

[0243] Example 7: New temperature adjustment profiles

[0244] Systems and methods of the present disclosure can be utilized to generate and apply new temperature adjustment profiles for the article of furniture, based at least in part on a user event.

[0245] In some embodiments, a user’s profile can comprise a preset temperature adjustment profile for the user’s article of furniture to be activated when the user is using the article of furniture (e.g., bed device). When the user provides one or more user events, whether past (e.g., exercise information, food consumption information, travel information, etc.) or future (e.g., about to take a nap, etc.), a controller (e.g., a computer processor) can modify the preset temperature adjustment profile to accommodate or match the user event(s). For example, when the user wishes to nap (e.g., during daytime) instead of taking a full night sleep, the controller may shorten the present temperature adjustment profile. In another example, when the user wishes to recover from an exercise, the controller may adjust the temperature profile to compensate for the offset in the user’s condition (e.g., different body temperature, different heartrate, different breathing rate, different sleep pattern, etc., as compared to a normal night’s sleep).

[0246] Example 8: New temperature adjustment profiles

[0247] Systems and methods of the present disclosure can be utilized to generate and apply new temperature adjustment profiles for the article of furniture, based at least in part on data associated with a user of the article of furniture during use of the system. In some embodiments, the data can comprise data collected prior to current use (or the next use) of the article of furniture by the user. For example, the data can comprise data collected during a previous night’s sleep session of the user (e.g., yesterday’s sleep session). Alternatively or in addition to, the data can comprise data collected during current use (e.g., current sleep session) of the article of furniture by the user. In some embodiments, the data can comprise one or more members (or one or more sub-members thereof) from the user factor, the sleep factor, the environment factor, any combination thereof, and / or any analysis thereof, as provided herein. Non-limiting example of such data can include sleep phases, sleep quality, sleep routine, duration of time slept, biological signals collected during sleep, environment signals collected during sleep, etc.).

[0248] In some embodiments, a user’s profile can comprise a preset temperature adjustment profile for the user’s article of furniture to be activated when the user is using the article of furniture (e.g., bed device). The temperature adjustment profile can have been used by the user during one or more prior sleep sessions (e.g., yesterday’s sleep session), and such temperature adjustment profile of the article of furniture can be modified in accordance with any one of the methods provided herein, e.g., to enhance sleep quality of the user during current sleep session (or the next sleep session).

[0249] In some embodiments, the modified temperature adjustment profile can be personalized (e.g., hyper-personalized) and responsive to the user, such that the user can sleep better than before. In some embodiments, the temperature adjustment profile can be modified based on review or analysis of the user’s sleep trends and personal patterns (e.g., minute by minute), to make temperature adjustments of the bed device to effect the user to get more time in one or more target sleep phases (e.g., REM sleep, deep sleep, light sleep). Such modification of the temperature adjustment profile can be performed in real-time by analyzing user sensor data obtained substantially in real-time by the user sensor of the bed device, e.g., to determine identity and / or quality of the user’s current sleep cycle or sleep phase. Alternatively or in addition to, the modification of the temperature adjustment profile can be performed by reviewing the user’s previous night’s sleep to learn what sleep phases the user was running low on. In some embodiments, in addition to analysis of current and / or past sleep phase or quality thereof, themodification of the temperature adjustment profile can be performed based on additional consideration such as the user’s biological sex, sleep routine, local weather, etc.

[0250] The systems and methods for generating, modifying, and / or applying new temperature adjustment profiles as provided herein are provided as examples, and such systems and methods can be utilized accordingly for generating, modifying, and / or applying new adjustment profiles for one or more environmental conditions or factors of the bed device (e.g., ambient temperature air quality, humidity, pressure, oxygen level, nitrogen level, noise, light, etc.).

[0251] Example 9: Snore detection model

[0252] Systems and methods of the present disclosure can be utilized to generate and apply new temperature adjustment profiles for the article of furniture, based at least in part on data associated with a user of the article of furniture during use of the system.

[0253] Locating individual snores in the time and frequency domain, rather than only detecting which minutes of piezo data contain snoring, allows for computing the intensity of each individual snore and provides for a more robust estimate of snoring intensity. Noise was detected due to movement. These noise segments were explicitly excluded to prevent these segments from biasing the intensity measurements. In order to do this, a model adapted from computer vision was utilized and specialized to the task of detecting snores. An example of bounding boxes generated by the computer vision model are shown in FIG. 3.

[0254] The model was adapted to the snore and noise detection task. The model was applied on 16 previously-unseen minutes of data from the validation set. The model identified which side the snoring vibrations originated from, placed a bounding box around each snore, and identified aggressors such as noise caused by sleeper movement and “other” vibrations such as speech or vibration alarms. In this visualization, bounding boxes for each identified snore were overlaid on top of a spectrogram representation of each minute of left / right piezo vibration data. The 16 different minutes of piezo data were laid out in a grid (FIG. 4). Each labeled bounding box was also accompanied by a confidence estimate.

[0255] After a number of iterations on training label generation, labeled task definitions were set. Bounding boxes were placed around each task definition: (i) Snore; (ii) Noise (large spikes most commonly caused by movement); and (iii) Other (vibrations that could be confused for snoring such as speech, wheezing, or vibration alarms). The model was trained via transfer learning from checkpoint pretrained for a computer vision task on the Common Objects in Context (COCO) dataset. The model was trained to accept spectrograms with frequency axis ranging from 0-250Hz, time axis ranging from 0-60 seconds, and channel axis coded as (left,right, dummy). The dummy channel was filled with zeros and required in order to match the standard RGB image channel dimensions to take advantage of the pretrained COCO weights. The model was trained to generate bounding boxes labeled with the following six classes: (1) 'snore_lef : 0; (2) 'snore_right' : 1; (3) 'noise_lef : 2; (4) 'noise_right' : 3; (5) 'other lef : 4; and (6) 'other right' : 5.

[0256] Example 10: Sleep-stage adjustment of temperature profiles

[0257] Systems and methods of the present disclosure can be utilized to generate and apply new temperature adjustment profiles for the article of furniture, based at least in part on data associated with a user of the article of furniture during use of the system.

[0258] The adjustment is based on research of sleep physiology and A / B testing of 14,000 members for 1.2 million hours of sleep. The adjustment accounts for three factors. (1) sleep stage: every minute, the adjustment algorithm optimizes the time spent in deep, light, and REM sleep by updating temperature in real-time based on current sleep stage. A user can increase the time spent in REM and deep sleep via warmer temperatures in REM sleep and cooler temperatures in deep sleep. Light sleep is similar to deep sleep (they are both non-REM sleep stages), where cooler temperatures promote light sleep, especially in the second half of the night.(2) Sleep deficiency: the algorithm looks at the previous night’s sleep to see what sleep stages the user needs more time in for the current night. (3) Demographics: the algorithm further adjusts each temperature based on the user’s biological sex and age.

[0259] A logic tree was developed to combine the above factors. Early phase comprised the first half of the night and late phase comprised the second half of the night. Units corresponded to temperature and were on a non-linear scale. When close to zero, each +1 corresponded to +1 °C. Farther from 0 (for example, 10) corresponded to greater change in temperature, where 10 can correspond to +15 °C.

[0260] Deep sleep

[0261] For deep sleep, if a user had normal deep sleep last night, a change of 0 to -0.5 was given during Early phase. For those users less than 50 years of age, a change of -0.5 was given to help promote deep sleep. For those users greater than 50 years of age, a change of 0 was given because older adults need warmer deep sleep temperatures and are more sensitive to temperature changes. The result was +4 min / night (+2 hours / per month) in Early phase deep sleep.

[0262] If a user had below-average deep sleep last night, a change of -0.5 (>50 years) to -1 (<50 years) during Early phase was given. The result was cooler temperatures of -1 have beenshown to improve deep sleep by 3 minutes / night or 1.5 hours / month for those with below- average deep sleep.

[0263] During Late phase, there is very little deep sleep (typically <10 minutes). A +0.5 change was applied to users given that core temperatures are warmer during Late Phase. The result was across the tested users, these temperature profiles for both normal and below-average deep sleep lead to 9% fewer nights with low deep sleep.

[0264] REM sleep

[0265] If a user had normal REM sleep the previous night, a temperature change of +0.5 was given in both Early and Late phases. If a user had below-average REM sleep last night, a temperature change of +1 was given in Early phase and either +0.5 (>50 years of age) or +1 (<50 years of age) in Late phase. For those users greater than 50 years of age lower REM temperature changes were given since they are more sensitive to temperature changes. The Late phase temperature profiles were selected to prevent the loss of REM sleep by 2 min / night. The results for both normal and below-average REM nights were increasing initial phase REM sleep by 3-5 minutes / night or 2.5 hours / month. Across the users, there are 19% fewer nights with low initial REM sleep.

[0266] Light sleep

[0267] Light sleep is similar to deep sleep in that cooler temperatures can promote light sleep. More light sleep can occur in the second half of the night, and so a -0.5 change is applied to light sleep in Late phase. The result was improvement in light sleep for users by 11 minutes / night or 5.5 hours / month.

[0268] In Early Phase, light sleep is set at +0 for everyone except for women greater than 50 years of age who get a -0.5 change. This is because most women greater than 50 years of age get hot flashes during light sleep, so preemptive cooling during light sleep can help minimize waking up due to hot flashes.

[0269] Wake time

[0270] During Early phase, cooler temperatures for men were tested and warmer temperatures for women were tested when they are awake. Women get a +0.3 change and men get a +0 change. The result is men have 1.5 less minutes / night awake, and women have 0.5 less minutes / night awake. During Late phase, men and women both have a +0 change so that users experience the temperature they selected when they wake up. Additionally, people can be more susceptible to temperature changes in the second-half vs. first-half of the night.

[0271] Total sleep time (TST)

[0272] As a result of increasing time in each of these sleep stages, total sleep time is increasing for those users with normal TST (greater than 6.75 hours of sleep) and below-average TST (less than 6.75 hours of sleep). The result is those users with normal TST see +10 minutes TST / night, or +5 hours TST / month. Those users with below-average TST see +20 minutes TST / night or +10 hours TST / month.

[0273] In the decision tree, there are 3 main parts: (1) initial deep recommendation, (2) initial REM recommendation, and (3) final REM recommendation. A user will get a separate recommendation for each of these phases. All of these parts will have a different value for each branch of the decision tree.

[0274] Potential splits in the tree can include: (1) age group, (2) biological sex, (3) low initial REM sleep, defined as less than 15% REM sleep in the first 4 hours of sleep after sleep onset + 20 minutes from the previous night, (4) low final REM sleep, defined as less than 20% REM sleep after the first 4 hours of sleep after sleep onset + 20 minutes from the previous night, (5) low initial deep sleep, defined as less than 20% deep sleep in the first 4 hours of sleep after sleep onset + 20 minutes from the previous night, (6) normal initial REM sleep, defined as greater than 15% REM sleep in the first 4 hours of sleep after sleep onset + 20 minutes from the previous night, (7) normal final REM sleep, defined as greater than or equal to 20% REM sleep after the first 4 hours of sleep after sleep onset + 20 minutes from the previous night, and (8) normal initial deep sleep, defined as greater than or equal to 20% deep sleep in the first 4 hours of sleep after sleep onset + 20 minutes from the previous night.

[0275] Example 11: Snoring Mitigation via Machine Learning

[0276] Many people struggle to sleep due to snoring, whether it be their own or their partners. Loud snoring can be extremely disruptive to a good night’s sleep, and snoring has also been linked to anything from sleep apnea to the common cold.

[0277] Machine learning was leveraged to detect snoring as it happens, with 93% accuracy. The sensors inside the bed device detect snoring due to the vibrations picked up by the sensors that span the entire width of the bed device. These sensors are located at chest level and pick up vibrations from snoring through the chest wall. These vibrations are then fed into one of algorithms to analyze a user’s (or second user’s) snores. Through this algorithm, the data are analyzed by a neural network trained using data from more than 2,500 known snorers, to provide accurate information to users about their sleep every night.

[0278] While many products in the market use an app to detect snoring via a microphone, unique sensors within the bed device allow for detection of the vibrations from the user’s body during snoring, rather than sound. By not relying on sound, it is determined which side of the bedthe snoring is coming from, making it easier to provide reliable sleep data to users in the morning.

[0279] When users wake up in the morning, users will see a comprehensive report of their snoring, with details including when they snored, for how long, and even how intensely - all without using a wearable device (FIG. 5).

[0280] To make snore mitigation more effortless for the user, we determined that an adjustable section of the bed device was determined to be a seamless approach to the problem. With the bed device described herein, detection and also reduction and / or stopping of snoring was accomplished using the movement of the bed device.

[0281] The bed device’s adjustable section gently elevates the bed when snoring is detected - in real time and without disruption. The adjustable section was designed to be silent and gentle in movement to avoid waking either user when making adjustments. This was achieved by using custom motors and control systems to move the bed device with a low minimum speed. The gradual angle change moves at a nearly imperceptible rate, supporting our ultimate goal of keeping users asleep throughout the night, undisturbed.

[0282] Another advantage of the bed device described herein is that it reacts to snoring in real-time, making instant adjustments. A user can review how the bed device reduced snoring each morning during sleep via a health report.

[0283] Clinical Study for Snoring Reduction

[0284] A clinical was performed which show snoring was reduced by 45% in users.

[0285] Snoring detection and mitigation was tested in a clinical study that captured over 200 nights of data while people slept on the bed device. The study revealed that participants saw up to a 31% decrease in snoring when the bed device’s head angle was raised to 6 degrees, and up to a 45% decrease in snoring when raised to 15 degrees.

[0286] To assess how the bed device impacts snoring duration, 39 subjects (33 male, 6 female, 38 ± 12 y) were tested for 281 nights on the bed device. To understand typical snoring patterns, snoring data were collected for 3-5 nights where the bed device was flat the entire night. Subjects then slept for 3 nights at angles of 6, 10, or 15 degrees, where the head of the bed device section would automatically rise to one of these angles when snoring was detected.

[0287] All snore data used for these analyses was collected via a ground truth microphonebased snore detection application. The subject’s mobile device, with the application, was placed on a bedside table one foot from the person’s head (when lying in bed). The application provides the total percentage of time the person snored during the night. This percentage of nightly snoring was used to calculate the percent change in snoring for each subject as: (snore % -baseline snore %) / (baseline snore %). Then, the average percent change in snoring across all subjects was calculated. For 76% of subjects, higher angles further reduced snoring.EMBODIMENTS

[0288] The following non-limiting embodiments provide illustrative examples of the invention, but do not limit the scope of the invention.

[0289] Embodiment 1. A computer-implemented method comprising:(a) receiving a user sensing data associated with a user, wherein the user sensing data is of a first data type and is generated by a user sensor while the user is using an article of furniture;(b) applying the user sensing data as an input to a machine learning model that is trained to determine a target condition of the subject, wherein the machine learning model is trained based on a ground truth data comprising a training sensing data of a second data type, the first data type and the second data type being different data types;(c) determining, using the machine learning model, the target condition of the subject from the user sensing data, wherein the determining comprises comparing a feature derived from the user sensing data and a feature derived from the training sensing data; and(d) generating an instruction for the article of furniture to change a condition of the article of furniture based on the determined target condition of the subject in (c), optionally wherein:(1) the training sensing data comprises a sensing data of the second data type associated with one or more individuals other than the user; and / or(2) the training sensing data is derived from clinical data; and / or(3) the machine learning model is not trained based on data of the first data type; and / or(4) the comparing in (c) comprises one or more of:(i) converting at least a portion of the user sensing data to the second data type; and / or(ii) converting at least a portion of the training sensing data to the first data type; and / or(iii) converting at least a portion of the user sensing data to a third data type, wherein the third data type is different from the first and second data types and is informative of the target condition; and / or(5) the generating in (d) is performed substantially in real-time; and / or(6) the machine learning model is further trained to determine a target degree of change of the condition of the article of furniture based on the determined target condition; and / or(7) the method comprises, subsequent to (d):(A) receiving an additional user sensing data that is generated by the user sensor subsequent to generating the user sensing data;(B) determining, using the machine learning model, a change in the target condition of the user from the additional user sensing data; and(C) generating an additional instruction for the article of furniture to reduce at least a portion of the change of the condition of the article of furniture; and / or(8) the first data type is the pressure data, and wherein the second data type is the audio data; and / or(9) the article of furniture comprises at least one actuator configured to induce movement of at least a portion of the article of furniture, wherein the instruction is for directing the at least one motor to induce the movement; and / or(10) the movement is vibration; and / or(11) the movement is a relative movement between at least two different sections of the article of furniture; and / or(12) the target condition comprises a sleep disordered breathing, and wherein the change in the condition of the article of furniture is to effect reduced level of the sleep disordered breathing of the subject; and / or(13) the sleep disordered breathing is selected from the group consisting of snoring, apnea, and hypopnea; and / or(14) the user sensor is part of the article of furniture; and / or(15) the article of furniture is a mattress or a mattress cover; and / or(16) the steps (a)-(d) are occurring over the user’s current sleep; and / or(17) the second data type is image data, and wherein the first data type is nonimage data; and / or(18) the image data is two-dimensional image data; and / or(19) the first data type is electrical signal data, pressure data, or any combination thereof; and / or(20) the second data type is image data, and where-in the first data type is electrical signal data; and / or(21) the user sensor is a piezoelectric sensor; and / or(22) the machine learning model is configured to convert the user sensing data to the second data type, thereby generating a converted user sensing data; and / or(23) the machine learning model is configured to analyze the con-verted user sensing data to determine the target condition of the user; and / or(24) the machine learning model is configured to (i) convert the user sensing data to the second data type and (ii) analyze the converted user sensing data to determine the target condition of the user; and / or(25) the target condition comprises a sleep disorder; and / or(26) the target condition comprises snoring.

[0290] Embodiment 2. A system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the one or more computer processors, implements the method of Embodiment 1, optionally wherein:(1) the system further comprises the user sensor; and / or(2) the system further comprises the article of furniture; and / or(3) the system further comprises the at least one actuator; and / or(4) wherein the at least one actuator is coupled to or is a part of the article of furniture.

[0291] Embodiment 3. A computer-implemented method comprising:(a) providing, from a database, a predetermined adjustment profile for a condition of an article of furniture of a user, the predetermined adjustment profile comprising a target condition and a target duration for the target condition, wherein the predetermined adjustment profile is executable by a regulating unit that is operatively coupled to the article of furniture and configured to regulate the condition of the article of furniture during a sleep session of the user;(b) receiving a user input comprising at least one user event;(c) automatically modifying the target duration for the target condition based at least in part on the at least one user event, to generate a modified adjustment profile of the condition of the article of furniture; and(d) storing, in the database, the modified adjustment profile for use by the regulating unit,optionally wherein:(1) the at least one user event is not collected during a prior sleep session of the user; and / or(2) the at least one user event comprises exercise information, food consumption information, or travel information; and / or(3) the at least one user event comprises a past event; and / or(4) the at least one user event comprises a future event; and / or(5) the future event comprises a nap event; and / or(6) the target condition is a target temperature of the article of furniture; and / or(7) the target condition is a target condition of an environment of the article of furniture; and / or(8) the predetermined adjustment profile comprises an additional target condition and an additional target duration for the additional target condition, wherein the method further comprises automatically modifying the additional target duration for the additional target condition based at least in part on the at least one user event, wherein the target condition and the additional target condition are different; and / or(9) the method further comprises monitoring a quality of sleep of the user during or subsequent to executing the modified adjustment profile by the regulating unit; and / or(10) the monitoring comprises analyzing one or more biological signals of the user, wherein the one or more biological signals are generated by a user sensor during the at least the portion of the usage; and / or(11) the user sensor is part of the article of furniture; and / or(12) the method further comprises:(A) directing the regulating unit to execute at least a portion of the modified adjustment profile for a usage of the article of furniture by the use;(B) subsequent to (A), monitoring a quality of at least a portion of the usage;(C) subsequent to (B), modifying at least a portion of the modified adjustment profile based on the monitored quality, to generate a different adjustment profile that is usable for a subsequent usage of the article of furniture by the user; and(D) storing, in the database, the different predetermined adjustment profile; and / or(13) the method further comprises:(A) directing the regulating unit to execute at least a portion of the modified adjustment profile for a usage of the article of furniture by the user;(B) subsequent to (A) and during the usage, monitoring a quality of at least a portion of the usage; and(C) subsequent to (B) and during the usage, modifying the modified adjustment profile based on monitored quality, to generate a different adjustment profile; and(D) directing the regulating unit to execute at least a portion of the different adjustment profile for at least a portion of a remainder of the usage; and / or(14) the user input is provided by the user via a graphical user interface (GUI) of a user application provided on a user device; and / or(15) the user input is retrieved from the database or an additional database associated with the user; and / or(16) the predetermined adjustment profile is specifically generated for each individual user; and / or(17) the user sensor is part of the article of furniture; and / or(18) the article of furniture is a mattress or a mattress cover.

[0292] Embodiment 4. A system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the one or more computer processors, implements the method of Embodiment 3, optionally wherein:(1) the system further comprises the user sensor; and / or(2) the system further comprises the article of furniture; and / or(3) the system further comprises the regulating unit; and / or(4) at least a portion of the regulating unit is coupled to or is a part of the article of furniture.

[0293] Embodiment 5. A computer-implemented method for controlling a condition of an article of furniture, the method comprising:(a) determining, by a computer algorithm, a target condition of a user based on a user sensing data generated by a user sensor while the user is using the article of furniture, wherein the computer algorithm comprises a machine learning model trained to determine the target condition of the subject based on a training data, wherein the training data and the user sensing data are different data types; and(b) generating an instruction for the article of furniture to change the condition of the article of furniture based on the determined target condition of the user, optionally wherein:(1) the determining and the generating are performed over a same use of the article of furniture; and / or(2) the determining and the generating are performed substantially in real-time; and / or(3) the training data comprises an image data, and wherein the user sensing data comprises a non-image data; and / or(4) the image data comprises a two-dimensional image data; and / or(5) the user sensing data comprises electrical signal data, pressure data, or any combination thereof; and / or(6) the training data comprises an image data, and wherein the user sensing data comprises an electrical signal data; and / or(7) the user sensor comprises a piezoelectric sensor; and / or(8) the computer algorithm is configured to convert the user sensing data to a same data type as the training data, thereby generating a processed user sensing data; and / or(9) the machine learning model is configured to analyze the processed user sensing data to determine the target condition of the user; and / or(10) the processed user sensing data comprises a single image data comprising a plurality of channels comprising (i) a first channel indicative of a first user sensing data from a first user sensor of the article of furniture and (ii) a second channel indicative of a second user sensing data from a second user sensor of the article of furniture; and / or(11) the first user sensing data and the second user sensing data are associated with different users of the article of furniture; and / or(12) the target condition comprises a sleep disorder selected from the group consisting of insomnia, hypersomnia, sleep apnea, circadian rhythm sleep disorder, restless legs syndrome, and sleep deficiency; and / or(13) the target condition comprises snoring; and / or(14) the training data comprises a data associated with one or more individuals; and / or(15) the data associated with one or more individuals does not comprise the user; and / or(16) the training data is derived from clinical data; and / or(17) the condition of the article of furniture comprises a movement of at least a portion of the article of furniture, wherein the article of furniture comprises at least one actuator configured to induce the movement; and / or(18) the instruction is configured to direct at least one motor to induce the movement; and / or(19) the movement comprises a vibration; and / or(20) the movement comprises an angular and / or height adjustment of the portion of the article of furniture; and / or(21) wherein the actuator is coupled to the article of furniture; and / or(22) the condition of the article of furniture comprises a temperature of at least a portion of the article of furniture; and / or(23) the article of furniture is a bed device; and / or(24) the article of furniture is a mattress or mattress cover; and / or(25) the user sensor is part of the article of furniture; and / or(26) the method further comprises (i) determining, using the machine learning model, a change in the target condition of the user based on an additional user sensing data; and (ii) generating an additional instruction for the article of furniture to change the condition of the article of furniture based on the change in the target condition of the user; and / or(27) the determining comprises comparing at least a portion of the user sensing data to at least a portion of the additional user sensing data; and / or(28) the comparing comprises identifying a decreased degree of the target condition of the user; and / or(29) the additional user sensing data is generated by the user sensor; and / or(30) the additional user sensing data is generated by a second user sensor.

[0294] Embodiment 6. A system for controlling a condition of an article of furniture, the system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the processor, implements the method of Embodiment 5.

[0295] Embodiment 7. A computer-implemented method for controlling a condition of an article of furniture, the method comprising:(a) determining a sleep disorder of a user of the article of furniture when a threshold indicator of a user sensing data is identified, wherein the user sensing data is generated by a user sensor while the user is using the article of furniture, and wherein the threshold indicator is based on at least one user factor; and(b) generating an instruction for the article of furniture to change the condition of the article of furniture based on the determined sleep disorder of the user, optionally, wherein:(1) the determining is based on a machine learning model trained to identify the threshold indicator from the user sensing data; and / or(2) the determining is performed in absence of a machine learning model; and / or(3) the determining is based on a time-frequency analysis; and / or(4) the time-frequency analysis comprises Fourier transformation, Laplace transformation, or Wavelet transformation; and / or(5) the time-frequency analysis comprises Fourier transformation, wherein the Fourier transformation comprise one or more members selected from the group consisting of a Discrete Fourier Transformation (DFT), a Fast Fourier Transformation (FFT), a Principal Component Analysis (PCA) transformation, an Inverse Discrete Fourier Transformation, and an Inverse Fast Fourier Transformation; and / or(6) the at least one user factor comprises a plurality of users factors; and / or(7) the at least one user factor comprises one or more members selected from the group consisting of biological sex, age, weight, health condition, exercise information, and diet information (or food consumption information); and / or(8) the sleep disorder comprises snoring; and / or(9) the determining and the generating are performed over a same use of the article of furniture; and / or(10) the determining and the generating are performed substantially in real-time; and / or(11) the condition of the article of furniture comprises a movement of at least a portion of the article of furniture, wherein the article of furniture comprises at least one actuator configured to induce the movement; and / or(12) the movement comprises a vibration, an angular adjustment, a height adjustment, or a combination thereof; and / or(13) the article of furniture is a bed device; and / or(14) the article of furniture is a bed mattress or a mattress cover; and / or(15) the user sensor is part of the article of furniture; and / or(16) the method further comprises (i) determining a change in the sleep disorder of the user based on an additional user sensing data; and (ii) generating an additional instruction for the article of furniture to change the condition of the article of furniture based on the change in the sleep disorder of the user; and / or(17) the comparing comprises identifying a decreased degree of the sleep disorder of the user.

[0296] Embodiment 8. A system for controlling a condition of an article of furniture, the system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the processor,implements the method of Embodiment 7.

[0297] Embodiment 9. A computer-implemented method for controlling a condition of an article of furniture of a user, the method comprising:(a) analyzing a quality of a target sleep phase of the user from a prior sleep session of the user; and(b) automatically modifying a predetermined adjustment profile of the condition of the article of furniture based on the analyzed quality of the target sleep phase, to generate a modified adjustment profile of the condition of the article of furniture, wherein the automatically modifying comprises changing a target value of the condition that is associated with a target sleep phase of the user, and wherein the modified adjustment profile is executable by the article of furniture to regulate the condition of the article of furniture, optionally, wherein:(1) the article of furniture comprises a regulating unit configured to regulate the condition of the article of furniture based on the modified adjustment profile; and / or(2) the method further comprises storing, in a database, the modified adjustment profile for use by the article of furniture; and / or(3) the prior sleep session of the user was on the article of furniture operated based on the predetermined adjustment profile; and / or(4) the target sleep phase is selected from the group consisting of REM sleep, deep sleep, and light sleep; and / or(5) the analyzing comprises determining that a duration of the target sleep phase is less than a target duration of the target sleep phase; and / or(6) the prior sleep session is an immediately preceding sleep session; and / or(7) the predetermined adjustment profile was utilized to regulate the condition of the article of furniture during the prior sleep session; and / or(8) the analyzing is based on a sensor data associated with the user collected by a user sensor during the prior sleep session; and / or(9) the user sensor is part of the article of furniture and / or(10) the condition is a temperature of the article of furniture, and wherein the regulating unit is a temperature regulating unit; and / or(11) the temperature regulating unit is coupled to the article of furniture; and / or(12) the method further comprises sending at least a portion of the modified adjustment profile to the regulating unit; and / or(13) the method further comprises repeating (a) and (b) a plurality of times for a plurality of different target sleep phases of the user; and / or(14) the method further comprises performing (a) and (b) prior to or subsequent to determining a presence of the user on the article of furniture; and / or(15) the article of furniture is a bed device selected from the group consisting of a mattress, a pillow, and a cover thereof; and / or(16) the prior sleep session comprises a single sleep session, and wherein the method further comprises analyzing the quality of the target sleep phase from the entirety of the sleep session; and / or(17) the prior sleep session comprises a single sleep session, and wherein the method further comprises analyzing the quality of the target sleep phase from a fraction and not the entirety of the single sleep session; and / or(18) the fraction comprises a portion of a first half of the single sleep session; and / or(19) a duration of the fraction is less than 80% of a duration of the singles sleep session; and / or(20) the predetermined adjustment profile comprises a plurality of sleep cycles, each sleep cycle comprising a same target sleep phase, and wherein the target value of the condition associated with the same target sleep phase is automatically modified to a same value for different sleep cycles; and / or(21) the predetermined adjustment profile comprises a plurality of sleep cycles, each sleep cycle comprising the same target sleep phase, and wherein the target value of the condition associated with the same target sleep phase is automatically modified differently for different sleep cycles; and / or(22) the predetermined adjustment profile comprises different types of the target sleep phase, and wherein the target value is automatically modified to a same value regardless of the different types of the target sleep phase; and / or(23) the predetermined adjustment profile comprises different types of the target sleep phase, and wherein the target value is automatically modified differently depending on the different types of the target sleep phase; and / or(24) (1) the target value is automatically modified if the quality of the target sleep phase is below a predetermined threshold and (2) the target value is not automatically modified if the quality of the target sleep phase is at or above the predetermined threshold; and / or(25) the target value is automatically modified differently depending on whether (1) the quality of the target sleep phase is below a predetermined threshold or (2) the quality of the targetsleep phase is at or above the predetermined threshold; and / or(26) the target value is automatically modified by different degrees depending on whether (1) or (2) is met; and / or(27) the target value is automatically modified by a greater degree when (1) is met as compared to when (2) is met; and / or(28) the method further comprises automatically modifying the target value based on a combination of a plurality of factors selected from the group consisting of the analysis of the quality of the target sleep phase, age of the user, biological sex of the user, geolocation of the user, and health condition of the user, wherein different combinations of the plurality of factors correspond to different modifications of the target value; and / or(29) the plurality of factors comprises three or more factors; and / or(30) the plurality of factors comprises (i) the quality of the target sleep phase and (ii) how the age of the user compares relative to a predetermined threshold age or range thereof; and / or(31) the predetermined threshold age is greater than or equal to about 30; and / or(32) the predetermined threshold age is less than or equal to about 60; and / or(33) the plurality of factors comprises (i) the quality of the target sleep phase and (ii) whether the biological sex is male or female.

[0298] Embodiment 10. A system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the one or more computer processors, implements the method of Embodiment 9.

[0299] Embodiment 11. A computer-implemented method for controlling a condition of an article of furniture of a user, the method comprising:(a) receiving a user input comprising at least one user event; and(b) automatically modifying a predetermined adjustment profile of the condition of the article of furniture based on the at least one user event, to generate a modified adjustment profile of the condition of the article of furniture, wherein the automatically modifying comprises changing a target duration for subjecting at least a portion of the article of furniture to a target value of the condition, and wherein the modified adjustment profile is executable by the article of furniture to regulate the condition of the article of furniture, optionally, wherein:(1) the at least one user event is not collected during a prior sleep session of the user; and / or(2) the at least one user event comprises exercise information, food consumptioninformation, travel information, or any combination thereof; and / or(3) the at least one user event comprises a past event; and / or(4) the at least one user event comprises a future event; and / or(5) the future event comprises a nap event; and / or(6) the condition is a temperature of the article of furniture; and / or(7) the condition is a condition of an environment of the article of furniture; and / or(8) the condition of the environment is a humidity or a temperature; and / or(9) the method further comprises monitoring a quality of sleep of the user during or subsequent to executing the modified adjustment profile by the article of furniture or the regulating unit thereof; and / or(10) the monitoring comprises analyzing one or more biological signals of the user; and / or(11) the one or more biological signals are generated by a user sensor during a portion of use of the article of furniture by the user; and / or(12) the user sensor is part of the article of furniture; and / or(13) the method further comprises: (i) directing the article of furniture or the regulating unit thereof to execute at least a portion of the modified adjustment profile for a usage of the article of furniture by the user; (ii) monitoring a quality of the user’s usage; and (iii) modifying at least a portion of the modified adjustment profile based on the monitored quality, to generate a different adjustment profile that is usable for (1) a subsequent usage of the article of furniture by the user, (2) a remainder portion of the usage, or (3) both; and / or(14) the user input is provided by the user via a graphical user interface (GUI) of a user application provided on a user device; and / or(15) the user input is retrieved from the database or an additional database associated with the user; and / or(16) the predetermined adjustment profile is specifically generated for each individual user; and / or(17) the article of furniture is a bed device; and / or(18) the article of furniture is a mattress or a mattress cover.

[0300] Embodiment 12. A computer computer-implemented method for controlling a condition of an article of furniture, the method comprising: changing a condition of the article of furniture as an alarm to wake up a user of the article of furniture, wherein the changing is based on (i) a target sleep duration of the user and (ii) a time when the user falls asleep on the article of furniture or when the user begins the current use of the article of furniture,optionally, wherein:(1) the changing is based on the time when the user falls asleep on the article of furniture; and / or(2) the changing is based the time when the user begins the current use of the article of furniture; and / or(3) the method further comprises determining the time based on a user sensing data generated by a user sensor, optionally wherein the user sensor is a part of the article of furniture; and / or(4) the changing is performed upon determining that the user has reached or is expected to reach the target sleep duration during the current use; and / or(5) the method further comprises assessing or tracking an actual sleep duration of the user during the current use, wherein an onset of the actual sleep duration of the user is determined based on the time, optionally wherein the assessing or tracking is performed a plurality of times throughout the current use; and / or(6) the condition of the article of furniture comprises a movement or vibration of at least a portion of the article of furniture; and / or(7) the condition of the article of furniture comprises a temperature of at least a portion of the article of furniture; and / or(8) the method further comprises retrieving the target sleep duration from a database associated with the user; and / or(9) the method further comprises receiving the target sleep duration from the user via a graphical user interface (GUI) operatively coupled to the article of furniture; and / or(10) the method further comprises turning off the changing or the alarm based on determining that the user is not present on the article of furniture, a current time is after the user’s predetermined alarm time, or both.

[0301] Embodiment 13. A system comprising: a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the one or more computer processors, implements the method of Embodiment 11 or 12.

[0302] Systems and methods of the present disclosure may be combined with or modified by additional systems comprising an article of furniture (e.g., a bed device) and methods of use thereof. For example, systems and methods of detecting a biological signal or a condition (e.g., a sleep disorder) of a user of an article of furniture, regulating a temperature or configuration of an article of furniture, regulating a biological signal or condition (e.g., a sleep disorder) of the useron an article of furniture, regulating operation of other devices operatively coupled to the article of furniture are described in U.S. Patent Publication No. 2015 / 0351556 (“BED DEVICE SYSTEM AND METHODS”), U.S. Patent Publication No. 2016 / 0128488 (“APPARATUS AND METHODS FOR HEATING OR COOLING A BED BASED ON HUMAN BIOLOGICAL SIGNALS”), U.S. Patent Publication No. 2017 / 0135882 (“ADJUSTABLE BEDFRAME AND OPERATING METHODS FOR HEALTH MONITORING”), U.S. Patent Publication No. 2017 / 0135632 (“DETECTING SLEEPING DISORDERS”), U.S. Patent Publication No. 2020 / 0405998 (“SLEEP POD”), and U.S. Patent Publication No. 2021 / 0315389 (“SYSTEMS AND METHODS FOR REGULATING A TEMPERATURE OF AN ARTICLE OF FURNITURE”), each of which is incorporated herein by reference in its entirety.

[0303] While preferred embodiments of the present invention have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. It is not intended that the invention be limited by the specific examples provided within the specification. While the invention has been described with reference to the aforementioned specification, the descriptions and illustrations of the embodiments herein are not meant to be construed in a limiting sense. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from the invention. Furthermore, it shall be understood that all aspects of the invention are not limited to the specific depictions, configurations or relative proportions set forth herein which depend upon a variety of conditions and variables. It should be understood that various alternatives to the embodiments of the invention described herein may be employed in practicing the invention. It is therefore contemplated that the invention shall also cover any such alternatives, modifications, variations or equivalents. It is intended that the following claims define the scope of the invention and that methods and structures within the scope of these claims and their equivalents be covered thereby.

Claims

CLAIMSWHAT IS CLAIMED IS:

1. A computer-implemented method for controlling a condition of an article of furniture, the method comprising:(a) determining, by a computer algorithm, a target condition of a user based on a user sensing data generated by a user sensor while the user is using the article of furniture, wherein the computer algorithm comprises a machine learning model trained to determine the target condition of the subject based on a training data, wherein the training data and the user sensing data are different data types; and(b) generating an instruction for the article of furniture to change the condition of the article of furniture based on the determined target condition of the user.

2. The method of claim 1, wherein the determining and the generating are performed over a same use of the article of furniture.

3. The method of any one of claims 1 or 2, wherein the determining and the generating are performed substantially in real-time.

4. The method of claim 1 or 2, wherein the training data comprises an image data, and wherein the user sensing data comprises a non-image data.

5. The method of claim 4, wherein the image data comprises a two-dimensional image data.

6. The method of any one of claims 1-4, wherein the user sensing data comprises electrical signal data, pressure data, or any combination thereof.

7. The method of any one of claims 1-6, wherein the training data comprises an image data, and wherein the user sensing data comprises an electrical signal data.

8. The method of any one of claims 1-7, wherein the user sensor comprises a piezoelectric sensor.

9. The method of any one of claims 1-8, wherein the computer algorithm is configured to convert the user sensing data to a same data type as the training data, thereby generating a processed user sensing data.

10. The method of claim 9, wherein the machine learning model is configured to analyze the processed user sensing data to determine the target condition of the user.

11. The method of claim 9, wherein the processed user sensing data comprises a single image data comprising a plurality of channels comprising (i) a first channel indicative of a first user sensing data from a first user sensor of the article of furniture and (ii) asecond channel indicative of a second user sensing data from a second user sensor of the article of furniture.

12. The method of claim 11, wherein the first user sensing data and the second user sensing data are associated with different users of the article of furniture.

13. The method of any one of claims 1-12, wherein the target condition comprises a sleep disorder selected from the group consisting of insomnia, hypersomnia, sleep apnea, circadian rhythm sleep disorder, restless legs syndrome, and sleep deficiency.

14. The method of any one of claims 1-13, wherein the target condition comprises snoring.

15. The method of any one of claims 1-14, wherein the training data comprises a data associated with one or more individuals.

16. The method of claim 15, wherein the data associated with one or more individuals does not comprise the user.

17. The method of any one of claims 1-16, wherein the training data is derived from clinical data.

18. The method of any one of claims 1-17, wherein the condition of the article of furniture comprises a movement of at least a portion of the article of furniture, wherein the article of furniture comprises at least one actuator configured to induce the movement.

19. The method of claim 18, wherein the instruction is configured to direct at least one motor to induce the movement.

20. The method of claim 18 or 19, wherein the movement comprises a vibration.

21. The method of claim 18 or 19, wherein the movement comprises an angular and / or height adjustment of the portion of the article of furniture.

22. The method of any one of claims 18-21, wherein the actuator is coupled to the article of furniture.

23. The method of any one of claims 1-22, wherein the condition of the article of furniture comprises a temperature of at least a portion of the article of furniture.

24. The method of any one of claims 1-22, wherein the article of furniture is a bed device.

25. The method of any one of claims 1-22, wherein the article of furniture is a mattress or mattress cover.

26. The method of any one of claims 1-25, wherein the user sensor is part of the article of furniture.

27. The method of any one of claims 1-26, further comprising (i) determining, using the machine learning model, a change in the target condition of the user based on an additional user sensing data; and (ii) generating an additional instruction for the article of furniture to change the condition of the article of furniture based on the change in the target condition of the user.

28. The method of claim 27, wherein the determining comprises comparing at least a portion of the user sensing data to at least a portion of the additional user sensing data.

29. The method of claim 28, wherein the comparing comprises identifying a decreased degree of the target condition of the user.

30. The method of claim 27 or 28, wherein the additional user sensing data is generated by the user sensor.

31. The method of claim 27 or 28, wherein the additional user sensing data is generated by a second user sensor.

32. A system for controlling a condition of an article of furniture, the system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the processor, implements the method of any one of claims 1-31.

33. A computer-implemented method for controlling a condition of an article of furniture, the method comprising:(a) determining a sleep disorder of a user of the article of furniture when a threshold indicator of a user sensing data is identified, wherein the user sensing data is generated by a user sensor while the user is using the article of furniture, and wherein the threshold indicator is based on at least one user factor; and(b) generating an instruction for the article of furniture to change the condition of the article of furniture based on the determined sleep disorder of the user.

34. The method of claim 33, wherein the determining is based on a machine learning model trained to identify the threshold indicator from the user sensing data.

35. The method of claim 33, wherein the determining is performed in absence of a machine learning model.

36. The method of claim 33, wherein the determining is based on a time-frequency analysis.

37. The method of claim 36, wherein the time-frequency analysis comprises Fourier transformation, Laplace transformation, or Wavelet transformation.

38. The method of claim 36, wherein the time-frequency analysis comprises Fourier transformation, wherein the Fourier transformation comprise one or more members selected from the group consisting of a Discrete Fourier Transformation (DFT), a Fast Fourier Transformation (FFT), a Principal Component Analysis (PCA) transformation, an Inverse Discrete Fourier Transformation, and an Inverse Fast Fourier Transformation.

39. The method of any one of claims 33-38, wherein the at least one user factor comprises a plurality of users factors.

40. The method of any one of claims 33-39, wherein the at least one user factor comprises one or more members selected from the group consisting of biological sex, age, weight, health condition, exercise information, and diet information (or food consumption information).

41. The method of any one of claims 33-40, wherein the sleep disorder comprises snoring.

42. The method of any one of claims 33-41, wherein the determining and the generating are performed over a same use of the article of furniture.

43. The method of any one of claims 33-42, wherein the determining and the generating are performed substantially in real-time.

44. The method of any one of claims 33-43, wherein the condition of the article of furniture comprises a movement of at least a portion of the article of furniture, wherein the article of furniture comprises at least one actuator configured to induce the movement.

45. The method of claim 44, wherein the movement comprises a vibration, an angular adjustment, a height adjustment, or a combination thereof.

46. The method of any one of claims 33-45, wherein the article of furniture is a bed device.

47. The method of any one of claims 33-45, wherein the article of furniture is a bed mattress or a mattress cover.

48. The method of any one of claims 33-47, wherein the user sensor is part of the article of furniture.

49. The method of any one of claims 33-48, further comprising (i) determining a change in the sleep disorder of the user based on an additional user sensing data; and (ii) generating an additional instruction for the article of furniture to change the condition of the article of furniture based on the change in the sleep disorder of the user.

50. The method of claim 49, wherein the comparing comprises identifying a decreased degree of the sleep disorder of the user.

51. A system for controlling a condition of an article of furniture, the system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the processor, implements the method of any one of claims 33-50.

52. A computer-implemented method for controlling a condition of an article of furniture of a user, the method comprising:(a) analyzing a quality of a target sleep phase of the user from a prior sleep session of the user; and(b) automatically modifying a predetermined adjustment profile of the condition of the article of furniture based on the analyzed quality of the target sleep phase, to generate a modified adjustment profile of the condition of the article of furniture, wherein the automatically modifying comprises changing a target value of the condition that is associated with a target sleep phase of the user, and wherein the modified adjustment profile is executable by the article of furniture to regulate the condition of the article of furniture.

53. The method of claim 52, wherein the article of furniture comprises a regulating unit configured to regulate the condition of the article of furniture based on the modified adjustment profile.

54. The method of claim 52 or 53, further comprising storing, in a database, the modified adjustment profile for use by the article of furniture.

55. The method of any one of claims 52-54, wherein the prior sleep session of the user was on the article of furniture operated based on the predetermined adjustment profile.

56. The method of claim 52, wherein the target sleep phase is selected from the group consisting of REM sleep, deep sleep, and light sleep.

57. The method of claim 52 or 53, wherein the analyzing comprises determining that a duration of the target sleep phase is less than a target duration of the target sleep phase.

58. The method of any one of claims 52-57, wherein the prior sleep session is an immediately preceding sleep session.

59. The method of any one of claims 52-58, wherein the predetermined adjustment profile was utilized to regulate the condition of the article of furniture during the prior sleep session.

60. The method of any one of claims 52-59, wherein the analyzing is based on a sensor data associated with the user collected by a user sensor during the prior sleep session.

61. The method of claim 60, wherein the user sensor is part of the article of furniture.

62. The method of any one of claims 52-61, wherein the condition is a temperature of the article of furniture, and wherein the regulating unit is a temperature regulating unit.

63. The method of claim 62, wherein the temperature regulating unit is coupled to the article of furniture.

64. The method of any one of claims 52-63, further comprising sending at least a portion of the modified adjustment profile to the regulating unit.

65. The method of any one of claims 52-64, further comprising repeating (a) and (b) a plurality of times for a plurality of different target sleep phases of the user.

66. The method of any one of claims 52-65, further comprising performing (a) and (b) prior to or subsequent to determining a presence of the user on the article of furniture.

67. The method of any one of claims 52-66, wherein the article of furniture is a bed device selected from the group consisting of a mattress, a pillow, and a cover thereof.

68. The method of any one of claims 52-67, wherein the prior sleep session comprises a single sleep session, and wherein the method further comprises analyzing the quality of the target sleep phase from the entirety of the sleep session.

69. The method of any one of claims 52-67, wherein the prior sleep session comprises a single sleep session, and wherein the method further comprises analyzing the quality of the target sleep phase from a fraction and not the entirety of the single sleep session.

70. The method of claim 69, wherein the fraction comprises a portion of a first half of the single sleep session.

71. The method of claim 69, wherein a duration of the fraction is less than 80% of a duration of the singles sleep session.

72. The method of any one of claims 52-71, wherein the predetermined adjustment profile comprises a plurality of sleep cycles, each sleep cycle comprising a same target sleep phase, and wherein the target value of the condition associated with the same target sleep phase is automatically modified to a same value for different sleep cycles.

73. The method of any one of claims 52-71, wherein the predetermined adjustment profile comprises a plurality of sleep cycles, each sleep cycle comprising the same target sleep phase, and wherein the target value of the condition associated with the same target sleep phase is automatically modified differently for different sleep cycles.

74. The method of any one of claims 52-71, wherein the predetermined adjustment profile comprises different types of the target sleep phase, and wherein the target value is automatically modified to a same value regardless of the different types of the target sleep phase.

75. The method of any one of claims 52-71, wherein the predetermined adjustment profile comprises different types of the target sleep phase, and wherein the target value is automatically modified differently depending on the different types of the target sleep phase.

76. The method of any one of claims 52-75, wherein (1) the target value is automatically modified if the quality of the target sleep phase is below a predetermined threshold and (2) the target value is not automatically modified if the quality of the target sleep phase is at or above the predetermined threshold.

77. The method of any one of claims 52-75, wherein the target value is automatically modified differently depending on whether (1) the quality of the target sleep phase is below a predetermined threshold or (2) the quality of the target sleep phase is at or above the predetermined threshold.

78. The method of claim 77, wherein the target value is automatically modified by different degrees depending on whether (1) or (2) is met.

79. The method of claim 78, wherein the target value is automatically modified by a greater degree when (1) is met as compared to when (2) is met.

80. The method of any one of claims 52-79, further comprising automatically modifying the target value based on a combination of a plurality of factors selected from the group consisting of the analysis of the quality of the target sleep phase, age of the user, biological sex of the user, geolocation of the user, and health condition of the user, wherein different combinations of the plurality of factors correspond to different modifications of the target value.

81. The method of claim 80, wherein the plurality of factors comprises three or more factors.

82. The method of claim 81, wherein the plurality of factors comprises (i) the quality of the target sleep phase and (ii) how the age of the user compares relative to a predetermined threshold age or range thereof.

83. The method of claim 82, wherein the predetermined threshold age is greater than or equal to about 30.

84. The method of claim 82, wherein the predetermined threshold age is less than or equal to about 60.

85. The method of claim 80, wherein the plurality of factors comprises (i) the quality of the target sleep phase and (ii) whether the biological sex is male or female.

86. A system comprising a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the one or more computer processors, implements the method of any one of the claims 52-85.

87. A computer-implemented method for controlling a condition of an article of furniture of a user, the method comprising:(a) receiving a user input comprising at least one user event; and(b) automatically modifying a predetermined adjustment profile of the condition of the article of furniture based on the at least one user event, to generate a modified adjustment profile of the condition of the article of furniture, wherein the automatically modifying comprises changing a target duration for subjecting at least a portion of the article of furniture to a target value of the condition, and wherein the modified adjustment profile is executable by the article of furniture to regulate the condition of the article of furniture.

88. The method of claim 87, wherein the at least one user event is not collected during a prior sleep session of the user.

89. The method of claim 87 or 88, wherein the at least one user event comprises exercise information, food consumption information, travel information, or any combination thereof.

90. The method of any one of claims 87-89, wherein the at least one user event comprises a past event.

91. The method of any one of claims 87-89, wherein the at least one user event comprises a future event.

92. The method of claim 91, wherein the future event comprises a nap event.

93. The method of any one of claims 87-92, wherein the condition is a temperature of the article of furniture.

94. The method of any one of claims 87-92, wherein the condition is a condition of an environment of the article of furniture.

95. The method of claim 94, wherein the condition of the environment is a humidity or a temperature.

96. The method of any one of claims 87-95, further comprising monitoring a quality of sleep of the user during or subsequent to executing the modified adjustment profile by the article of furniture or the regulating unit thereof.

97. The method of claim 96, wherein the monitoring comprises analyzing one or more biological signals of the user.

98. The method of claim 97, wherein the one or more biological signals are generated by a user sensor during a portion of use of the article of furniture by the user.

99. The method of claim 98, wherein the user sensor is part of the article of furniture.

100. The method of any one of claims 87-99, further comprising: (i) directing the article of furniture or the regulating unit thereof to execute at least a portion of the modified adjustment profile for a usage of the article of furniture by the user; (ii) monitoring a quality of the user’s usage; and (iii) modifying at least a portion of the modified adjustment profile based on the monitored quality, to generate a different adjustment profile that is usable for (1) a subsequent usage of the article of furniture by the user, (2) a remainder portion of the usage, or (3) both.

101. The method of any one of claims 87-100, wherein the user input is provided by the user via a graphical user interface (GUI) of a user application provided on a user device.

102. The method of any one of claims 87-101, wherein the user input is retrieved from the database or an additional database associated with the user.

103. The method of any one of claims 87-102, wherein the predetermined adjustment profile is specifically generated for each individual user.

104. The method of any one of claims 87-103, wherein the article of furniture is a bed device.

105. The method of any one of claims 87-104, wherein the article of furniture is a mattress or a mattress cover.

106. A computer computer-implemented method for controlling a condition of an article of furniture, the method comprising:changing a condition of the article of furniture as an alarm to wake up a user of the article of furniture, wherein the changing is based on (i) a target sleep duration of the user and (ii) a time when the user falls asleep on the article of furniture or when the user begins the current use of the article of furniture.

107. The method of claim 106, wherein the changing is based on the time when the user falls asleep on the article of furniture.

108. The method of claim 106 or 107, wherein the changing is based the time when the user begins the current use of the article of furniture.

109. The method of any one of claims 106-108, further comprising determining the time based on a user sensing data generated by a user sensor, optionally wherein the user sensor is a part of the article of furniture.

110. The method of any one of claims 106-109, wherein the changing is performed upon determining that the user has reached or is expected to reach the target sleep duration during the current use.

111. The method of claim 110, further comprising assessing or tracking an actual sleep duration of the user during the current use, wherein an onset of the actual sleep duration of the user is determined based on the time, optionally wherein the assessing or tracking is performed a plurality of times throughout the current use.

112. The method of any one of claims 106-111, wherein the condition of the article of furniture comprises a movement or vibration of at least a portion of the article of furniture.

113. The method of any one of claims 106-112, wherein the condition of the article of furniture comprises a temperature of at least a portion of the article of furniture.

114. The method of any one of claims 106-113, further comprising retrieving the target sleep duration from a database associated with the user.

115. The method of any one of claims 106-114, further comprising receiving the target sleep duration from the user via a graphical user interface (GUI) operatively coupled to the article of furniture.

116. The method of any one of claims 106-115, further comprising turning off the changing or the alarm based on determining that the user is not present on the article of furniture, a current time is after the user’s predetermined alarm time, or both.A system comprising: a computer processor and a computer memory coupled thereto, wherein the computer memory comprises a machine executable code that, upon execution by the one or more computer processors, implements the method of any one of claims 87-116.-Wi

Images