Work and rest behavior intervention method and implementation device thereof

Abstract

The invention relates to a method for intervening work and rest behaviors, which specifically realizes behavior transformation through a multi-dimensional state perception and mandatory sound intervention mechanism, and comprises a state detection module, a light detection module, a posture detection module, a time judgment module, a main control unit module and a sound output and information display module. The device judges and recognizes whether the user stays up late, stays in bed or evades or not through a multi-dimensional state at the time points of falling asleep and getting up set by the user, and triggers inevitable sound intervention if behavior requirements are not met. Intervention removal needs to meet a specific state or a limited cancel condition, and the device operates periodically to assist a user to establish regular work and rest habits.

Description

I. TECHNICAL FIELD

[0001] The present application provides an intelligent health management device for sleep-wake behavior intervention, aiming to identify whether the user has the behavior of staying up late, sleeping in or avoiding through multi-dimensional state perception and mandatory sound intervention mechanism at the user's set sleep and wake-up time points, and to promote the user to complete the behavior change through technical means, so as to establish regular sleep-wake habits. II. BACKGROUND

[0002] The existing sleep-wake management methods mainly rely on mobile phone applications, smart bracelets or smart light control devices, etc. Such products usually remind the user to sleep or wake up through timing reminders, vibration prompts or light changes. However, these methods generally have the following shortcomings:

[0003] 1. Dependence on user self-consciousness

[0004] Mobile phone alarms or App reminders can be easily turned off or ignored by the user, lack of mandatory, and are difficult to truly change the habit of staying up late or sleeping in.

[0005] 2. Lack of state recognition

[0006] Most existing solutions are only based on time-triggered reminders and cannot determine whether the user has really gone to bed, is still using electronic products, or has left the bed, so the reminder effect is limited.

[0007] 3. Easy to avoid

[0008] Users can bypass the reminder by moving the device, blocking the sensor, turning off the power, etc., resulting in ineffective intervention and inability to form long-term regularity.

[0009] 4. Lack of periodic constraints

[0010] Existing products usually provide one-time reminders, and the user can cancel or modify the settings at any time, lack of periodic execution and cancellation time limit, and are difficult to form continuous behavior correction.

[0011] 5. Intervention effect is weakened

[0012] Most existing reminder methods are mild prompts with limited sound or vibration intensity, and usually allow the physical button to be turned off, so the user is prone to develop dependence or inertia, and cannot achieve the effect of mandatory intervention.

[0013] 6. Privacy and complexity issues

[0014] Some smart wearable devices rely on heart rate, sleep monitoring and other physiological data, involving privacy collection, and the device is complex to wear, with limited user acceptance.

[0015] Based on the above shortcomings, the prior art is difficult to effectively constrain user behavior at key time points, cannot prevent the case of "people in bed but not sleeping" or "alarm clock ringing but continuing to sleep", and thus is difficult to help users truly establish regular work and rest habits. IV. SUMMARY

[0016] To overcome the above problems, the present application proposes an intelligent health management device for work and rest behavior intervention, aiming to guide users to complete the sleep and get-up behavior at the set time point through multi-dimensional state perception and mandatory sound intervention mechanism, so as to establish regular work and rest habits.

[0017] The device comprises a state detection module, an ambient light detection module, a posture detection module, a time judgment module, a main control unit, a sound output module and an information display module.

[0018] The system judges whether the user continues to use electronic products in bed at the sleep time point through "whether in bed + light state", and prevents the user from continuing to use electronic products in bed; the system judges whether the user continues to lie in bed or avoids getting up at the get-up time point through "whether getting off the bed + posture stability", and prevents the user from continuing to lie in bed or avoiding getting up. If the conditions are not met, the sound intervention is triggered, and the sound cannot be turned off through the physical button, and is only terminated under the conditions of meeting the cancellation conditions or a limited number of manual cancellations.

[0019] The device operates in cycles, and each day contains two intervention links of sleep and getting up, the number of cancellations is limited, and the cycle can be renewed after the end of the cycle. The mechanism effectively helps users gradually form regular work and rest habits by limiting avoidance behavior and periodic execution.

[0020] V. ADVANTAGEOUS EFFECTS

[0021] The present application can effectively constrain user behavior at key time points through multi-dimensional state perception and mandatory intervention mechanism, and gradually help users form regular work and rest habits, and has the following remarkable effects:

[0022] ● Establishment of regular work and rest The device automatically triggers intervention at the sleep and get-up time points set by the user, avoids disrupting the work and rest rhythm due to staying up late or lying in bed, and continuous use can help users gradually form fixed sleep and get-up rules.

[0023] ● Prevention of procrastination in bed Through the dual judgment of "in bed state + light state", the device can identify the case that the user has already lain down but is still using electronic products, prevent the procrastination behavior of "people in bed but not sleeping", and ensure that the sleep time is not delayed.

[0024] ● Prevention of lying in bed and avoiding the get-up section The combination of state detection and posture detection avoids the user's escape from intervention by continuing to lie in bed or moving the device, and ensures that the user really gets up and leaves the bed at the set time point.

[0025] ● Behavior guidance natural intervention condition release requires users to actively adjust the environment and behavior, such as turning off the lighting, putting down electronic products, entering the bed or leaving the bed. In the process of meeting the conditions, the user naturally completes the transition of the work and rest behavior.

[0026] ● Periodic constraint helps habit formation device to run by period, daily contains two intervention links of falling asleep and getting up, and the number of manual cancellation is limited. The periodic execution and limitation mechanism can effectively prevent the user from relying on the cancellation function, thereby strengthening the formation of regular work and rest.

[0027] ● Simple technology and privacy protection device only detects "in bed / not in bed" and environmental light state, and does not involve specific user physical information; posture detection only monitors the state of the device itself,

[0028] Avoid privacy leakage. The overall structure is simple, which is convenient for popularization and application. Six, ORNAMENTAL DESCRIPTION

[0029] Figure 1 : The structure diagram of the device of the application shows the connection relationship between the state detection module, the environmental light detection module, the posture detection module, the time judgment module, the main control unit, the sound output module and the information display module.

[0030] Figure 2 : Control logic flow chart, showing the triggering condition, judgment process and intervention release mode of the sleep intervention and getting up intervention. Seven, PREFERRED EMBODIMENT

[0031] System overall structure:

[0032] ● State detection module: microwave radar, used for "in bed / not in bed" binary detection of bed area and output of binary signal.

[0033] ● Environmental light detection module: photosensitive sensor, used for collecting environmental light intensity.

[0034] ● Posture detection module: three-axis gyroscope and three-axis accelerometer, used for detecting that the device is rotated or instantaneous acceleration event occurs.

[0035] ● Time judgment module: system real-time clock and timing trigger logic.

[0036] ● Main control unit (MCU): responsible for state machine, logic judgment, parameter reading and writing, event recording and man-machine interaction control.

[0037] ● Sound output module: loudspeaker or buzzer, volume range 70-100dB (factory setting), used for outputting intervention sound.

[0038] • Information display module: screen and optional voice announcement unit for displaying parameters, events and status information and providing voice prompts.

[0039] • Non-volatile memory (NVM): holds factory parameters, posture data, cycle settings, cancellation records, dynamic light reference and event logs.

[0040] • Power supply: built-in rechargeable battery, external charging supported; the device has no physical power-off button or equivalent interface that can be directly cut off by the user.

[0041] Parameter strategy and write protection

[0042] • Key parameters that are fixed in NVM at factory and write-protected during the cycle include but are not limited to:

[0043] o Sound volume range: 70-100 dB;

[0044] o Posture event threshold: rotation angle threshold 30°; acceleration threshold 0.5g;

[0045] o Light darkness initial threshold: factory slightly high value, e.g. 30-50 lux;

[0046] o Minimum allowed value for execution cycle: 7 days;

[0047] o Structural protection parameters: sound output area protrusion size and layout.

[0048] • The user can set the sleep time point, wake-up time point, cycle length (≥7 days) before the cycle starts, but once the cycle starts, the above parameters will be frozen until the end of the cycle. Only the volume level can still be adjusted within the specified range.

[0049] • The number of cancellations is automatically allocated by the system according to the cycle length (e.g. 2 times per 7 days), written into NVM by the MCU and counted, which cannot be modified or reset by the user.

[0050] Cycle initialization and posture reference recording

[0051] • Reference collection trigger: at the start of the cycle, the MCU prompts the user to place the device in a suitable position (e.g. fixed stand or bedside plane) and perform "posture reference collection" (e.g. 60-120 seconds).

[0052] • Sampling and calculation:

[0053] o Continuously sample three-axis gyroscope and three-axis accelerometer data, perform basic denoising and stability verification (angular velocity close to zero, acceleration vector stable).

[0054] o Calculate and fix the following reference data:

[0055] ■Attitude reference: Quaternions or Euler angles, representing the device's rest orientation;

[0056] ■Acceleration reference vector: Contains gravity component, used for subsequent attitude calculation and deviation judgment;

[0057] ■Noise statistics parameters: Short-term variance or standard deviation of angular velocity / acceleration on each axis, as baseline for anti-fluctuation threshold.

[0058] ●Write and freeze: Reference and statistics parameters are written to NVM and frozen with the cycle; throughout the cycle, the device's "movement / rotation" determination will be referenced to this reference, combined with the threshold

[0059] and time window for judgment.

[0060] Sleep intervention process and dynamic light calibration:

[0061] 1. Trigger: The system clock reaches the user-set sleep time point, and the MCU enters the sleep judgment process.

[0062] 2. Preliminary judgment:

[0063] ○Microwave radar returns "in bed / not in bed" binary signal;

[0064] ○Photosensitive sensor collects current light and compares it with the current effective darkness threshold (preferably using the calibrated on-site darkness reference, if not, using the factory initial threshold).

[0065] 3. Sound trigger: If the radar determines "not in bed" or the light is stronger than the effective threshold, the MCU drives the sound output module to output intervention sound at the set gear (70-100 dB); the sound does not respond to any physical key-off request.

[0066] 4. Intervention maintenance and release:

[0067] ○Automatic release: When the radar returns "in bed" and the light is below the current effective threshold at the same time, the MCU stops sound output;

[0068] ○Manual cancellation: If the remaining cancellation times in this cycle are >0, the user can initiate a cancellation, and the MCU records and deducts the remaining times; cancellation is a one-time, non-retractable record operation, only terminating the sound output of this sleep session.

[0069] 5. Dynamic light calibration (only automatically executed after sleep confirmation):

[0070] ○Sleep confirmation: The radar continuously returns "in bed" and the micro-motion feature meets the factory-set stability standard within the predetermined stable time window (e.g., 5-10 consecutive minutes), and the MCU determines that the user has fallen asleep;

[0071] ○ Data Acquisition and Consolidation: Within a stable window (e.g., 2–3 minutes) after sleep confirmation, sample the photosensor and take the average value as the "on-site darkness baseline"; form a new effective darkness threshold with "baseline + factory margin" (e.g., +5–10 lux) and write it into the NVM for priority use in subsequent cycles; if the acquisition fails, retain the original factory threshold; this calibration process is only automatically triggered upon sleep confirmation and cannot be manually triggered or changed by the user.

[0072] Wake-up intervention process and postural event determination:

[0073] 1. Trigger: When the system clock reaches the user-defined wake-up time, the MCU enters the wake-up judgment process.

[0074] 2. Judgment Item:

[0075] ○ The microwave radar returns a binary signal indicating "still in bed / already out of bed";

[0076] The attitude detection module compares the real-time attitude with the "periodic attitude reference" and determines whether a rotation or instantaneous acceleration event has occurred based on a preset threshold and time window.

[0077] (For example, rotation ≥30° or instantaneous acceleration ≥0.5g is valid within the window).

[0078] 3. Event trigger determination (based on baseline and time window):

[0079] ○ When the deviation relative to the attitude reference reaches a preset threshold and meets the continuous or percentage conditions within the set time window, it is determined to be an "avoidance of related events". The MCU writes the event record to the NVM and displays the peak value and timestamp on the display module.

[0080] If an avoidance-related event occurs and no sound has been emitted yet, the MCU will immediately trigger or increase the sound output; if sound has already been emitted, the intervention will be maintained or intensified until the condition is lifted or canceled.

[0081] 4. Intervention termination:

[0082] ○Automatic release: When the radar returns to "off the bed" and the device is in a stable attitude state (returning to the allowable static attitude range relative to the reference, and no over-threshold event occurs again within the set time window), the sound output stops;

[0083] ○ Manual Cancellation: One cancellation is allowed and recorded if the number of cancellation attempts has not been exhausted; cancellation is a one-time operation and cannot be undone.

[0084] Attitude event details and display:

[0085] ●Thresholds and Events: Rotational deviation threshold of 30° (any axis) relative to the "Periodic Attitude Reference", acceleration deviation threshold of 0.5g (instantaneous peak); thresholds are written to NVM at the factory and are protected. Events are logged after they are determined according to the threshold and time window rules.

[0086] ●Event Log: Each event includes information such as event type, peak value (angle or g value relative to the baseline), timestamp, occurrence stage (falling asleep / waking up), and remaining cancellation count at that time, which is written to NVM for auditing purposes.

[0087] ●Display and prompts:

[0088] When an event occurs, the information display module displays the event peak (relative to the baseline), timestamp, and current remaining cancellation count in real time, and provides a voice prompt: "Rotation / movement detected, determined to be an avoidance related event, intervention continues."

[0089] ○ The main interface or period summary page displays the cumulative number of events in this period, the peak value of each event, and the timestamp;

[0090] The screen can display the current attitude value and its deviation from the reference in real time, or indicate the attitude stability / instability status in a graphical way, so that users can confirm whether the device has returned to the allowable static attitude.

[0091] Overall state machine process and exception handling:

[0092] ●State machine diagram: Standby state → Sleep judgment state → Sleep intervention state → (Sleep confirmation → Dynamic calibration) → Standby state → Wake-up judgment state → Wake-up intervention state → Standby state.

[0093] ●Sensor malfunction:

[0094] ○ When the photosensitive or radar malfunctions, the system enters conservative mode: it continues to use the factory initial threshold and the radar binary conclusion to perform intervention, and the attitude module still makes judgments based on the fixed attitude reference; at the same time, the display module prompts the sensor fault and records the log.

[0095] ○ When the attitude module malfunctions, prioritize intervention and issue a fault warning, and record the fault log; if necessary, prompt for repair at the end of the cycle or during equipment maintenance.

[0096] ●Power and Power Off: NVM saves key parameters, cancels records, dynamic ray references, attitude references and event logs without loss when power is off; after restarting, the records are restored to the appropriate state.

[0097] Power behavior and continuous sound emission rules:

[0098] ●Power Supply: Built-in rechargeable battery with support for external charging. The device lacks a user-operable power-off button or equivalent interface to cut off the power supply; therefore, users cannot directly shut down the device using a button. The device operates continuously when powered (internal or external). It stops operating due to power failure when the battery is depleted or the external power supply is disconnected and the battery has no remaining charge.

[0099] ● Continuous Sound Output Rule: Once the trigger condition is met, sound output will continue until any of the following conditions occur and are determined by the MCU:

[0100] ○ The corresponding automatic release condition is met and confirmed; or

[0101] ○ The user initiates a manual cancellation within the current period and has not yet exhausted their cancellation attempts; or

[0102] ○ The device is powered off (the battery is depleted or the external power supply is disconnected and the battery has no remaining charge).

[0103] ●Recovery after power failure: If sound stops due to a power failure, the device will resume its recovery status according to the NVM record after power is restored; if the triggering conditions are still met, the device will output sound again after power is restored and self-test and judgment are completed.

[0104] ●Safety Control: Peak sound level is capped at 100dB; to reduce exposure to continuous high-intensity sound, the duty cycle or intermittent playback strategy can be preset at the factory. Specific duty cycle parameters are written at the factory.

[0105] It is written to NVM and protected from writes.

[0106] Unavoidable implementation details and privacy protection:

[0107] ●Unavoidability: Factory-fixed thresholds and parameters, write protection within the cycle, automatic allocation of cancellation counts by the system according to the cycle, structural protrusions to prevent physical obstruction, physical buttons not having the function of cutting off sound, and "cycle start attitude reference recording + relative reference time window judgment" together ensure that intervention is difficult for users to avoid through normal operation.

[0108] ● Privacy protection: The microwave radar only outputs a binary signal indicating "in bed / out of bed" and necessary micro-motion stability indicators; the system does not store or report heart rate, respiration, images or other personal physiological information; illumination data is only used for environmental determination and threshold calibration and is not transmitted as personal identification information.

[0109] Periodic logging, auditing, and user interface

[0110] ●The cycle summary interface displays: cycle start and end dates, number of canceled / remaining times, event log for this cycle (time, type, peak relative to the baseline), fixed field darkness baseline value, attitude baseline establishment timestamp, and battery power status.

[0111] ● Log Export: The device can export event summaries for auditing under controlled physical or authorized interfaces. The exported content does not include personal physiological data. The export function is limited and is explicitly authorized and recorded on the device.

Claims

1. A method for intervening in daily routines, characterized in that, Includes the following steps: • At the user's set bedtime, collect data on whether the user is in bed and the ambient light intensity; • When the detection results indicate that the user is not in bed or the ambient light is strong, the system enters a sleep intervention state and triggers sound output; • At the user-defined wake-up time, collect data on whether the user gets out of bed and whether the device's posture exceeds a preset threshold range; • When the detection result indicates that the user is still in bed or the device posture exceeds the preset threshold range, the system will enter the wake-up intervention state and trigger sound output; • In the intervention state, the sound output remains active and does not respond to the physical button to turn it off; The sound output will only terminate if any of the following conditions are met: The system detects either "in bed and in dim light" or "out of bed and in a position within a preset threshold range"; The user initiated a manual cancellation within the current period and has not exhausted the cancellation attempts; Equipment power failure; The method is executed according to a preset cycle, each cycle includes two intervention stages: falling asleep and waking up, and the number of cancellations is automatically allocated and limited according to the cycle; • During the period, key parameters are frozen and cannot be modified or reset by the user.

2. The method according to claim 1, characterized in that, After the user confirms that they are asleep, dynamic light calibration is automatically performed, and the average light value within a stable window is collected as the on-site darkness benchmark and used preferentially in subsequent cycles.

3. The method according to claim 1 or 2, characterized in that, The attitude detection includes acquiring and freezing the attitude reference at the beginning of the cycle and freezing it during the cycle. When the rotation or acceleration relative to the reference exceeds a preset threshold and meets the time window condition, it is determined to be an avoidance behavior.

4. The method according to any one of claims 1 to 3, characterized in that, The status detection module is a microwave radar, used to output a binary signal indicating "in bed / out of bed".

5. The method according to any one of claims 1 to 4, characterized in that, The ambient light detection module is a photosensitive sensor used to collect ambient light intensity and compare it with a preset or dynamically calibrated threshold.

6. The method according to any one of claims 1 to 5, characterized in that, The attitude detection module includes a three-axis gyroscope and a three-axis accelerometer, used to detect the rotation angle and instantaneous acceleration of the device relative to a periodic attitude reference.

7. The method according to any one of claims 1 to 6, characterized in that, The sound output module is a speaker or buzzer with a volume range of 70-100dB and does not respond to the power-off request of the physical button.

8. The method according to any one of claims 1 to 7, characterized in that, The main control unit is a microcontroller (MCU), used to execute state machine logic, parameter freezing, cancellation count, and event logging.

9. The method according to any one of claims 1 to 8, characterized in that, The external structure of the sound output module has a protrusion to prevent the sound output area from being blocked, thereby ensuring the effective propagation of the interference sound.

Images