Ai posture correction system for customized spinal care and method using same

Abstract

The present invention relates to an AI posture correction system for customized spinal care and may comprise: a smart seat (100) including a frame unit (110) configured to, when pressure is received from a body, continuously apply upward and inward compression forces for active stabilization support, a cushion unit (120) including a cushioning member such that compression resistance is generated when in direct contact with the body, a sensor unit (130) positioned between the cushion unit (120) and the frame unit (110) and having a plurality of pressure sensors distributed in a pressure region that receives pressure from the body, and a control unit (140) that transmits, to a computing apparatus (200), pressure data sensed by the sensor unit (130); and a computing apparatus (200) that calculates, on the basis of the pressure data, a first center point that is a center point of pressure, calculates a difference value between the first center point and a second center point that is a center point of the pressure region, recognizes a posture pattern on the basis of the difference value, and outputs an AI-customized operation related to posture correction from the posture pattern.

Description

AI Posture Correction System for Customized Spinal Care and Method Using the Same

[0001] The present invention relates to an AI posture correction system and a method for customized spinal care.

[0002] More specifically, the invention relates to an AI posture correction system and a method thereof, wherein a computing device collects pressure data generated at a part in contact with the user's body while monitoring the user's posture in real time, performs AI analysis based on the collected data and monitoring results, and outputs AI-customized movements related to posture correction.

[0003] In modern society, many people spend long hours sitting while living or working. If the sitting posture is incorrect, prolonged sitting can have a negative impact on the spine and musculoskeletal system, potentially leading to various health problems such as back pain, scoliosis, and cervical disc herniation. To prevent or alleviate these issues, many researchers and companies have been conducting studies on posture correction devices.

[0004] Traditional posture correction devices have been provided in forms that attach to the user's body or force them to maintain a specific posture. However, these devices have disadvantages, such as being uncomfortable to wear or potentially restricting the user's daily life. Furthermore, some mechanical devices restrict user movement, which can cause long-term discomfort, and are often difficult for users to wear continuously.

[0005] Therefore, the present invention aims to solve the problems of existing technology and propose a smart seat and a method that provide a more sophisticated and effective posture correction function.

[0006] Domestic Patent Publication 10-2011-0117193, which serves as the basis, discloses "a method and apparatus for mechanically correcting posture."

[0007] The present invention aims to propose a method for monitoring a user's posture in real time through a smart seat and correcting the posture based on pressure data. Through this, the user can correct their posture in a more convenient and effective manner, and the main challenge to be solved is to minimize the inconveniences associated with existing technologies.

[0008] An AI posture correction system for customized spinal care to solve the aforementioned problem may include: a cushion part equipped with a cushioning member to generate compression resistance when in direct contact with the body; a frame part configured to continuously apply upward and inward compressive force for active stabilization support when receiving pressure from the body; a sensor part located between the cushion part and the frame part, in which a plurality of pressure sensors are distributed in a pressure area receiving pressure from the body; and a control part that transmits pressure data sensed by the sensor part to a computing device; and an AI posture correction system including a computing device that calculates a first center point, which is the center point of pressure, based on the pressure data, calculates a difference value between the first center point and a second center point, which is the center point of the pressure area, recognizes a posture pattern based on the difference value, and outputs an AI customized movement related to posture correction from the posture pattern.

[0009] The effects that may occur when using the embodiments of the present invention are listed.

[0010] An AI posture correction system according to one embodiment of the present invention can collect pressure data generated in real time from a part in contact with the user's body and perform posture correction movements based on this data. The smart seat presented in claim 1 has the characteristic of being able to output movements necessary for posture correction by calculating the difference between the center point of pressure and the center point of the pressure area. Through this, the user can quickly recognize and correct postural imbalances and receive assistance in maintaining a correct posture while sitting for a long time.

[0011] An AI posture correction system according to one embodiment of the present invention can measure pressure distribution more precisely by lowering the sensitivity of a sensor located in the central region within the pressure area. This enables more accurate analysis of pressure data generated at the center and provides the advantage of more detailed monitoring of the user's posture state. Consequently, by collecting accurate pressure data for each part of the user's body, more effective posture correction is possible.

[0012] An AI posture correction system according to one embodiment of the present invention can enhance the efficiency of posture correction by differentiating the sensitivity of pressure sensors located in the buttocks and thigh regions. This method is advantageous for preventing deformation of the spine or pelvis by finely distinguishing and analyzing the pressure applied to different parts of the user's body. Through this, excessive burden applied to specific areas during posture correction can be reduced, and it helps maintain the health of the musculoskeletal system in the long term.

[0013] An AI posture correction system according to one embodiment of the present invention activates a vibration motor based on a pressure difference to provide a warning signal when the user's posture is incorrect. Through this, the user receives immediate feedback and can quickly correct an incorrect posture if they unwittingly adopt it. This real-time feedback function can be of great help to the user's continuous posture correction.

[0014] An AI posture correction system according to one embodiment of the present invention includes a function that utilizes a plurality of vibration motors to operate the motor closest to the user's center of pressure. Through this, the user can receive more specific and accurate posture correction signals and receive direct feedback regarding the areas where postural imbalances occur. As a result, the user can monitor and correct their posture status more efficiently.

[0015] An AI posture correction system according to one embodiment of the present invention enhances the structural flexibility of a smart seat and enables efficient placement of a pressure sensor by forming a groove in the film portion of the sensor portion that is symmetrical to the curved shape of the frame portion. This allows for accurate pressure measurement tailored to various body shapes and maintains comfort even during prolonged use.

[0016] An AI posture correction system according to one embodiment of the present invention is configured in a symmetrical and separated form, thereby providing the advantage of being able to measure pressure on various body parts more independently. This structure is effective in detecting and correcting asymmetry in pressure distribution, and allows the user to receive more precise posture correction feedback. Furthermore, due to the symmetrical and separated form, the problem of the sensor being damaged by tension applied to the sensor can be improved.

[0017] FIG. 1 is a diagram illustrating the components of the AI ​​posture correction system of the present invention.

[0018] FIG. 2 is a perspective view for explaining the external appearance of the smart sheet of the present invention.

[0019] FIG. 3 is an exploded perspective view for explaining the sensor part of the smart sheet of the present invention.

[0020] FIG. 4 is a drawing showing a pressure sensor of a sensor unit according to one embodiment of the present invention.

[0021] FIG. 5 is a drawing illustrating examples of a pressure region, a first center point, and a second center point according to an embodiment of the present invention.

[0022] FIG. 6 is a drawing illustrating examples of a buttocks region and a thigh region inside a pressure region according to one embodiment of the present invention.

[0023] FIG. 7 is a drawing illustrating an example for explaining a haptic part according to an embodiment of the present invention.

[0024] FIG. 8 is a drawing illustrating an example for explaining the film portion of a sensor portion according to one embodiment of the present invention.

[0025] FIG. 9 is a drawing illustrating a specific example of a posture correction system that can be implemented in a computing device according to one embodiment of the present invention.

[0026] FIGS. 10 to 15 are drawings illustrating specific examples of a system incorporating IoHT-based AI for spinal care that can be implemented in a computing device according to an embodiment of the present invention.

[0027] In the present invention, various descriptions are provided to facilitate an understanding of the invention. However, it is evident that these embodiments can be practiced without such specific descriptions.

[0028] The term "or" is intended to mean an inclusive "or" rather than an exclusive "or." That is, unless otherwise specified or evident from the context, "X uses A or B" is intended to mean one of the natural inclusive substitutions. In other words, if X uses A; if X uses B; or if X uses both A and B, "X uses A or B" may apply to any of these cases.

[0029] Additionally, the term "and / or" as used in the present invention should be understood to refer to and include all possible combinations of one or more of the listed related items.

[0030] Additionally, the terms "include" and / or "include" shall be understood to mean that the relevant features and / or components are present. However, the terms "include" and / or "include" shall be understood not to exclude the presence or addition of one or more other features, components and / or groups thereof.

[0031] Additionally, unless otherwise specified or the context does not clearly indicate a singular form, the singular in the present invention and claims should generally be interpreted to mean "one or more."

[0032] Finally, the term “at least one of A or B” should be interpreted to mean “a case including only A,” “a case including only B,” and “a case combined with the composition of A and B.”

[0033] The embodiments presented below are provided so that those skilled in the art can easily understand and practice the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the present invention. Accordingly, the present invention is not limited to the specific embodiments presented, but should be interpreted to the maximum extent that is consistent with these principles and novel features.

[0034]

[0035] FIG. 1 is a diagram illustrating the components of the AI ​​posture correction system of the present invention.

[0036] FIG. 2 is a perspective view for explaining the external appearance of the smart sheet (100) of the present invention.

[0037] The present invention relates to an AI posture correction system (10) comprising a smart seat (100) for customized spinal care and a computing device (200), and a method thereof. The smart seat (100) proposed in the present invention is preferably placed on a chair or cushion and is intended for a user to sit on the smart seat (100). It provides a technology capable of monitoring and correcting the user's posture by collecting pressure data generated in real time at the part in contact with the user's body and analyzing it. To address the issue that existing posture correction devices cause inconvenience in the user's daily life or are uncomfortable to wear, the smart seat (100) of the present invention is designed to correct posture in a more sophisticated and user-friendly manner using sensors and a control device.

[0038] The smart seat (100) provided in the present invention detects the pressure distribution of the body through a pressure sensor and transmits this to a computing device (200) to calculate the center point of pressure and the center point of the pressure area. It operates by recognizing the user's posture pattern based on this difference value and outputting an AI-customized motion that can correct the user's posture from the posture pattern. This method provides real-time feedback to the user, helping them maintain a correct posture even when sitting for a long time.

[0039] In addition, the present invention differentiates the sensitivity of pressure sensors by body part to enable more accurate measurement of pressure generated in areas such as the buttocks and thighs, thereby preventing deformation of the spine and pelvis. Furthermore, the efficiency of posture correction is enhanced by providing an immediate warning through a real-time feedback system utilizing a vibration motor if the user adopts an incorrect posture or sits for longer than a set time.

[0040] Accordingly, the smart seat (100) of the present invention has the advantage of improving the problems of existing posture correction devices and providing a more convenient and effective posture correction solution to the user by providing real-time posture monitoring, sophisticated pressure data analysis, and user-customized feedback functions.

[0041] The frame portion (110) is a key element that provides structural stability to the smart seat (100) and serves to effectively support and distribute pressure exerted by the user's body. The frame portion (110) is designed to continuously apply upward and inward compressive force, which helps the user maintain a stable and balanced posture. The frame portion (110) is generally made of a durable material so that it undergoes minimal deformation even with prolonged use and maintains elasticity under repeated pressure.

[0042] The design of the frame portion (110) affects the pressure distribution of the entire seat and is organically linked with the sensor portion (130) that collects data related to the user's posture. By the frame portion (110) efficiently adjusting the pressure in response to the movement of the user's body, it can evenly distribute the force applied by the body to the seat and contribute to reducing the excessive burden on body parts caused by an unbalanced posture.

[0043] The sensor unit (130) is an important component of the smart seat (100) that detects and analyzes the pressure distribution of the user's body in real time. This sensor detects pressure changes occurring between the user and the seat and can monitor the user's posture or movement based on the data. Through this, the seat can perform various functions to help correct the user's posture or maintain a comfortable state.

[0044] The sensor unit (130) is mainly composed of a sensing device such as a pressure sensor and measures the intensity and distribution of pressure applied to a specific area. The sensor unit (130) is distributed throughout the entire sheet so that it can finely detect pressure changes occurring in various parts of the body. Through this, the movement of the user's body can be recognized and the body's balance can be measured. The movement and balance measured in this way can serve as a basis for suggesting exercises to the user to improve their balance in the future.

[0045] Additionally, the sensor unit (130) can transmit collected pressure data to a computing device (200) to contribute to analyzing the user's posture or movement patterns. Through this, the smart seat (100) can continuously induce the user to maintain a posture suitable for them and provide warnings or notifications if necessary.

[0046] Additionally, the sensor unit (130) can transmit collected data to a central processing unit to contribute to analyzing the user's posture or movement patterns. Through this, the smart seat (100) can continuously induce the user to maintain a posture suitable for them and provide warnings or notifications if necessary.

[0047] The control unit (140) includes a communication unit, a battery unit, and a processor, and can supply power to the sensor unit (130) and transmit pressure data to an external computing device (200) through the communication unit. In this process, a series of functions can be performed by the processor of the control unit (140) by performing operations predefined in the firmware. However, since the processor of the control unit (140) is designed to reduce the unit cost of the product and to promote mass distribution, it does not perform complex computational operations such as the first center point and the second center point, and can focus on the role of transmitting pressure data so that such operations can be processed by the external computing device (200).

[0048] In addition, the computing device (200), which includes a communication unit (230), a processor (210), a memory unit (220), and a display unit, is a digital device that performs communication and includes a display function. Any digital device equipped with memory means and equipped with a microprocessor to have computational capabilities, such as a desktop computer, a laptop computer, a workstation, a PDA, a web pad, or a mobile phone, can be adopted as the computing device (200) according to the present invention. Furthermore, although it is described as a computing device (200) below, the processes described below may be executed through an app (program) installed on the computing device (200).

[0049] The computing device (200) can provide the user with exercise information necessary for posture correction or notifications regarding imbalances occurring in the current posture through a display unit. Through this, the user can identify their physical condition in real time and make necessary adjustments, and the system can induce more effective posture correction. The information provided through the display unit of the computing device (200) can be designed to be intuitively visualized so that the user can easily understand it, and can contribute to the improvement of the user's posture by providing continuous feedback through a notification function.

[0050] The computing device (200) can perform the role of processing and analyzing pressure data received from the control unit (140). This device has processing capabilities capable of performing complex calculation tasks such as a first center point and a second center point, and can receive only simple data from the control unit (140) and process various calculations and analyses based on it.

[0051] The computing device (200) can be connected to an internal or external server to enable real-time data processing, and may include the ability to visualize result data (see FIG. 9 described later) or, if necessary, transmit additional control commands back to the control unit (140). Through these roles, the accuracy and efficiency of the system can be increased, and meaningful information can be provided to the user.

[0052] FIG. 3 is an exploded perspective view for explaining the sensor part (130) of the smart sheet (100) of the present invention.

[0053] FIG. 3 is an exploded perspective view for explaining the sensor unit (130) of the smart seat (100), visually showing the structural arrangement of the smart seat (100). The sensor unit (130) may be positioned between the cushion unit (120) and the frame unit (110) of the smart seat (100), and the sensor unit (130) may perform the role of detecting the user's pressure and collecting data. The wiring (134) of the sensor unit (130) is connected to the control unit (140), through which pressure data can be transmitted to the control unit (140). The control unit (140) may perform the role of transmitting this data to a computing device (200), thereby enabling the overall system to operate organically.

[0054]

[0055] FIG. 4 is a drawing showing a pressure sensor of a sensor unit (130) according to one embodiment of the present invention.

[0056] The distribution of pressure sensors in the sensor unit (130) includes a structural arrangement designed to precisely detect and analyze the user's body pressure. The pressure sensors measure the pressure distribution occurring in various parts of the body, thereby enabling real-time monitoring of the balance state of the user's posture.

[0057] Pressure sensors can generally be distributed in the following ways:

[0058] 1. Even distribution: The pressure sensors of the sensor unit (130) are evenly distributed across the surface of the seat and detect pressure in all areas where the user's body comes into contact. This method is useful for determining the overall pressure distribution and can evenly measure pressure occurring in various parts of the body. For example, when the user sits evenly on the seat, the pressure sensors detect pressure evenly in all areas of the buttocks, thighs, and lower body. This allows real-time monitoring of whether the user's body is sitting symmetrically or if pressure is concentrated in a specific area. This method is effective for maintaining overall body balance and is advantageous for quickly identifying and correcting imbalances in the sitting posture.

[0059] 2. Zone-based distribution: The sensor unit (130) can be concentrated in specific zones of the seat, and these zones are designed to correspond to specific parts of the body, for example, the buttocks and thighs. This method helps to perform posture correction more precisely by measuring the pressure distribution in major parts of the body. For example, if sensors are zone-based in the buttocks area and the thigh area, the pressure applied to the two areas can be measured individually. This method can help to accurately determine the angle of the spine and the alignment of the pelvis, and can monitor in real time the situation where strain is applied to specific parts of the spine and induce appropriate corrective movements. This zone-based analysis is advantageous for providing customized posture correction functions for each body part.

[0060] 3. Differentiated Distribution of Center and Periphery: Pressure sensors can be positioned differently in the center and periphery of the seat. By placing high-sensitivity sensors in the center and low-sensitivity sensors in the periphery, the design can be made to respond sensitively to pressure changes in the center. This placement offers the advantage of more accurately measuring pressure concentration. For example, if excessive pressure is applied to the center due to a user sitting in an incorrect posture, the sensor can immediately detect this and send a warning to the user. On the other hand, because the periphery has relatively lower sensitivity, posture correction can be performed by focusing more on subtle changes in the center. This method enables more precise posture correction by allowing the system to recognize and correct even minute changes in posture when a user sits for a long time.

[0061] These pressure sensors are individually arranged in the sensor unit (130), but a wiring structure is required to transmit their data to the control unit (140). Signals generated from the pressure sensors are transmitted to the control unit (140) through the wiring (134), and the control unit (140) is responsible for analyzing these signals or transmitting them to a computing device (200).

[0062] The method for calculating the pressure area in the present invention includes a process in which a pressure sensor detects pressure generated at the part where the body comes into contact with the seat and processes the data. Pressure sensors distributed in the sensor unit (130) measure pressure changes applied when a user sits on the seat in real time, and through this data, the area where the body comes into contact can be identified. Each sensor detects pressure changes in a specific part according to its location, and a pressure distribution is formed through this.

[0063] In this process, pressure values ​​measured by the sensors are first collected, and then only those sensors that receive pressure above a certain threshold are activated to define the area in contact with the body. Since these activated sensors detect pressure at different locations, the entire area in contact with the body can be visually represented based on their location information and pressure data. The pressure zone is established based on the coordinates of the sensors to which pressure is applied, playing a crucial role in specifically identifying the parts of the user's body that come into contact.

[0064] For example, the pressure area may include a virtual circular sub-pressure area for each pressure sensor, and the pressure area may refer to an area formed by merging all said sub-pressure areas. In this case, each sub-pressure area may be set to be inversely proportional or proportional to the sensing sensitivity of the corresponding pressure sensor. Specifically, the fact that the size of the sub-pressure area may be set to be inversely proportional or proportional to the sensing sensitivity of the pressure sensor means that the range of each sub-pressure area is adjusted according to the sensitivity of the sensor.

[0065] First, taking the example where the sub-pressure area is set inversely proportional to the sensing sensitivity, a sensor with higher sensing sensitivity will have a narrower sub-pressure area. For instance, if a pressure sensor located on the buttocks has very high sensitivity, the range it can accurately detect becomes narrower, and consequently, the sub-pressure area forms a small circle. Conversely, a sensor with low sensing sensitivity tends to detect a wider range, so its sub-pressure area may form a large circle. This method is designed to allow high-sensitivity sensors to finely detect pressure in specific areas, providing an environment that can respond sensitively to specific pressure changes.

[0066] Secondly, taking the example where the sub-pressure area is set proportionally to the sensing sensitivity, the higher the sensing sensitivity, the wider the sub-pressure area is set. For instance, a high-sensitivity sensor located on the thigh can accurately detect pressure occurring in that area over a wide range, so its sub-pressure area can be formed as a large circle. Conversely, a low-sensitivity sensor detects pressure within a narrow range, so its sub-pressure area is set as a small circle. This method is advantageous for grasping the pressure distribution of the entire body at a glance, as a high-sensitivity sensor can detect a wider range of pressure distribution.

[0067] For example, it can be assumed that when a user sits on the seat, sensors with different sensitivities are placed on the buttocks and thighs, respectively. It can be assumed that the sensor located on the buttocks has low sensitivity and a wider sub-pressure area, while the sensor located on the thighs has high sensitivity and a narrower sub-pressure area. The sub-pressure areas detected by these sensors are merged according to their respective sizes to form a final total pressure area. Through this, the pressure distribution of the part of the user's body in contact with the seat can be analyzed more precisely, and differentiated pressure measurements based on sensor sensitivity become possible.

[0068] The first center point corresponds to the center of gravity of the pressure area, which is calculated based on the distribution of pressure detected by each pressure sensor when a user sits on the seat. The method for calculating the first center point involves the process of mathematically deriving the center of gravity within the pressure-applied area based on data collected from the pressure sensors. The method for calculating the first center point can be described as follows.

[0069] First, based on the pressure data collected from each pressure sensor, the coordinates of each sensor and the pressure value applied to it are measured. Pressure sensors are typically placed at regular intervals on the surface of the sheet, and each sensor has specific location coordinates. The pressure value detected by each sensor represents the body pressure generated at that coordinate.

[0070] The centroid of the total pressure distribution is calculated by assigning weights to the pressure values ​​collected from each sensor. In this process, the location coordinates of each sensor and the pressure value applied to it are used together. Generally, the higher the pressure value, the greater the influence the sensor's location has on the calculation of the total centroid.

[0071] The first center point is calculated using a weighted average based on the pressure values ​​and coordinates measured by all sensors. Specifically, the pressure values ​​of all sensors are multiplied by their corresponding coordinates, the results are summed, and the sum is divided by the total sum of the pressure values. Through this process, the center of gravity of the pressure region relative to the point where pressure is concentrated—that is, the first center point—is derived.

[0072] The method for calculating the second center point is the process of determining the center of the entire area where pressure values ​​are distributed, based on the pressure values ​​detected by the pressure sensors. This center point serves as an important indicator to identify where the body's pressure is concentrated when the user is seated on the seat.

[0073] To explain in detail, multiple pressure sensors distributed across the seat detect pressure generated at the points of contact with the user's body. Each sensor has a unique location and collects the pressure values ​​detected at that position. The process of calculating the second center point involves applying different weights based on the pressure values ​​detected by each sensor to find the average point where the pressure is most distributed across the area where all sensors are located.

[0074] For example, when different pressure values ​​are detected in areas such as the buttocks or thighs, the location of the sensor detecting higher pressure has a greater impact on the overall center point calculation. Conversely, the location of the sensor detecting relatively lower pressure has less impact on the center point calculation. Through this process, the average center point of the entire pressure distribution—that is, the second center point—is derived, centering on the point where the sensors with high pressure are located.

[0075] In one embodiment, a method may be applied in which the sensing sensitivity of a corresponding pressure sensor is increased as it moves further away from a first center point within a pressure area. The first center point represents the location where pressure is most concentrated when a user sits on the seat, and the sensitivity is adjusted by measuring the distance between sensors based on this point. This method maintains low sensitivity for sensors close to the center point, as they can detect relatively large pressures, while increasing sensitivity as they move further away from the center point to detect small pressure changes.

[0076] In one embodiment, since pressure sensors located far from the first center point may be less sensitive to minute pressure changes compared to the center, increasing the sensitivity of these sensors allows for the detection of more precise pressure changes. For example, it is possible to more accurately measure minute pressure differences occurring in areas where the user's weight is not concentrated, which can be effective in detecting minute changes in posture or imbalances.

[0077] In another embodiment, by increasing the sensitivity of sensors located far from the first center point, even small pressure changes occurring in the periphery can be reacted to quickly. This enables the detection of not only pressure concentrated at the center but also minute movements of the body or changes in posture. This method provides more precise data for user posture correction and can assist in monitoring and correcting posture in real time.

[0078]

[0079] FIG. 5 is a drawing illustrating examples of a pressure region, a first center point, and a second center point according to an embodiment of the present invention.

[0080] The computing device (200) can calculate the difference between the first center point and the second center point. Specifically, the computing device (200) can first calculate the coordinates of the two center points, derive the coordinates of the first center point and the second center point based on the pressure values ​​and positions detected by each sensor, and then calculate the difference between these two coordinates. The difference can be expressed as the distance between the two center points, through which the user's body can be evaluated in a balanced state while sitting on the seat and the posture pattern can be recognized.

[0081] If the difference calculated shows a large gap between the first and second center points, it indicates that the user is sitting asymmetrically on the seat or that pressure is excessively concentrated on specific areas. In other words, it corresponds to an unbalanced postural pattern.

[0082] Conversely, when the two center points nearly coincide, it indicates that the body's weight is evenly distributed and the posture is balanced. In other words, the posture pattern corresponds to a balance pattern.

[0083] The computing device (200) of the present invention can analyze difference values ​​to recognize posture patterns (imbalance patterns, balance patterns, etc.) and output AI-customized movements related to posture correction from the posture patterns. That is, it can provide AI posture correction feedback for spinal care to the user.

[0084] For example, let's assume that a user sits on the smart seat (100). At this time, the smart seat (100) detects the pressure exerted by each part of the user's body on the seat in real time through pressure sensors. Pressure sensors collect various values ​​centered on the buttocks and thighs, and the computing device (200) calculates a first center point and a second center point based on this data.

[0085] First, the first center point is the point in the body where the most pressure is applied. For example, if a user is sitting in a tilted position to the right, applying more pressure to the right buttock, the first center point will be located near the right buttock.

[0086] The second center point is the average center of the total pressure applied by the user to the seat. For example, if the pressure distribution across the entire body is evenly distributed across the thighs and buttocks when the user sits on the seat, the second center point will be located at the center of the pressure area.

[0087] Now, the computing device (200) calculates the difference in coordinates between the first center point and the second center point. If the difference between the two center points is large, it means that the user is sitting with the seat tilted to the right, and thus pressure is excessively concentrated (an unbalanced pattern) on a specific part. For example, if the first center point is located near the right buttocks and the second center point is close to the center of the seat, this difference indicates that the user's posture is asymmetrical.

[0088] Conversely, when the user is sitting in a balanced posture, the first and second center points nearly coincide (balance pattern). For example, when pressure is evenly distributed between the buttocks and thighs, the difference between the two center points becomes very small, which indicates that the user is sitting in a balanced posture.

[0089] The computing device (200) can analyze these difference values ​​and send a warning or provide real-time feedback if the user is in an unbalanced posture. For example, it can give the user a notification saying "Correct your posture tilted to the right" or give a signal through vibration that posture correction is needed. Through this, the user can immediately recognize and correct their posture. In addition, based on these difference values, it can recommend exercises to the user to correct the collapsed center of gravity.

[0090] For example, one can assume a situation where a user sits for a long time in a posture tilted to the right. According to data detected by a pressure sensor, the first center point is skewed toward the right buttocks, and the second center point is located in the center of the seat. A large difference between these two center points indicates that the user is sitting in an unbalanced posture to the right, which may cause long-term strain on the spine and pelvis. In this situation, the computing device (200) can analyze this difference value and recommend customized exercises to restore the body's balance. Through this, the user can restore a balanced posture and maintain the health of the spine and pelvis in the long term.

[0091] Additionally, as another embodiment, a situation may be assumed where a user sits on the same seat for more than a preset time (e.g., 1 hour). In this case, the computing device (200) can use a pressure sensor of the smart seat (100) to determine whether the user is sitting on the smart seat (100), and if so, for how long. If it is confirmed that the user has been sitting on a single seat (100) for more than a preset time (e.g., 1 hour), the computing device (200) can give the user a notification saying, "You have been sitting for too long. Please stand up for a moment," or cause them to stand up via vibration, or give a signal that stretching or movement is needed. Through this, the user can immediately recognize that they have been sitting for a long time and correct their posture.

[0092] Additionally, as an additional embodiment, when a user sits on the smart seat (100) and crosses their legs, the pressure distribution changes according to the user's body posture. Generally, the motion of crossing the legs shifts the body's center of gravity in the direction of the twist, that is, toward the raised legs, and at the same time, the center of gravity tends to shift backward. Based on this, the computing device (200) can perform posture analysis through the following procedure.

[0093] First, when a user sits with their legs crossed, the pressure applied to the seat may be more concentrated on the buttocks on the side of the crossed legs. For example, if a user sits with their left leg crossed over their right leg, most of the pressure is concentrated on the right buttocks and right thighs. Based on the pressure data at this time, the computing device (200) calculates the first center point to a position shifted to the right. On the other hand, the second center point will still be located in the center, as it is the average center of the pressure applied to the entire seat.

[0094] In this situation, a difference between the two center points occurs, and this difference reflects the body's center of gravity shifted to the right. At the same time, as the user's upper body tilts slightly backward due to the leg-crossing motion, the first center point can move to the rear of the seat in the vertical direction as well as the horizontal direction. The computing device (200) can calculate this vertical difference as well and recognize that the user's posture is tilted backward with the legs crossed.

[0095] The computing device (200) can detect the asymmetry of the posture and the shift in the center of gravity, and then provide feedback to the user. For example, the UI can provide a warning message such as, "You are currently sitting with your right leg crossed. This posture can strain your spine, so please uncross your leg and maintain a balanced posture." At this time, the computing device (200) can activate a vibration motor to correct the user's posture or provide feedback in an appropriate direction to indicate which side the body is tilted to depending on the leg's crossing.

[0096] In addition, since sitting with legs crossed for a long time can cause imbalances in the pelvis and spine, the computing device (200) can suggest customized exercises to prevent related long-term health problems. For example, when sitting with legs crossed to the right, the computing device (200) can suggest specific stretching movements with a notification saying, "Correct your leg-crossing posture and recommend pelvic stretching exercises." Through this, the user can become aware of their physical condition and contribute to maintaining long-term health.

[0097] In conclusion, based on the shift in the center of gravity caused by the leg-crossing motion, the computing device (200) can calculate the asymmetric pressure distribution of the body and, based on this, provide real-time feedback and exercise suggestions so that the user can maintain a balanced posture.

[0098] Meanwhile, the computing device (200) can acquire collected data including pressure data received from the smart seat (100), sensing data acquired from other advanced sensors (e.g., waist belt, smart watch, etc.), evaluation data input by the user, vision data related to the user's posture acquired through an external camera (e.g., 3D motion tracking), and goals input by the user.

[0099] Additionally, the processor (210) of the computing device (200) can recognize a posture pattern using the collected data as an input value through a posture pattern recognition AI algorithm (recognition of repeating patterns in a series of period-unit data based on Fourier transform). The recognized posture pattern includes a posture pattern at a certain point in time, but may also include a posture pattern accumulated over a certain period.

[0100] The processor (210) of the computing device (200) can derive posture habit information or spinal health information based on posture patterns accumulated over a certain period of time. At this time, the posture habit information, referring to FIG. 9 to be described later, can distinguish between the time spent maintaining a proper posture (e.g., 76.8 minutes (64%)) or the time spent maintaining a skewed posture (e.g., 43.2 minutes (36%)) during a measurement time (e.g., 120 minutes), and can include information (time, ratio, etc.) regarding the time spent skewed by direction (e.g., 8 directions) relative to the user.

[0101] At this time, the sum of the skewed times for all 8 directions (6 + 5 + 4.5 + 5.5 + 5 + 6 + 4 + 7.2) may correspond to the time of maintaining the skewed posture (e.g., 43.2 minutes).

[0102] Additionally, the computing device (200) can derive spinal health information based on accumulated posture patterns. The spinal health information can be obtained by inputting the posture patterns into a pre-learned spinal health change prediction AI algorithm, and may include probability information such as “there is an 80% probability of back pain occurring within 3 weeks.”

[0103] Next, the computing device (200) can output an AI-customized movement through the posture habit information or spinal health information, etc. At this time, an AI-customized exercise recommendation algorithm learned based on posture cluster data, existing analysis data (posture improvement degree, goal, etc.), recommended exercise programs and scores, etc. can be used (the AI-customized exercise recommendation algorithm can be examined in more detail in FIG. 13, which will be described later).

[0104] Specifically, the computing device (200) can derive a user-customized movement (AI-customized movement) through a display unit using an AI exercise recommendation algorithm with the posture habit information or spinal health information as input values. Here, the user-customized movement (AI-customized movement) may include at least one recommended exercise and information regarding the time, sequence, etc., for each recommended exercise.

[0105] The processes in the aforementioned computing device (200) can also be verified through FIGS. 10 to 15, which will be described later.

[0106]

[0107] FIG. 6 is a drawing illustrating examples of a buttocks region and a thigh region inside a pressure region according to one embodiment of the present invention.

[0108] The method for calculating the buttocks region within the pressure domain is a process of specifically distinguishing and analyzing the area by measuring the pressure applied to the buttocks when a user sits on the seat. This method consists of a procedure that uses pressure sensors to define the buttocks region based on the location where the part of the user's body corresponding to the buttocks contacts the seat and the pressure applied thereto.

[0109] First, pressure sensors installed on the seat collect pressure data in real time when the user sits down. By analyzing the collected data, it is possible to identify which parts of the body are subjected to pressure. Since the buttocks are the body's primary point of gravity, they are the area that exerts the greatest pressure when sitting on the seat. High pressure values ​​detected by the sensors can be interpreted as corresponding to the buttocks area.

[0110] The key factors in the process of calculating the buttocks region are as follows. First, the buttocks region can be determined based on pressure values ​​measured by sensors that exceed a certain level. For example, this method involves collecting the coordinates of sensors where pressure is detected above a certain threshold and establishing that set of coordinates as the buttocks region. In this case, the buttocks region can be expanded to include surrounding sensors, centering on the area where the highest pressure is applied.

[0111] Second, the buttock area can be specifically identified based on the distribution of pressure values. This method calculates the contact range of the buttocks by considering the central point where pressure is most concentrated and the surrounding pressure distribution. Since the buttocks exhibit a pressure distribution pattern different from that of the thighs, a boundary line for the buttocks distinct from the thighs can be established by analyzing data detected by sensors. This boundary line is dynamically defined based on changes in the sensors' pressure values, thereby enabling a clear distinction of the buttock area.

[0112] Third, the coordinates and size of the buttocks region can be calculated by analyzing the shape and magnitude of the pressure distribution. Generally, since the buttocks exert pressure over a wide area, the shape of the pressure distribution can be analyzed to estimate its area, which can then be set as the buttocks region.

[0113] The method for calculating the thigh area involves detecting the pressure applied to the thighs when a user sits on the seat using pressure sensors and analyzing the contact points based on this data. This process is distinct from the gluteal area, and since the thighs are located at the front of the lower body, they exhibit a different pressure distribution pattern.

[0114] First, pressure sensors distributed across the seat collect the user's pressure data in real time. While the thighs apply relatively lower pressure compared to the buttocks, the pressure is distributed over a wide area. Therefore, when calculating the thigh region, it is important to detect areas that have a continuous pressure distribution extending beyond a certain range, even though the pressure values ​​are smaller than those of the buttocks.

[0115] The method for calculating the thigh area can be explained as follows.

[0116] First, it is necessary to identify the portion corresponding to the thigh among the pressure values ​​measured by the sensors. Since the thigh exerts relatively less pressure compared to the buttocks, the thigh region can be defined based on pressures below a specific threshold. In other words, the thigh region can be defined based on a certain range of pressure values ​​detected by the sensors. For example, high pressure values ​​detected in the buttocks region can be excluded, and the area applying lower pressure can be considered the thigh.

[0117] Second, since the thighs come into greater contact with the front part of the seat, the thigh area can be calculated by considering the sensor locations and pressure distribution. The area where the thighs make contact can be extracted based on the pressure values ​​detected by sensors located at the front of the seat. In this process, the wide contact area of ​​the thighs is identified based on the coordinates of the sensors, and the set of these coordinates is set as the thigh area.

[0118] Third, by analyzing the distribution of pressure values, a boundary line distinguishing the thighs and buttocks can be established. The buttocks apply relatively greater pressure to the sheet, while the thighs have lower pressure distributed over a wider area. By utilizing this difference, the boundary between the buttocks and thighs can be established based on changes in pressure values, and the thigh area can be calculated in detail.

[0119] Finally, the thigh region is calculated based on the pressure distribution widely spread across the front part of the seat. Since the thigh covers the lower front of the body and has a large surface area in contact with the seat, the thigh region can be accurately derived by analyzing this pressure distribution.

[0120] Generally, the buttocks are the area where the body's center of gravity is concentrated, so they exert relatively high pressure when sitting on the seat. On the other hand, the thighs exert relatively lower pressure compared to the buttocks, but they transmit pressure distributed over a wider area to the seat. Considering these differences in pressure characteristics between body parts, it is necessary to set the sensing sensitivity differently.

[0121] In one embodiment, since the pressure sensor in the buttocks area detects relatively high pressure, the sensing sensitivity can be set low. This prevents the detection of unnecessary minute changes within a range where changes in the pressure exerted by the buttocks on the seat can be sufficiently detected, thereby preventing excessive fluctuations in data and allowing the focus to be on measuring only large changes in the pressure exerted by the buttocks. For example, since the buttocks account for the majority of the pressure when a user sits, there is relatively little need to measure the pressure in the buttocks area with high sensitivity.

[0122] On the other hand, since the thigh area has a large contact surface with the seat and is subjected to relatively low pressure, the pressure sensor in the thigh region can be set to a high sensitivity to precisely detect pressure changes. This allows for the sensitive measurement of even small pressure fluctuations in the thighs, enabling a more precise understanding of how the body is distributed across the seat and whether it is well-balanced. For example, although pressure changes in the thighs are subtle, they can provide important information about the user's posture or weight distribution.

[0123] In another embodiment, the computing device (200) can dynamically adjust the sensitivity of each sensor while monitoring pressure changes in the buttocks and thigh areas in real time according to changes in the user's posture. When the user changes posture and more pressure is applied to the thigh area, the sensitivity of the thigh sensor can be set higher to detect minute pressure changes more precisely, and conversely, when more pressure is applied to the buttocks area, the sensitivity of the buttock sensor can be lowered to detect only large pressure changes.

[0124]

[0125] FIG. 7 is a drawing illustrating an example for explaining a haptic part according to an embodiment of the present invention.

[0126] The haptic unit is a device that provides physical tactile feedback to the user. Based on data detected by a pressure sensor, it provides immediate feedback through tactile signals, such as vibration, when the user assumes an incorrect posture or sits for longer than a preset time (e.g., 1 hour). The haptic unit primarily includes mechanical devices such as vibration motors and helps the user recognize pressure changes or postural imbalances occurring in specific parts of the body.

[0127] In the case of the smart seat (100), the haptic unit monitors the user's pressure distribution and provides a warning signal via vibration when the user assumes an improper posture or when pressure is not evenly distributed (or when sitting for longer than a preset time). For example, if the user sits for a long time in a posture leaning to one side, the haptic unit generates vibration in that area to notify the user to correct their posture. This vibration signal is immediately transmitted to the user, allowing them to quickly recognize and correct the incorrect posture they unconsciously maintain.

[0128] The haptic unit may include a single vibration motor or multiple vibration motors, and the number and position of the motors may be set differently depending on the pressure distribution. If there are multiple vibration motors, the motors closer to the areas of the body where pressure is applied unevenly may be activated based on data detected by the pressure sensor, thereby providing more specific and accurate feedback to the user.

[0129] In the case where a single vibration motor is provided in the haptic unit, only one vibration motor is installed in the seat. In this method, the control unit (140) calculates the difference between the first center point and the second center point based on data collected from the pressure sensor. If this difference exceeds a set threshold, one vibration motor operates to send a signal to the user to correct the incorrect posture. For example, if the user's body is tilted to the right, the pressure sensor detects more pressure on the right buttocks, and as a result, a difference occurs between the two center points. If this difference exceeds the threshold, the vibration motor operates to provide an overall warning. This method has a simple structure and is advantageous for providing overall feedback, but it may have the disadvantage that it is difficult to know specifically where the posture imbalance occurred.

[0130] When multiple vibration motors are provided in the haptic unit, vibration motors are installed in various parts to provide more detailed feedback. The control unit (140) calculates the difference between the first center point and the second center point through the data of the pressure sensor, and if the difference exceeds a threshold value, it activates the vibration motor closest to the first center point. For example, if the user is in a posture tilted to the right, the vibration motor located on the right buttocks is activated so that the user immediately recognizes that the weight is shifted to the right. This method can clearly convey the specific location where the postural imbalance occurred, allowing the user to know more accurately which part of the body needs to correct their posture. Although the method using multiple vibration motors provides specific and detailed feedback, it may have the disadvantage of increasing the complexity of the structure and installation costs.

[0131]

[0132] FIG. 8 is a drawing illustrating an example for explaining the film portion of a sensor portion (130) according to one embodiment of the present invention.

[0133] The film portion is a major component of the sensor portion (130) and serves to support the pressure sensor so that it can operate stably and to accurately detect various forms of body pressure. The film portion is made of a flexible and durable material and has structural characteristics that allow it to adapt to the curved shape of the sheet.

[0134] The primary purpose of the film section is to facilitate the proper placement of pressure sensors along the curves of the sheet. Since the sheet surface features various curves tailored to the user's body shape, the film section is designed to be flexible so that the sensors can accurately detect body pressure even when deformed to conform to these shapes. This flexibility is essential for accurately transmitting pressure at the points where the body contacts the sheet.

[0135] In addition, the film portion must possess not only durability but also sufficient flexibility to ensure user comfort even during prolonged use. If the film portion is too rigid or thick, it may cause discomfort to the user; conversely, if it is excessively thin or weak, there is a possibility of wear or damage during repeated use. Therefore, the film portion is composed of a material that considers the balance between durability, flexibility, and user convenience.

[0136] In conclusion, the film portion is made of a flexible material designed with the precise placement of the pressure sensor, long-term durability, and user convenience in mind, and plays an important role in accurately detecting changes in body pressure and transmitting them to the sensor.

[0137] In one embodiment, a case can be considered in which the frame portion (110) of the smart seat (100) is designed to fit the curves of the human body. The frame portion (110) is made of a curved surface to distribute pressure across the buttocks and thighs of the body. A film portion of the sensor portion (130) is positioned to fit this curved surface, and the film portion is made of a flexible material so that it can be freely deformed along the curved shape of the frame portion (110).

[0138] At this time, the film portion has a plurality of grooves (133) formed symmetrically along the curved surface of the frame portion (110). These grooves (133) serve to help the film portion adhere more closely to the curved surface. That is, the grooves (133) formed in the film portion match the curved structure of the frame portion (110), thereby allowing the film portion to be more stably coupled with the frame portion (110). For example, if the frame portion (110) has a structure that supports the buttocks and thighs of the human body, the film portion is naturally positioned along the grooves (133) so that pressure sensors can be positioned precisely aligned with that area.

[0139] Multiple pressure sensors are placed on the film portion and are evenly distributed along the groove (133) on the curved surface. Because the film portion is flexible, the sensors can accurately detect the pressure applied by the user's body to the frame portion (110) without distortion. For example, even if the pressure is concentrated on the buttocks or thighs in accordance with the curvature of the body when the user is sitting, the pressure sensors can provide accurate data as the film portion deforms to fit the curve of the corresponding area.

[0140] In one embodiment, a situation can be assumed in which the smart seat (100) detects pressure corresponding to the buttocks and thighs of the user's body. In this case, the film section is configured to be divided into a first film section (130-1) and a second film section (130-2) so as to detect the buttocks and thighs separately. The first film section (130-1) is positioned on the area corresponding to the right buttocks and thighs of the user's body, and the second film section (130-2) corresponds to the left buttocks and thighs. The two film sections are arranged in a symmetrical structure so that when the user sits on the seat, the pressure on the left and right sides of the body can be detected individually.

[0141] At this time, the first film section (130-1) and the second film section (130-2) are each independently equipped with pressure sensors, and each film section is designed in a left-right symmetrical shape to fit the curved surface of the frame section (110). For example, if the user's body is tilted to the right, more pressure will be applied to the first film section (130-1), and relatively less pressure will be applied to the second film section (130-2). Through this, the two film sections can detect the pressure difference between the left and right sides of the body, and this data can transmit a command message to the user terminal to sit upright from the current posture for posture correction.

[0142] Additionally, since the two film sections are separated, each film section can deform individually and detect pressure. This is advantageous for collecting accurate data even when the left and right sides of the body are not completely identical or when more pressure is applied to one side of the body depending on the posture. For example, if a user sits with their left leg crossed, the pressure applied to the second film section (130-2) increases, and the pressure on the first film section (130-1) decreases relatively. Based on this information, the smart seat (100) can recognize the user's postural imbalance and send a correction signal.

[0143] Meanwhile, methods utilizing a smart seat according to an embodiment of the present invention may be implemented by using a program on an external computing device (200) or by a processor included inside the smart seat. For example, the program may calculate a first center point, which is the center point of pressure, based on pressure data collected from the sensor unit (130) of the smart seat (100). Subsequently, it may calculate the difference value between this first center point and a second center point, which is the center point of the pressure area. The program may analyze this difference value to determine the action required for correcting the user's posture and output the corresponding action to guide the user. That is, it may transmit a command message to the user terminal to sit in a correct posture from the current posture.

[0144]

[0145] FIG. 9 is a drawing illustrating a specific example of a posture correction system that can be implemented in a computing device (200) according to one embodiment of the present invention.

[0146] When a user sits on the smart seat (100), pressure generated from the body is transmitted to the sensor unit (130) of the seat. This sensor unit includes a plurality of pressure sensors, which measure the distribution of pressure applied by the user's body to the seat. A computing device (200) receives this data and calculates the center point of pressure (first center point). The first center point represents the balance of the overall force applied by the user's body to the seat.

[0147] At the same time, a second center point, which is the structural center of the sheet itself, is also pre-set. The computing device (200) calculates the difference value between the first center point and the second center point, and this value serves as the basis for analyzing the user's posture. For example, if the user's body is tilted to the right, the first center point moves to the right, and a difference occurs with the second center point.

[0148] If this difference value exceeds a threshold, the processor (210) of the computing device (200) can provide a UI to the user through the display unit to notify the user of this. At this time, the measurement time (e.g., 120 minutes), the time maintained in a normal posture (e.g., 76.8 minutes, 64%), the time maintained in a skewed posture (e.g., 43.2 minutes, 36%), etc., may be provided.

[0149] In addition, as can be seen in the drawing, this UI provides intuitive feedback to the user and can clearly indicate which direction the body is visually leaning in the case of a skewed posture and what posture correction is needed (e.g., the position of colored text indicating a specific direction).

[0150] For example, as can be seen in Fig. 9, the skewed posture relative to the user on the screen can be classified into 8 directions (e.g., top, top right, right, bottom right, bottom, bottom left, left, top left, etc.), and the UI can display how much the user was skewed in which direction during the measurement time.

[0151] Referring to Fig. 9, it can be seen that the user maintained a body tilted toward the upper left for 7.2 minutes and a body tilted toward the lower side for 5 minutes.

[0152] According to one embodiment of the present invention, the computing device (200) may display different visual warning indicators depending on the proportion (or time) of the skewed posture. For example, the degree of the skewed posture may be divided into four stages, and based on the total measurement time, if the proportion is less than 4%, stage 0 (no color indication); if it is 4% or more, stage 1 (one color indication starting from the innermost); if it is 5% or more, stage 2 (two colors indication starting from the innermost); and if it is 6% or more, stage 3 (three colors indication).

[0153] Additionally, the computing device (200) may provide feedback to the user through vibration. When a threshold is exceeded, a vibration motor built into the smart seat is activated to notify the user that they are maintaining an incorrect posture. This gives the user the opportunity to immediately recognize and correct their posture.

[0154] Specifically, according to one embodiment of the present invention, the computing device (200) may provide vibration feedback when the difference value between the aforementioned first center point and second center point exceeds a threshold (based on the physical degree of the skewed posture). Additionally, according to another embodiment, vibration feedback may be provided when the proportion of time during which the posture is skewed in a specific direction during the total measurement time exceeds a threshold (based on time).

[0155] According to another embodiment, the computing device (200) may provide vibration feedback when it detects that a user has been sitting for a certain period of time (e.g., 1 hour) or longer (sitting on the smart seat).

[0156] In conclusion, the computing device (200) can calculate the deviation of posture based on data collected from the sensor and, based on this, provide visual UI feedback to the user through an app or vibration notification through a smart seat to help correct the posture.

[0157]

[0158] FIGS. 10 to 15 are drawings illustrating specific examples of a system (hereinafter, system) incorporating an IoHT (Internet of Health Things)-based AI for spinal care that can be implemented in a computing device (200) according to one embodiment of the present invention.

[0159] According to Fig. 10, a computing device (200, mobile app in the drawing) can acquire collected data in real time, such as daily posture habit data and exercise posture habit data, through a posture measurement sensor (smart seat, etc.) and the vision of a camera.

[0160] Additionally, the computing device (200) can evaluate the user's exercise performance results through the acquired collected data, derive a posture pattern, and output an AI-customized motion related to posture correction.

[0161] In some cases, it may include a function to readjust the program as needed. Through this, the user continuously receives a personalized exercise program, and the computing device (200) plays a role in helping the user maintain a correct posture and promote spinal health.

[0162] At this time, the smart seat of the present invention may be used as a sensor for measuring posture, and through this, a computer device can acquire data corresponding to daily posture habit data. Additionally, the computing device (200) can use camera vision to measure the user's exercise posture or body balance and acquire exercise posture data, and then use an artificial intelligence / machine learning module based on the daily posture habit data and exercise posture habit data to predict posture-related data and output it in the form of a UI. Accordingly, the computing device (200) can suggest customized exercises, which will be explained in detail later.

[0163] According to FIG. 11, the computing device (200) first performs a step (S300) of recognizing the user's posture and behavioral patterns through a smart seat. Subsequently, the computing device (200) executes a step (S310) of analyzing and predicting changes in posture and spinal health status based on the collected data. Finally, based on the analysis results, it proceeds to a step (S320) of recommending a customized exercise program to the user using an artificial intelligence (AI) model.

[0164] For example, the computing device (200) collects various data (collected data) as previously mentioned and proposes customized exercises based on this, but the customized exercises may be recommended differently by analyzing exercise performance evaluation and posture improvement relative to the goal. For example, the computing device (200) may recommend a cat pose considering the user's posture, but if the posture is corrected later and the need for it decreases, the recommended cat pose may be excluded from the recommendation.

[0165] Additionally, the computing device (200) can evaluate the user's exercise performance for a recommended exercise (e.g., cat pose), and if the exercise performance is low, it can recommend another exercise (e.g., lying down stretching) through an AI module, etc. Here, the exercise performance evaluation can be determined through the measurement of exercise posture (motion) via a camera, score reports, etc.

[0166] That is, when multiple recommended exercises are distinguished based on the difficulty of the exercise, etc., the computing device (200) can recommend another exercise that is set to be easier to follow when the user's exercise performance level is low, and can recommend the same exercise or another exercise that is set to be more difficult to follow when the exercise performance level is high.

[0167] According to FIG. 12, the computing device (200) performs the function of recognizing posture patterns and predicting posture changes and spinal health based on data collected from sensors, user evaluations, and camera vision. The collected data analyzes patterns over time through a Fourier Transform and can detect posture changes that occur repeatedly.

[0168] First, ① the user's daily posture habits are analyzed through posture pattern recognition machine learning (AI). The results of this analysis can visually output repetitive patterns to provide the user with information such as, for example, "There is an 80% chance that an incorrect posture will occur after 3 PM."

[0169] Next, ② an AI model that predicts changes in posture and spinal health is operated, and the burden applied to the spine due to changes in the user's posture is calculated to predict future health conditions. Through this, the user can identify their posture habits and receive corrective information and customized exercise programs provided by the computing device (200).

[0170] According to FIG. 13, the process of implementing an AI customized exercise recommendation algorithm in a computing device (200) is described. This system operates by analyzing posture patterns, such as changes and improvements in the user's posture, based on data (collected data) such as sensors, user goals, and camera vision, and recommending a customized exercise program based on this.

[0171] First, ① an analysis of the user's posture changes and improvement is conducted. AI analyzes data collected from sensors and camera vision to recognize changes in the user's posture patterns and identify areas requiring improvement. This enables posture correction tailored to the user's individual goals.

[0172] Additionally, ② similarity-based exercise recommendations are made through cluster clustering. The computing device (200) clusters posture data of various users to form groups of users with similar posture problems and recommends effective exercises within those groups. Through this, a personalized exercise program is provided, and exercises suitable for each user are recommended.

[0173] In addition, ③ an exercise performance evaluation is conducted. The computing device (200) evaluates how accurately the user is performing the recommended exercise and analyzes the results to provide feedback on the exercise performance. Through this, the user can check their exercise performance status, and the computing device (200) can reflect this in subsequent exercise recommendations.

[0174] Finally, ④ the computing device (200) generates a customized exercise recommendation list based on the collected data. This process is achieved through clustering based on user data with similar posture patterns, and allows the user to continuously receive appropriate exercises based on the results of the exercise performance.

[0175] According to FIG. 14, one embodiment of the present invention describes a system using an unsupervised machine learning clustering algorithm and an exercise effect evaluation algorithm. This system is designed to provide more sophisticated analysis and feedback in user posture correction and exercise recommendation.

[0176] First, the first part describes user grouping using an unsupervised learning clustering algorithm. A computing device (200) collects various user data and, based on this, forms user groups with similar goals by applying a K-means clustering technique through unsupervised learning. This grouped data can be used for posture correction and recommending exercise programs. For example, users with the same posture correction goal are grouped into a single cluster and recommended an optimized exercise program tailored to them.

[0177] The second part deals with an exercise effect evaluation algorithm. A computing device (200) evaluates the user's exercise performance results by combining existing analysis data (posture improvement, goal), camera vision data, and previously recommended exercise program data. This algorithm scores the exercise performance and evaluates it with a score in the range of [-1, 1]. Through this, the user can quantitatively check how effective the recommended exercise is or what areas need improvement.

[0178] According to FIG. 15, one embodiment of the present invention relates to an algorithm (Factorization Machine) that provides an optimal list of recommended exercises. This algorithm describes a system that provides a customized exercise program with high accuracy by training an AI model based on various data.

[0179] First, existing analysis data (posture improvement, goals, etc.), evaluation scores of recommended exercise programs, and posture cluster data are used as input. These data reflect the user's exercise performance status and goals, helping the AI ​​model recommend the optimal exercise program.

[0180] Secondly, the Factorization Machine algorithm processes multidimensional data to propose the most suitable list of exercises for the user. This algorithm analyzes various variables, including the user's previous exercise performance data, and recommends personalized exercises optimized for the individual user. For example, exercises that received high ratings in a user's previous exercise program are more likely to be recommended again. Conversely, exercises that received low ratings may be excluded from new recommendation lists or suggested in an improved form.

[0181] This algorithm also continuously collects data through feedback on the user's exercise performance, enabling it to provide more sophisticated recommendation lists over time. This maximizes the user's exercise results and provides the ability to continuously improve personalized programs.

[0182]

[0183] Meanwhile, additional embodiments that are not depicted in the drawings will be described.

[0184] In one embodiment, the smart seat (100) can continuously detect whether a user is sitting through a pressure sensor. The computing device (200) can detect if the user is away from the seat for a certain period of time (e.g., 10 minutes or more) and automatically switch the system to a power-saving mode or an off state. When the user returns and sits down, the computing device (200) can detect this through the pressure sensor and automatically turn the system back on.

[0185] For example, after 10 minutes have passed since the user left the seat, the computing device (200) can temporarily suspend all monitoring functions of the smart seat with a message saying, "The user has left the seat. The system is switching to power saving mode." The moment the user returns and sits down, the pressure sensor detects this, and the computing device (200) automatically turns on and starts collecting user data again. This allows for energy saving and efficient management of the system's continuous operation.

[0186] Through these functions, the smart seat (100) can continuously monitor the user's physical condition and provide real-time feedback in various ways, as well as create a healthy and efficient work environment for the user.

[0187]

[0188] For the above, it should be understood that various embodiments of the present invention can be modified in various ways by those skilled in the art, and that various applications and modifications are possible within the essential scope thereof.

Claims

As an AI posture correction system for customized spinal care, A frame part (110) configured to continuously apply upward and inward compressive force for active stabilization support when subjected to pressure by the body, A cushion part (120) equipped with a cushioning member to generate compression resistance when in direct contact with the body, A sensor part (130) located between the cushion part (120) and the frame part (110), wherein a plurality of pressure sensors are distributed in a pressure area receiving pressure from the body, and A smart sheet (100) comprising a control unit (140) that transmits pressure data sensed by the sensor unit (130) to a computing device (200); and An AI posture correction system characterized by including a computing device (200) that calculates a first center point, which is the center point of pressure, based on the above pressure data, calculates a difference value between the first center point and a second center point, which is the center point of the above pressure area, recognizes a posture pattern based on the difference value, and outputs an AI customized motion related to posture correction from the posture pattern. In Article 1, The above computing device is, An AI posture correction system characterized by recognizing a posture pattern through a posture pattern recognition AI algorithm using collected data including the pressure data as input, deriving posture habit information or spinal health information based on the accumulated posture pattern, and outputting the AI ​​customized motion. In Article 2, The above posture habit information is, During the measurement period, distinguish between i) the time spent maintaining an upright posture or ii) the time spent maintaining a skewed posture, and An AI posture correction system characterized by including information on time biased by direction based on the user. In Article 1, The above pressure region includes a central region inside, and The sensing sensitivity of the pressure sensors distributed in the above central region is, It is lower than the sensing sensitivity of pressure sensors distributed outside the central area, and The above pressure region includes a gluteal region and a thigh region internally, and The sensing sensitivity of the pressure sensor distributed in the above thigh region is, An AI posture correction system characterized by having a lower sensing sensitivity than that of pressure sensors distributed in the above-mentioned buttock area. In Article 1, The smart sheet (100) above is, It further includes a haptic part located between the cushion part (120) and the frame part (110) and equipped with a vibration motor within the pressure area, and The control unit (140) operates the vibration motor when the difference value of the second center point exceeds a predetermined threshold point, and The above haptic part has a plurality of vibration motors distributed within the pressure area, and The AI ​​posture correction system is characterized in that the control unit (140) operates the vibration motor closest to the first center point when the difference value of the second center point exceeds a predetermined threshold. In Article 1, The sensor unit (130) is provided with a plurality of pressure sensors on a flexible film portion, and AI posture correction system, characterized in that the film portion is formed with a plurality of grooves (133) that are symmetrical according to the curved shape of the frame portion (110). In Article 6, AI posture correction system characterized by the above film portion being composed of a first film portion (130-1) and a second film portion (130-2) that are symmetrical to each other and separated. A method for correcting posture performed by an AI posture correction system including a computing device (200) and a smart seat (100), The smart sheet (100) above is A frame portion (110) configured to continuously apply upward and inward compressive force for active stabilization support when subjected to pressure by the body; A cushion part (120) equipped with a cushioning member to generate compression resistance when in direct contact with the body; A sensor unit (130) located between the cushion unit (120) and the frame unit (110), wherein a plurality of pressure sensors are distributed in a pressure area receiving pressure from the body; and A control unit (140) that transmits pressure data sensed by the sensor unit (130) to an external computing device (200); Includes, The above method is, (a) A step in which the computing device (200) calculates a first center point, which is the center point of pressure, based on the pressure data; (b) a step in which the computing device (200) calculates the difference value between the first center point and the second center point, which is the center point of the pressure region; and (c) A step in which the computing device (200) recognizes a posture pattern based on the difference value and outputs or supports an action related to posture correction from the posture pattern; A method characterized by including

Images