Motion support system and motion support method

Abstract

To provide a motion support system and a motion support method capable of improving an effect of rehabilitation by motivating a user to actively participate in the rehabilitation.SOLUTION: A motion support system includes: motion detection unit 3 that are attached respectively to a plurality of users 1 and each detect a motion of at least a part of a body of the user 1; a control target 2; and a control unit that controls the control target 2. The control unit controls the control target 2 so as to simulate a motion of a first user 1a selected from the users 1. In addition, the control unit controls the control target 2 so as to simulate a motion of a second user 1b different from the first user 1a when a predetermined period elapses.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to an exercise support system and an exercise support method that encourage users, such as elderly people and people requiring care, to exercise when undergoing rehabilitation. [Background technology]

[0002] It is known that as people age, muscle strength and flexibility decline, and physical balance and walking ability decline. The incidence of dementia also increases with age, and the risk of developing the condition increases particularly for people over 65 years old. Rehabilitation (hereinafter sometimes referred to as "rehab") is an effective countermeasure against age-related muscle loss and dementia.

[0003] Moderate rehabilitation exercise improves physical functions such as balance. Furthermore, cognitive functions such as memory and attention are stimulated through rehabilitation, activating brain function. It is considered especially important for the elderly to gain mental fulfillment by participating in social activities, but declines in physical function often limit participation in these activities. Rehabilitation is seen as enabling the elderly to participate in society by improving physical function, leading to an improvement in their quality of life (QOL). Against this background, the importance of rehabilitation is growing in today's rapidly aging society, and there is a need for practical application of technologies that further enhance its effectiveness.

[0004] As a technology for assisting users in their rehabilitation, for example, an upper limb rehabilitation training system is known that uses a depth camera to capture the user's movements, thereby making the acquired data more accurate and objective and simplifying recording and storage. The upper limb function evaluation device included in this system includes a display, a depth camera, and a central processing unit. The depth camera is for capturing the user's movements, the display is for displaying the pilot movements and the user's movements, and the central processing unit is connected to the display and the depth camera, respectively (Patent Document 1).

[0005] According to Patent Document 1, the central processing unit determines whether the degree of completion of the user's actions meets the requirements in the evaluation table, allowing the user to obtain the results of the evaluation report themselves without receiving significant assistance or cooperation from doctors. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Publication No. 2022-536439 DISCLOSURE OF THE INVENTION [Problem to be solved by the invention]

[0007] In the technology disclosed in Patent Document 1, a pilot's movement is displayed on a display, and the user is guided to imitate the displayed pilot's movement. The user's imitating image is then displayed on the display. However, the user only sees the pilot's movement and their own image on the display, which is considered insufficient to motivate the user to actively engage in rehabilitation.

[0008] The present invention was devised to solve these problems of the prior art, and its purpose is to provide an exercise support system that can motivate users to actively participate in rehabilitation and improve the effectiveness of rehabilitation. [Means for solving the problem]

[0009] The present invention, which has been made to solve the above problems, is an exercise support system comprising: a motion detection unit that is worn by each of a plurality of users and detects the movement of at least a part of the user's body; a control object; and a control unit that controls the control object, wherein the control unit controls the control object based on the output of the motion detection unit worn by the first user selected from the plurality of users so as to imitate the movement of the first user. This allows the selected first user to make an effort to have the control object imitate his or her own movements, thereby improving the effectiveness of rehabilitation.

[0010] Furthermore, in the present invention, when a predetermined period of time has elapsed, the control unit controls the control object based on the output of the movement detection unit worn by the second user so as to imitate the movements of a second user different from the first user. This provides multiple users with an opportunity to operate the control object, improving the motivation of each user and making it possible to improve the effectiveness of rehabilitation.

[0011] In addition, in the present invention, the control unit randomly selects the second user from among the plurality of users, thereby enabling all users participating in rehabilitation to participate in the operation of the controlled object fairly.

[0012] In addition, in the present invention, the control unit causes the control target to perform a predetermined action, and selects the first user or the second user based on the movements of the plurality of users while the control target is performing the predetermined action. This allows each user to desire to be imitated by the control target, thereby making it possible to encourage each user to participate more actively in rehabilitation.

[0013] The present invention also includes a notification unit, and the control unit outputs instruction information instructing the users to move via the notification unit, and selects the user having the highest correlation between the instruction information and the output of the movement detection unit as the second user. This allows each user 1 to obtain the right to operate the control target 2 by making a movement in accordance with the instruction information, thereby allowing each user to participate more actively in rehabilitation.

[0014] Furthermore, in the present invention, the control unit selects the user whose slower or faster movement, or whose smaller or larger displacement is detected by the movement detection unit as the first user or the second user, thereby making it possible to allow each user 1 to participate in rehabilitation fairly even if the degree of disability of each user 1 is different.

[0015] Furthermore, the present invention includes a notification unit, and the control unit notifies the plurality of users of information about the specific user whose movements are imitated by the controlled object via the notification unit. This makes the users who are notified of their name or nickname even more eager to hear their own name or nickname uttered by the controlled object, and also stimulates the competitive spirit of other users 1, making it possible to encourage each user to participate more actively in rehabilitation.

[0016] The present invention also provides an exercise support method in which a motion detection unit attached to each of a plurality of users detects the motion of at least a part of the body of the users, and a control target is controlled based on the motion of at least a part of the body of a first user selected from the plurality of users so as to imitate the motion of the first user, thereby enabling the selected first user to make an effort to make the control target imitate his or her own motion, thereby improving the effect of rehabilitation.

[0017] Furthermore, the present invention controls the control object based on the movement of at least a part of the body of a second user different from the first user after a predetermined period of time has elapsed, so as to imitate the movement of the second user. This provides multiple users with an opportunity to operate the control object, improving the motivation of each user and making it possible to improve the effectiveness of rehabilitation. [Effects of the Invention]

[0018] In this way, according to the present invention, the user is motivated to actively participate in rehabilitation, and the effect of rehabilitation can be improved. [Brief explanation of the drawings]

[0019] [Figure 1] 1A and 1B are explanatory diagrams showing a usage mode of an exercise support system S1 according to a first embodiment of the present invention. [Figure 2] FIG. 1 is a diagram showing the configuration of an exercise support system S1 according to a first embodiment of the present invention. [Figure 3] 1A and 1B are diagrams showing the configuration of the motion detection unit 3. [Figure 4] 1 is a flowchart showing the operation of the exercise support system S1 according to the first embodiment of the present invention. [Figure 5] FIG. 1 is a diagram showing the configuration of an exercise support system S1 according to a second embodiment of the present invention. [Figure 6] FIG. 10 is a diagram showing the configuration of an exercise support system S1 according to a fourth embodiment of the present invention. [Figure 7] 10 is a flowchart showing the operation of an exercise support system S1 according to a fifth embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0020] (First embodiment) A first embodiment of the present invention will be described below with reference to the drawings. FIGS. 1(A) and 1(B) are explanatory diagrams showing how an exercise support system S1 according to the first embodiment of the present invention is used. The exercise support system S1 is installed in a rehabilitation department of a hospital, a nursing home, or the like. In FIG. 1, reference numeral 1 denotes a user who receives rehabilitation using the exercise support system S1 under the supervision of a doctor, physical therapist, or the like (hereinafter sometimes referred to as "doctor, etc."). User 1 may be, for example, an elderly person or a person requiring care.

[0021] Reference numeral 2 denotes a control target in the exercise support system S1, which in this case is, for example, a robot or a robot toy. By making the control target 2 a tangible robot, the user 1 develops a sense of familiarity with the control target 2, which makes it possible to improve the effectiveness of rehabilitation. In the following description, the side of the control target 2 as seen visually in FIG. 1 may be referred to as the front, the opposite direction as the rear, the direction of the right arm 2AR as the right, the direction of the left arm 2AL as the left, the direction of the head 2HD as the up, and the opposite direction as the down.

[0022] A box-shaped right arm motion detection unit 3AR is attached to the wrist of the right arm 1AR of the user 1 using a wristband or the like. On the other hand, for the left arm 1AL, the motion detection unit 3 is configured, for example, in the shape of a stick, which is held by the user 1. Of course, the left arm motion detection unit 3AL may also be box-shaped and attached to the left arm 1AL using a wristband or the like.

[0023] Additionally, box-shaped motion detection units 3 (right leg motion detection unit 3LR, left leg motion detection unit 3LL) are attached to the left and right ankles of the user 1 via wristbands or the like. From the viewpoint of detecting the motion of the arm 1A and leg 1L with high accuracy, it is preferable that the motion detection units 3 are attached to parts of the body with a wide range of movement. From this viewpoint, it is preferable that the motion detection unit 3 that detects the motion of the arm 1A is attached to the wrist or held in the palm of the hand. Furthermore, it is preferable that the motion detection unit 3 corresponding to the leg 1L is attached to the ankle. On the other hand, the motion detection units 3 may be attached to, for example, the upper arm or thigh, as long as the detection accuracy is not impaired.

[0024] Here, the correspondence between each motion detection unit 3 and the body part where it is worn may be determined in advance. That is, in this embodiment, the body part where it should be worn is clearly indicated on the motion detection unit 3, such as "for arm (right)." Then, a doctor or the like instructs the user 1 to wear, for example, the right arm motion detection unit 3AR on the wrist of the right arm 1AR, and similarly wear the other motion detection units 3 on corresponding body parts. On the other hand, the correspondence between each motion detection unit 3 and the body part does not have to be determined in advance. In this case, the correspondence between each motion detection unit 3 and the body part where it is worn is acquired as described below.

[0025] Of course, depending on the content of rehabilitation, the movement detection unit 3 may be attached to, for example, either the arm 1A or leg 1L of the user 1, or either the right arm 1AR or left arm 1AL, or either the right leg 1LR or left leg 1LL. This makes it possible to detect movement limited to the parts of the body that the user 1 actually moves, depending on the content of rehabilitation.

[0026] The control target 2 is composed of a main body 2BD, a head 2HD, arms 2A (right arm 2AR, left arm 2AL), and legs 2L (right leg 2LR, left leg 2LL). Joints (not shown) are provided at the parts of the arms 2A corresponding to the shoulders 2p, elbows 2q, and wrists, and at the parts of the legs 2L corresponding to the parts that engage with the main body 2BD (base 2s of legs 2L), knees 2t, and ankles 2u. Motors are arranged near the joints, and the arms 2A and legs 2L are rotatable within a predetermined range around the joints.

[0027] FIG. 1(A) shows a state in which user 1 and control target 2 are standing motionless (so-called "at attention" posture). For example, if user 1 raises his right arm 1AR straight up from this position, this movement is detected by the movement detection unit 3 (here, right arm movement detection unit 3AR), and control target 2 raises its right arm 2AR straight up as shown in FIG. 1(B). Of course, if user 1 raises his right arm 1AR to shoulder height, control target 2 also raises its right arm 2AR to shoulder height 2p. Similarly, if user 1 moves his left arm 1AL, right leg 1LR, or left leg 1LL, the corresponding parts of control target 2 are displaced in accordance with the movement of user 1.

[0028] In this way, the controlled object 2 imitates (reproduces) the movements of the user 1 based on the output of the movement detection unit 3. In the following description, the movement of the controlled object 2 (movement imitating the movements of the user 1) based on the output of the movement detection unit 3 ("user movement information" to be described later) may be referred to as the "second movement mode." The controlled object 2 is also capable of reproducing the movement of swinging the arm 1A forward from a position where it is attached to the body (forward swing) and the movement of swinging the arm 1A backward and rotating it (backward swing). Furthermore, the movement detection unit 3 may be attached to the upper arm or thigh in addition to the wrist and ankle as described above. By increasing the number of parts for detecting movement, the controlled object 2 can imitate more complex movements.

[0029] On the other hand, the control target 2 is also capable of moving autonomously, regardless of the movement of the user 1. In the following description, the autonomous movement of the control target 2 (movement based on a "movement pattern" to be described later) may be referred to as a "first movement mode." In this way, in the first embodiment, a first movement mode in which the control target 2 moves autonomously and a second movement mode in which the control target 2 moves to imitate (reproduce) the movement of the user 1 based on the output of the movement detection unit 3 are set. The first movement mode and the second movement mode are switched over as the rehabilitation progresses.

[0030] Furthermore, the control target 2 is provided with a first imaging unit 13, a first notification unit 14, a second notification unit 15, a distance measurement unit 16, a switch unit 17, and a first environment sensor 21 on the main body 2BD or the head 2HD. Of these, the first notification unit 14, the second notification unit 15, and the switch unit 17 function as a so-called user interface.

[0031] FIG. 2 is a configuration diagram showing the configuration of an exercise support system S1 according to a first embodiment of the present invention. The configuration and operation of the exercise support system S1 will be described in detail below, using FIG. 2 in conjunction with FIG. 1. As shown in the figure, the exercise support system S1 is made up of a motion detection unit 3 and a control target 2. The motion detection unit 3 is made up of a right arm motion detection unit 3AR, a left arm motion detection unit 3AL, a right leg motion detection unit 3LR, and a left leg motion detection unit 3LL, which are worn by the user 1 as described above. In the first embodiment, it is assumed that there is a one-to-one correspondence between the user 1 and the control target 2 (i.e., one control target 2 is assigned to one user 1).

[0032] The controlled object 2 is composed of a first control unit 10, a first memory unit 11, a first communication unit 12, a first imaging unit 13, a first notification unit 14, a second notification unit 15, a switch unit 17, a first environmental sensor 21, a right leg drive unit 2LRa, a right leg motor 2LRb, a right leg 2LR, a right leg position detection unit 2LRc, a left leg drive unit 2LLa, a left leg motor 2LLb, a left leg 2LL, a left leg position detection unit 2LLc, a right arm drive unit 2ARa, a right arm motor 2ARb, a right arm 2AR, a right arm position detection unit 2ARc, a left arm drive unit 2ALa, a left arm motor 2ALb, a left arm 2AL, and a left arm position detection unit 2ALc. Here, each motor, such as the right leg motor 2LRb, is composed of, for example, a DC servo motor.

[0033] The first control unit 10 is composed of a CPU (Central Processing Unit) and the like, and operates according to a control program stored in a first storage unit 11 which is composed of a ROM (Read Only Memory), RAM (Random Access Memory), and the like. The first control unit 10 and other components are connected via a bus 20 and the like, and the first control unit 10 controls the other components via the bus 20 and the like. The first control unit 10 is provided with a timer (not shown) for measuring time.

[0034] The first storage unit 11 also includes a non-volatile memory (such as an EEPROM (Electrically Erasable Programmable Read-Only Memory)). This non-volatile memory stores a rehabilitation plan for the user 1. The rehabilitation plan includes a plurality of movement patterns to be referenced when operating the control target 2, and a plurality of target patterns as targets for the movement in rehabilitation. Different movement patterns and target patterns can be set for each individual user 1 based on the judgment of a doctor or the like.

[0035] The rehabilitation plan also includes a target time for the user 1 to exercise in one rehabilitation session (i.e., rehabilitation duration; hereinafter, this may be referred to as a "rehabilitation goal index"). Note that the target amount of exercise for the user 1 when undergoing rehabilitation may also be used as the rehabilitation goal index.

[0036] Here, the movement pattern includes, for example, items (parameters) related to the movement speed, movement cycle (rhythm), and displacement range (displacement angle) of the part (e.g., arm 2A) that is the movement target of the control target 2. The target pattern includes, for example, items (parameters) related to the movement speed, movement cycle (rhythm), and displacement range (displacement angle) of the part that is the movement target of the user 1.

[0037] Furthermore, the movement patterns include information such as sounds, voices, images, and videos provided to the user 1 via the first notification unit 14 and the second notification unit 15. The first control unit 10 combines a plurality of movement patterns, or changes or modifies a part of the movement patterns as necessary, to control the arms 2A, legs 2L, the first notification unit 14, and the second notification unit 15. Similarly, the target pattern may also be a combination of a plurality of patterns, or may be changed or modified depending on the environment in which the user 1 is undergoing rehabilitation.

[0038] The non-volatile memory also stores reference biometric indicators of the user 1 determined by a doctor or the like, personal information of the user 1 such as the gender, age, name, nickname, etc., identifiers (IDs) of each movement detection unit 3, text, video, and audio content provided to the user 1, etc. Here, the reference biometric indicators are indices set in advance for each user 1, and refer to the ranges of values ​​of the heart rate, respiratory rate, blood oxygen concentration, blood lactate concentration, etc. when the user 1 undergoes rehabilitation while maintaining an appropriate exercise load under the guidance of a doctor or the like, i.e., the appropriate ranges of these values. Note that the reference biometric indicators may be substituted with standard ranges of values ​​based on the gender, age, exercise load, etc. of the user 1, for example.

[0039] Furthermore, for each ID, a correspondence relationship between each body part of the user 1 (right arm 1AR, left arm 1AL, right leg 1LR, left leg 1LL) and each movement detection unit 3 is stored. Note that the correspondence relationship between each ID and each body part of the user 1 can be changed.

[0040] The first communication unit 12 includes a communication module (not shown) that complies with a short-range wireless standard, such as BLE (Bluetooth (registered trademark) Low Energy). The first communication unit 12 acquires an ID and movement information (hereinafter referred to as "user movement information") from each movement detection unit 3, and outputs this to the first control unit 10. Furthermore, the first communication unit 12 acquires information on the heart rate of the user 1 and the like (hereinafter referred to as "measured biological information") from some of the movement detection units 3, and outputs this to the first control unit 10.

[0041] The first imaging unit 13 includes an image sensor configured with a CMOS (Complementary Metal Oxide Semiconductor) or a CCD (Charge Coupled Device). The distance measuring unit 16 detects objects (including the user 1) around the controlled object 2 by using a ToF (Time of Flight) that measures the time it takes for light to travel from when it is emitted until the reflected light returns, or a LiDAR (Light Detection and Ranging) that performs optical scanning over a wide area using a pulsed laser or the like. As shown in the figure, the first imaging unit 13 and the distance measuring unit 16 are provided in front of the controlled object 2.

[0042] The control target 2 can move forward and backward and change its posture left and right and up and down by driving the legs 2L. That is, the control target 2 can change its imaging range by moving forward and backward or by panning and tilting. As a result, even if ToF is used as the distance measurement unit 16, it can detect the placement and surface condition of objects, etc. in the three-dimensional space in which the control target 2 is placed, essentially in the same way as LiDAR (i.e., it functions as a 3D-ToF). Based on the image data captured by the first imaging unit 13 or the 3D-ToF image obtained by the distance measurement unit 16, the first control unit 10 can recognize the relative positional relationship between the user 1 and the control target 2 (for example, whether they are facing each other directly).

[0043] The first notification unit 14 includes, for example, an LED (Light Emitting Diode). The LED is arranged in a position on the control target 2 that corresponds to a human eye. The first control unit 10 changes the lighting pattern of the first notification unit 14 to indicate to the user 1 (or a doctor, etc.), for example, whether the control target 2 is operating in the first operation mode or the second operation mode. Furthermore, the first notification unit 14 includes a speaker, and audio information is provided to the user 1 via the speaker.

[0044] The second notification unit 15 is configured with, for example, an LCD (Liquid Crystal Display) or an OLED (Organic Light Emitting Diode). For example, text information or an avatar (described later) is displayed on the second notification unit 15. The switch unit 17 is a push switch (toggle switch) placed on the top of the head of the control target 2. When the switch unit 17 is pressed, the control target 2 starts / stops operation. Note that the control mode of the control target 2 may be changed depending on the state of pressing the switch unit 17 (for example, long press or short press).

[0045] The first inertial sensor 18 is configured, for example, with a triaxial acceleration sensor and / or a gyro sensor. Here, the triaxial acceleration sensor outputs the direction and degree of speed change of the control target 2 (acceleration) for the three axes X, Y, and Z. The gyro sensor outputs the direction and speed at which the control target 2 is rotating (angular velocity) for the three axes X, Y, and Z. Generally, the gyro sensor detects the direction of movement, and the acceleration sensor detects the distance of movement. In this way, the first inertial sensor 18 outputs triaxial acceleration information and triaxial angular velocity information based on the movement of the control target 2.

[0046] The first control unit 10 detects the posture of the controlled object 2 based on the output of the first inertial sensor 18. Then, in a situation where the controlled object 2 moves forward or backward (walks), or stands on one leg, the first control unit 10 controls the controlled object 2 to ensure its balance and prevent it from falling over.

[0047] The first environmental sensor 21 may include, for example, a temperature and humidity sensor that measures the temperature (air temperature) and humidity of the external environment of the controlled object 2, a microphone that measures sound, an illuminance sensor that detects illuminance, a human presence sensor that detects infrared rays emitted by a person or the like, and an odor sensor that detects odors (or a gas sensor that detects a predetermined gas) (all of which are not shown). Of course, the controlled object 2 does not necessarily include all of these sensors. The output of the first environmental sensor 21 is converted into digital data by, for example, an A / D converter included in the first control unit 10.

[0048] The first control unit 10 recognizes the voice of the user 1 or the doctor, etc., acquired via a microphone. In the voice recognition, "phonemes" are extracted from the acquired voice information, converted into text, and then vocabulary information is identified. Based on this vocabulary information, for example, the start / stop of the control target 2 may be controlled. Furthermore, the first control unit 10 may perform, for example, frequency analysis on the voice information to distinguish at least between the user 1 and the doctor, etc., as speakers. Then, it may determine a priority for accepting instructions depending on the speaker (for example, giving priority to instructions from the doctor, etc.). Furthermore, it may change the movement pattern or target pattern based on instructions from the doctor, etc., for example.

[0049] The first control unit 10 may also select or change the movement pattern or target pattern based on the temperature and humidity of the external environment. Specifically, in an environment where heatstroke is a risk, the parameters of the target pattern may be modified to correspond to gentler exercise. Furthermore, the first control unit 10 may be controlled not to activate the controlled object 2 if the environment is dark and unsuitable for rehabilitation based on the output of the illuminance sensor, if the number of people detected by the human presence sensor is one (i.e., if no doctor or other person is present), or if the carbon dioxide concentration in the air is determined to be higher than a predetermined value based on the output of the odor sensor.

[0050] The following describes the components related to the movement of the controlled object 2. For simplicity, the following description will be given using the right leg 2LR as an example. The first control unit 10 determines control targets (driving direction, target position, target speed) based on the above-mentioned movement pattern, and outputs a control command to the right leg driving unit 2LRa. The right leg driving unit 2LRa determines driving conditions (for example, the ON duty ratio of PWM (Pulse Width Modulation) control) based on the control command. The right leg driving unit 2LRa includes a motor driver, and the motor driver supplies power to the right leg motor 2LRb based on the driving conditions. The right leg 2LR then moves (rotates) due to the driving force generated by the right leg motor 2LRb.

[0051] Here, a right leg position detector 2LRc is mechanically engaged with the right leg motor 2LRb. The right leg position detector 2LRc is composed of a so-called rotary encoder, which outputs a pulse signal as it rotates. Based on this pulse signal, the displacement speed and position of the right leg 2LR are detected. The rotary encoder may be either an incremental or absolute type, but in the case of an incremental type, an origin sensor is further provided to detect the origin position of the right leg 2LR.

[0052] The output of the right leg position detector 2LRc is input to the first controller 10. The first controller 10 measures the position of the right leg motor 2LRb (i.e., the right leg 2LR) based on the number of pulse signals output from the rotary encoder, and further measures the rotation speed of the right leg motor 2LRb (i.e., the moving speed of the right leg 2LR) based on the period of the pulse signals (e.g., rising edges). These are then compared with control target values ​​to perform, for example, PID (Proportional-Integral-Differential) control (feedback control). As a result, the right leg 2LR is displaced (rotated) in the specified drive direction to the target position at the target speed.

[0053] The left leg 2LL, right arm 2AR, and left arm 2AL are controlled in the same manner as the right leg 2LR. In the above description of the right leg 2LR, it was stated that the right leg 2LR is provided with a single right leg drive unit 2LRa, right leg motor 2LRb, and right leg position detector 2LRc. However, in reality, drive units, motors, and position detectors are provided at the shoulders 2p and elbows 2q of the arms 2A of the controlled object 2, and at the engagement portions of the legs 2L between the main body 2BD and the legs 2L (the bases 2s of the legs 2L), knees 2t, and ankles 2u, respectively, and the first control unit 10 controls all of these. Furthermore, the ankle 2u is provided with a mechanism (a swing mechanism including a joint and a balance control motor (not shown)) for maintaining left-right balance when the controlled object 2 is moved forward, etc., and the first control unit 10 drives the balance control motor based on the output of the first inertial sensor 18.

[0054] 3(A) and (B) are diagrams showing the configuration of the motion detection unit 3. Here, FIG. 3(A) shows the configuration of at least one of the right arm motion detection unit 3AR and the left arm motion detection unit 3AL. Also, FIG. 3(B) shows the configuration of the right leg motion detection unit 3LR and the left leg motion detection unit 3LL. Note that either the right arm motion detection unit 3AR or the left arm motion detection unit 3AL may adopt the configuration of FIG. 3(B). Specifically, the box-shaped motion detection unit 3 attached to the right arm 1AR shown in FIG. 1 may have the configuration of FIG. 3(A), and the stick-shaped motion detection unit 3(3AL) held in the left hand may have the configuration of FIG. 3(B).

[0055] 3(A), the motion detection unit 3 is composed of a second control unit 3a, a second storage unit 3b, a second communication unit 3c, a second inertial sensor 3d, and a biological index measurement unit 3e. Note that the biological index measurement unit 3e may be provided in either the right arm motion detection unit 3AR or the left arm motion detection unit 3AL.

[0056] The second control unit 3a is composed of a CPU etc. and operates according to a control program stored in a second storage unit 3b which is composed of a ROM, RAM etc. The second control unit 3a and other components are connected by, for example, a bus, and the first control unit 10 controls the other components via the bus etc. The second storage unit 3b further stores an identifier (ID) which represents each individual motion detection unit 3.

[0057] The second inertial sensor 3d is configured, for example, with a triaxial acceleration sensor and / or a gyro sensor, similar to the first inertial sensor 18 described above. Here, the triaxial acceleration sensor outputs the direction and degree of speed change of the arm 1A, etc. of the user 1 for three axes, i.e., X, Y, and Z. The gyro sensor outputs the direction and speed at which the arm 1A, etc. of the user 1 is rotating for three axes, i.e., X, Y, and Z. In this way, the second inertial sensor 3d detects the movement of the arm 1A and leg 1L of the user 1 and outputs triaxial acceleration information and / or triaxial angular velocity information based on the detected movement (hereinafter, sometimes referred to as "user movement information"). Note that the user movement information (particularly the triaxial acceleration information) can be used to determine the position, displacement amount, displacement speed, and movement vector of the arm 1A and leg 1L of the user 1 in the X, Y, and Z space.

[0058] The biometric measurement unit 3e measures at least one of the heart rate, respiratory rate, blood oxygen concentration, and blood lactate concentration of the user 1. The biometric measurement unit 3e is arranged on the side of the movement detection unit 3 facing the radial artery and is composed of an optical device that emits red light (λ=600nm to 750nm) or near-infrared light (λ=750nm to 1000nm) and a light receiving sensor (neither shown) that is sensitive to these wavelength bands. The heart rate, blood oxygen concentration, and blood lactate concentration can be detected based on the output of the light receiving sensor.

[0059] Furthermore, with regard to the heart rate and respiratory rate, for example, the distance measurement unit 16 or the first image capture unit 13 provided in the control target 2 can also function as the biological index measurement unit 3e. Specifically, the distance measurement unit 16 can detect the movement of the chest of the user 1 to measure the respiratory rate. Alternatively, the first image capture unit 13 may be configured with a stereo camera, and the chest movement may be detected from depth information acquired by the stereo camera to measure the respiratory rate. Alternatively, the pulse wave (heart rate) may be detected based on a moving image (green channel output) of the face, etc. of the user 1 captured by the first image capture unit 13.

[0060] The heart rate and respiratory rate are converted into values ​​per minute (heart rate / min, respiratory rate / min) by the first control unit 10. The exercise load of the user 1 can be estimated based on the measurement results of these biometric indicator measurement units 3e (hereinafter, sometimes referred to as "measured biometric information"). If the measured heart rate / min and respiratory rate / min exceed the above-mentioned reference biometric indicators, the exercise load of the user 1 is determined to be excessive. Furthermore, since the blood oxygen concentration or blood lactate concentration changes depending on the exercise situation, the average value of the blood oxygen concentration or blood lactate concentration over a predetermined period can also be used as measured biometric information.

[0061] The second communication unit 3c includes a communication module (not shown) that complies with a short-range wireless standard, such as BLE. The second control unit 3a acquires the ID and user movement information stored in the second storage unit 3b, and transmits this information to the control target 2 via the second communication unit 3c. Furthermore, the movement detection unit 3 (for example, the above-mentioned right arm movement detection unit 3AR) that includes the biological index measurement unit 3e also transmits measured biological information to the control target 2. This information is received by the first communication unit 12 of the control target 2 and passed to the first control unit 10.

[0062] As described above, in the first embodiment, the motion detection unit 3 includes an inertial sensor (second inertial sensor 3d), and the motion detection unit 3 transmits the output of the inertial sensor to a control unit (first control unit 10 of the control target 2) wirelessly (here, BLE). This eliminates the need for wiring such as wires between the user 1 and the control target 2, allowing the user 1 to freely exercise during rehabilitation.

[0063] The following provides an overview of the processing performed by the exercise support system S1. 1. First, present user 1 with a "movement example" (first operation mode), and motivate him by saying something like, "Come on, I'm going to move too, so move with me."

[0064] 2. The robot as the controlled object 2 stops its movement and waits for the user 1 to start moving. If the user 1 does not start moving, the "model movement" is presented again. Note that the "model movement" here is a trigger that causes the user 1 to start moving, and does not require the user 1 to "perfectly match the model movement."

[0065] 3. When user 1 starts to move, controlled object 2 imitates user 1's movements (second operation mode). At this time, user 1's heart rate, etc. are measured. The second operation continues for a predetermined period. Note that in the second operation mode, if the user's heart rate is monitored and it is determined that the exercise load is excessive, controlled object 2 notifies user 1 to reduce their movements and does not switch to the first operation mode.

[0066] 4. If it is determined that the exercise load of user 1 is appropriate, controlled object 2 selects a movement pattern based on user 1's rehabilitation plan and operates in the first movement mode. At this time, it motivates user 1 by saying, "Try imitating me and moving like me." It also emits a voice such as "One, two, one, two" to encourage user 1 to move in sync with controlled object 2. Note that in the second movement mode, the conditions for switching from the second movement mode to the first movement mode include that the heart rate, etc., is within a predetermined range and stable.

[0067] 5. While operating in the first operating mode, the movement detection unit 3 detects the movement of the user 1 and compares the movement of the user 1 with the target pattern. If there is a discrepancy between the movement of the user 1 and the target pattern (for example, if the user 1 cannot keep up with the movement of the controlled object 2), the movement pattern is changed to a slower movement or a pattern with a smaller displacement range, or the parameters of the movement pattern are modified. Also, if the heart rate, etc. is not within a predetermined range or is determined to be unstable, a movement pattern with a low exercise load is selected to maintain an appropriate exercise load. At this time, the parameters of the target pattern of the user 1 may be changed.

[0068] 6. If it is determined that the user 1's movements are following the selected movement pattern, the movement pattern is selected and corrected to be closer to the target pattern, and the exercise load is gradually increased. At this time, heart rate, etc. are also detected.

[0069] 7. When the movement of user 1 reaches the target pattern, controlled object 2 operates in the second operation mode and imitates the movement of user 1. At this time, the movement of user 1 is detected, and if the movement appears to have slowed down, the controlled object 2 switches to the first operation mode and tells user 1, "You're moving slower. Try a little harder." The heart rate and other parameters of user 1 are also measured in the second operation mode, and if the heart rate and other parameters are not within a predetermined range, a slower pattern is selected as the operation pattern.

[0070] Rehabilitation ends when the patient continues the movements in 8.7 for a predetermined period of time.

[0071] FIG. 4 is a flowchart showing the operation of the exercise support system S1 according to the first embodiment of the present invention. The operation of the exercise support system S1 will be described in detail below, using FIG. 4 in conjunction with FIG. 1, FIG. 2, and FIG. 3. The controlled object 2 is activated by pressing a switch unit 17 provided on the head 2HD, and each component becomes active. Meanwhile, at least when the controlled object 2 is activated, each movement detection unit 3 is activated and transmits user movement information to the controlled object 2. The first control unit 10 then acquires the user movement information and ID transmitted from each movement detection unit 3 via the first communication unit 12.

[0072] First, the first control unit 10 performs initial setting of the exercise support system S1 (ST00). In the following explanation, "ST" means a processing step. In the initial setting, the correspondence between the body parts of the user 1 (right arm 1AR, left arm 1AL, right leg 1LR, left leg 1LL) and each of the motion detection units 3 attached (or held) thereto is acquired.

[0073] In acquiring the correspondence between each body part and each motion detection unit 3, the first control unit 10 guides the user 1 via the first notification unit 14 (here, a speaker), such as "Raise your right arm." Then, the motion detection unit 3 that outputs user motion information with the largest change in the upward direction (z-axis direction) is recognized as the right arm motion detection unit 3AR. The first control unit 10 stores the ID received from this motion detection unit 3 in the first storage unit 11 as the ID of the right arm motion detection unit 3AR. In the same manner, correspondence between the motion detection units 3 and other body parts is acquired and stored in the first storage unit 11. In this embodiment, the exercise support system S1 becomes available at this stage.

[0074] However, in a case where the part of the body to which each movement detection unit 3 is to be attached is specified in advance and the user 1 or a doctor or the like attaches the movement detection unit 3 accordingly, there is no need to detect the correspondence between each body part and each movement detection unit 3. In this case, a predetermined setting (default) is adopted as the correspondence between each body part and each movement detection unit 3. In either case, if measured biological information is not transmitted from the right arm 1AR or left arm 1AL of the user 1, the first control unit 10 notifies the user via the first notification unit 14 or the second notification unit 15 that there is an error in the attachment state of the movement detection unit 3.

[0075] Next, the first control unit 10 presents a "model movement" to the user 1 (ST01). Specifically, the first control unit 10 extracts a predetermined movement pattern from the first storage unit 11, and controls the arms 2A, legs 2L, etc. of the control target 2 based on this movement pattern (first movement mode). Here, a "pattern that serves as a model movement for the user 1" is extracted as the movement pattern. This pattern includes, for example, parameters for causing the legs 2L of the control target 2 to step at a predetermined rhythm, and the control target 2 executes stepping.

[0076] When the first control unit 10 executes the model movement, it motivates the user 1 by emitting a voice such as "Now, try moving like this" via the first notification unit 14. Furthermore, while executing the model movement, the first control unit 10 also emits a voice such as "One, two, one, two" in synchronization with the model movement to make it easier for the user 1 to imitate the movements of the control target 2. In this way, the components of the control target 2 are controlled by the first control unit 10, but hereinafter, for simplicity of explanation, the description of the first control unit 10 as the entity executing the control will be omitted, and the control target 2, etc. may be used as the subject in the explanation.

[0077] After the model movement is performed, the control target 2 stops moving for a predetermined period (e.g., several seconds). At this time, the first control unit 10 prompts the user 1 to start moving by audio, such as "Now, move, and I'll move along with you." The first control unit 10 then determines whether the user 1 has started moving based on the user movement information received from each movement detection unit 3 (ST02). If the user 1 has started moving (Yes in ST02), the first control unit 10 controls the arms 2A and legs 2L of the control target 2 based on the user movement information, and as a result, the control target 2 moves to reproduce (imitate) the movement of the user 1 (second movement mode) (ST03).

[0078] If the user 1 does not start moving (No in ST02), the process proceeds to ST01, where the user 1 is prompted to start moving again. The number of times that the determination in ST02 is No may be counted, and if the number exceeds a predetermined number, the process may proceed to ST22 and end. Furthermore, the first control unit 10 may refer to the result of the above-mentioned voice recognition, select a movement pattern based on instructions from a doctor or the like, and move the control target 2. If the user 1 starts moving accordingly, the process may proceed to ST03.

[0079] As described above, in the first embodiment, the control unit (first control unit 10) operates the control target 2 in the first operation mode before operating it in the second operation mode. This allows the user 1 to be shown a model of movement, and makes it possible to guide the user 1 to move their body.

[0080] Furthermore, in the first embodiment, a notification unit (first notification unit 14) is provided to provide predetermined information to the user 1, and the controlled object 2 notifies the user 1 via the notification unit that the controlled object 2 will reproduce (imitate) the movement of the user 1 prior to operating in the second operation mode. This makes it possible to motivate the user 1 to start moving.

[0081] Now, the model movement executed by the control object 2 prior to the second operation mode is a trigger to start the user 1's movement, and does not require the user 1 to "perfectly match the model movement." In other words, the user 1 is merely asked to start some kind of movement. However, in the second operation mode following the model movement, even if the user 1 has been shown the model movement beforehand, the important point is that the user 1 can "control" the control object 2 at his or her own will. During the rehabilitation process, the user 1 realizes this and actively tries to influence and control the control object 2. In this way, a game-like element is added to the second operation mode, in which the user 1 acts as the controller and controls the control object 2.

[0082] Next, the first control unit 10 determines whether the measured biological information (measured by the biological index measurement unit 3e, as described above) received from some of the movement detection units 3 (here, the right arm movement detection unit 3AR or the left arm movement detection unit 3AL) is within an appropriate range (ST04). As described above, the first storage unit 11 stores appropriate ranges (reference biological indexes) of values ​​related to the user 1's heart rate, respiratory rate, blood oxygen concentration, blood lactate concentration, etc. On the other hand, some of the movement detection units 3 are provided with a biological index measurement unit 3e, and the movement detection unit 3 transmits the measured biological information measured by the biological index measurement unit 3e to the control target 2.

[0083] The first control unit 10 compares the measured biological information with the reference biometrics stored in the first storage unit 11. If the measured biological information is within the appropriate range of the reference biometrics (Yes in ST04), it determines that the exercise load of the user 1 is appropriate, and acquires the target pattern and movement pattern included in the rehabilitation plan from the first storage unit 11 (ST05). At this time, the first control unit 10 starts the timer described above, and starts measuring the time during which the user 1 is undergoing rehabilitation (hereinafter, sometimes referred to as "rehabilitation duration"). Note that the condition for determining Yes in ST04 may be "the measured biological information is within the appropriate range and is stable (the measured biological information is not fluctuating suddenly)."

[0084] As described above, the first embodiment includes a biological index measurement unit 3e that measures predetermined biological indexes of the user 1, and in the second operation mode, the control unit (first control unit 10) switches from the second operation mode to the first operation mode (ST07, which will be described later) based on the output of the biological index measurement unit 3e (when the measured biological information is within the appropriate range). This enables the user 1 to perform movements in accordance with the rehabilitation plan.

[0085] Here, the movement pattern selected in ST05 is a pattern for making the movement of the user 1 closer to the target pattern, and is determined based on the user movement information when the control target 2 is operating in the second movement mode (i.e., when the control target 2 is imitating the movement of the user 1). Here, the movement patterns are set in, for example, 10 stages ranging from movements that pose a low exercise load to the user 1 to movements of the target pattern (movements that pose a high exercise load).

[0086] The movement patterns are set so that the exercise load increases as the level increases from 1 to 10. For example, if the user movement information detected by the movement detection unit 3 is at the second level out of 10 levels, the third level is selected as the movement pattern. That is, the first control unit 10 selects a pattern with an exercise load that is one level higher than the movement pattern that most closely resembles the user movement information. In this way, in ST05, a movement pattern with an exercise load lower than the target pattern, which is the final movement goal of the user 1, is selected. This prevents the risk of a sudden increase in the exercise load of the user 1, allowing the user 1 to carry out rehabilitation without strain.

[0087] Machine learning is preferably used to select the movement pattern. That is, a learning model that learns the relationship between user movement information and movement patterns is constructed in advance based on the judgment of a doctor or the like. Then, in an actual usage scenario, the first control unit 10 inputs the acquired user movement information into the learning model and obtains the movement pattern as its output.

[0088] The output of the first environmental sensor 21 may be reflected in the selection of the movement pattern. Specifically, for example, by referring to the output of a temperature and humidity sensor included in the first environmental sensor 21, if the discomfort index of the environment in which the user 1 is undergoing rehabilitation is high, a movement pattern with a lower exercise load may be selected. In this case as well, machine learning can be suitably used.

[0089] As described above, in the first embodiment, the memory unit (first memory unit 11) stores a target pattern as a target for the movement of the user 1, and in the second operation mode, the control unit (first control unit 10) selects a movement pattern that brings the movement of the user 1 (user movement information) closer to the target pattern based on the output of the movement detection unit 3, and operates the controlled object 2 in the first operation mode based on the selected movement pattern. This makes it possible to select a movement pattern that does not deviate significantly from the movement that the user 1 is imitating, and then transition to the first operation mode.

[0090] On the other hand, if the measured biological information is not within the appropriate range (No in ST04), the first control unit 10 prompts the user 1 to slow down or speed up their movements via the first notification unit 14 (ST06), and proceeds to ST04 after a predetermined period of time has passed. Specifically, the first notification unit 14 (here, a speaker) reproduces a voice message such as "Try moving a little slower" or "Try moving a little faster." Of course, at this time, the notification may be made by displaying text or images on the second notification unit 15 instead of (or together with) the first notification unit 14.

[0091] As described above, the first embodiment includes a bioindicator measurement unit 3e that measures predetermined bioindicators of the user 1, and an alarm unit (first alarm unit 14) that provides predetermined information to the user 1, and in the second operation mode, the control unit (first control unit 10) notifies the user 1 via the alarm unit to suppress or promote the movement of the user 1 based on the output of the bioindicator measurement unit 3e (when the measured bioindicator information is not within the appropriate range). This makes it possible to start rehabilitation from a state in which an appropriate exercise load is applied to the user 1.

[0092] Next, the first control unit 10 controls the arms 2A and legs 2L of the control target 2 based on the selected movement pattern (i.e., in the first movement mode) (ST07). At this stage, the movement of the control target 2 switches from the "second movement mode" to the "first movement mode." At this time, the control target 2 urges the user 1 to imitate the movement of the control target 2, for example, by saying, "Next, imitate me," and also yells, "One, two, one, two," to make it easier for the user 1 to synchronize with the movement of the control target 2. In this way, it is possible to encourage the user 1 to actively engage in rehabilitation.

[0093] Thus, the first embodiment includes a movement detection unit 3 that detects movement of at least a part of the body of the user 1, a control object 2, a memory unit (first memory unit 11) that stores movement patterns to be executed by the control object 2, and a control unit (first control unit 10) that controls the control object 2, and the control unit switches between a first movement mode in which the control object 2 is operated based on the movement pattern, and a second movement mode in which the control object 2 is operated to reproduce the movement of the user 1 based on the output of the movement detection unit 3.

[0094] In step ST01, in which the above-described example movement (first operation mode) is presented, the user 1 imitates the movement of the control target 2. After the user 1 actually starts moving (Yes in ST02), the control target 2 imitates the movement of the user 1 (ST03 (second operation mode)). After that, the process shifts to the first operation mode based on the movement pattern (ST07). By switching the operation mode in this way, an "interaction" occurs between the user 1 and the control target 2 (exercise support system S1). In particular, in the second operation mode, the control target 2 (a tangible robot) imitates the user 1, which is thought to cause the user 1 to develop a strong attachment to the control target 2. By cultivating this attachment, the user 1 is strongly motivated to imitate (want to imitate) the movement of the control target 2 in the first operation mode, which follows the second operation mode.

[0095] The term "interaction" is generally interpreted as "when a human performs some kind of action (operation or behavior), the system responds accordingly." However, in this specification, "interaction" is understood to mean not only "when a human performs some kind of action (operation or behavior) and the system responds accordingly," but also "when the system performs some kind of action and the human responds accordingly." In other words, the present invention can build a deeper and stronger relationship between the user 1 and the controlled object 2 (exercise support system S1) than a typical "interaction."

[0096] Next, the first control unit 10 compares the rehabilitation duration based on the output of the timer with a rehabilitation goal index included in the rehabilitation plan (here, a target time for the user 1 to exercise in one rehabilitation session), and determines whether the rehabilitation duration has reached the rehabilitation goal index (ST08). If the rehabilitation duration has reached the rehabilitation goal index (No in ST08), it determines that the rehabilitation has ended, and moves the process to ST22, where the operation of the control target 2 is stopped.

[0097] As described above, the motion vector of the body part where the motion detection unit 3 is attached can be obtained from the user motion information. The first control unit 10 can calculate the amount of exercise for each body part of the user 1 from the scalar quantity of the motion vector, and can further calculate the cumulative exercise amount of the user 1 by summing the exercise amounts for each body part and accumulating this total value. The target exercise amount (determined by a doctor, etc.) required of the user 1 during rehabilitation can be set as a rehabilitation goal index, and the end of rehabilitation in ST08 can be determined by comparing this with the calculated cumulative exercise amount. In other words, if the cumulative exercise amount exceeds the rehabilitation goal index, rehabilitation can be considered to have ended, and processing can proceed to ST22.

[0098] If the rehabilitation duration has not reached the rehabilitation goal index (Yes in ST08), it is determined whether the measured biological information of user 1 is within the appropriate range (ST09). The process in ST09 is the same as ST04 described above, and therefore a description thereof will be omitted here.

[0099] If it is determined that the measured biological information of the user 1 falls within the range of the reference biological index (i.e., the exercise load is appropriate) (Yes in ST09), the first control unit 10 determines whether the movement of the user 1 has reached the movement pattern that is the basis of the movement of the control target 2, based on the user movement information received from the movement detection unit 3 (ST10). That is, in ST10, it determines whether the user 1 is appropriately imitating (following) the movement of the control target 2. On the other hand, if it is determined that the measured biological information does not fall within the appropriate range (No in ST09), the first control unit 10 determines whether the exercise load based on the measured biological information is excessive (ST11).

[0100] If the exercise load is excessive (Yes in ST11), the first control unit 10 selects a movement pattern that reduces the exercise load of the user 1 (ST13). Here, the movement pattern that reduces the exercise load refers to, for example, a movement pattern in which the displacement width of the arms 2A or legs 2L is smaller, a movement pattern in which the displacement speed is slower, a movement pattern in which the displacement period is longer, or a movement pattern that is a combination of these, compared to the movement pattern used to control the control target 2 at the time the processing of ST11 is executed.

[0101] Instead of selecting a movement pattern, the parameters that make up the movement pattern may be modified or changed. Also, a pattern with a low exercise load may be selected as the target pattern, and the parameters that make up the target pattern may be modified or changed. This allows rehabilitation to proceed while maintaining an appropriate exercise load on the user 1. Then, the process proceeds to ST07.

[0102] As described above, the first embodiment includes a biomarker measurement unit 3e that measures predetermined biomarkers of the user 1, and a control unit (first control unit 10) selects, modifies, or changes an action pattern or a target pattern based on the output of the biomarker measurement unit 3e. This makes it possible to maintain an appropriate exercise load for the user 1 and reduce the risk of blood pressure rise or fall, angina attacks, arrhythmia, etc.

[0103] On the other hand, if the exercise load is not excessive (No in ST11), the first control unit 10 prompts the user 1 to promote exercise (for example, to increase the swing amplitude of the arms 1A or legs 1L). Then, the process proceeds to ST08 (transition from X to X' shown in the figure).

[0104] If the first control unit 10 determines that the user 1's movement has not reached the "movement pattern" (No in ST10), it notifies the user 1 to encourage exercise, for example, by saying, "Your movements have slowed down. Try a little harder" (ST12), and proceeds to ST08 (transition from X to X' shown in the figure). On the other hand, if it determines that the user 1's movement has reached the "movement pattern" (Yes in ST10), it determines whether the user 1's movement has reached the target pattern (ST14). If the user 1's movement has not reached the target pattern (No in ST14), the first control unit 10 selects a movement pattern that brings the movement pattern closer to the target pattern (ST16). Then, it proceeds to ST07.

[0105] As described above, in the first embodiment, the memory unit (first memory unit 11) stores a target pattern as a target for the movement of the user 1, and in the first movement mode, the control unit (first control unit 10) selects a movement pattern that brings the movement of the user 1 (user movement information) closer to the target pattern based on the output of the movement detection unit 3, and operates the control target 2 in the first movement mode based on the selected movement pattern. This makes it possible to gradually increase the exercise load of the user 1 and improve the effect of rehabilitation.

[0106] When the movement of the user 1 reaches the target pattern (Yes in ST14), the first control unit 10 controls the arms 2A and legs 2L of the control target 2 in the second movement mode (ST17) so as to reproduce (imitate) the movement of the user 1. That is, in this situation, the movement of the control target 2 is switched from the first movement mode to the second movement mode.

[0107] Next, the first control unit 10 determines whether the measured biological information of the user 1 is within the appropriate range (ST18). This determination process is the same as that explained in ST09, so the explanation will be omitted here. If the measured biological information of the user 1 is not within the range of the reference biological index (No in ST18), the first control unit 10 selects the target pattern as the movement pattern (ST19), and proceeds to ST07.

[0108] On the other hand, if the measured biological information of the user 1 is within the range of the reference biological index (Yes in ST18), the first control unit 10 determines whether the movement of the user 1 is maintaining the target pattern based on the user movement information (ST20). If the movement of the user 1 is not maintaining the target pattern (No in ST20), the target pattern is selected as the movement pattern (ST19), and the process proceeds to ST07. If the determinations in ST18 and ST20 are No, the operation of the control target 2 is switched from the "second operation mode" to the "first operation mode" by the process in ST07.

[0109] Next, the first control unit 10 determines whether the rehabilitation duration has reached the rehabilitation goal index (ST21). The processing in ST21 is the same as that in ST08, so a description thereof will be omitted here. If the rehabilitation duration has reached the rehabilitation goal index (Yes in ST21), the first control unit 10 stops the movement of the control target 2 (ST22), and the rehabilitation ends. If the rehabilitation duration has not reached the rehabilitation goal index (No in ST21), the processing proceeds to ST18. Note that the target exercise amount of the user 1 may also be used as the rehabilitation goal index in ST21.

[0110] When switching the operation of the control target 2 from the first operation mode to the second operation mode, or from the second operation mode to the first operation mode, or when providing motivation to the user 1 by voice, the first control unit 10 may extract personal information of the user 1 from the first storage unit 11 and add the name or nickname of the user 1 when speaking. This makes it possible to further deepen the relationship between the user 1 and the control target 2.

[0111] The exercise support system S1 according to the first embodiment has been described in detail above. Based on the above description, the exercise support method according to the present invention is as follows. (1) The system stores an operation pattern to be executed by the controlled object 2, detects the movement of at least a part of the body of the user 1, and switches between a first operation mode in which the controlled object 2 is operated based on the operation pattern, and a second operation mode in which the controlled object 2 is operated to reproduce the movement of the user 1 based on the detected movement of the user 1, at a predetermined timing. (2) A target pattern is stored as a target for the movement of user 1, and in the first operation mode or the second operation mode, the movement of user 1 is compared with the target pattern, and based on the comparison result, an operation pattern that brings the movement of user 1 closer to the target pattern is selected, and based on the selected operation pattern, the controlled object 2 is operated in the first operation mode.

[0112] A modification (first modification) of the first and second operation modes described above will be described below with reference to FIGS. 1 and 2. In the first modification, the first control unit 10 detects the relative positional relationship between the user 1 and the control target 2 based on an image obtained by the first imaging unit 13 or a 3D-ToF image (hereinafter, sometimes referred to as "image, etc.") obtained by the distance measurement unit 16. That is, the first imaging unit 13 or the distance measurement unit 16 functions as a relative position detection unit. Of course, the relative positional relationship between the user 1 and the control target 2 may be detected by referring to the outputs of both the first imaging unit 13 and the distance measurement unit 16.

[0113] The first control unit 10 determines whether the user 1 and the control target 2 are facing each other directly, based on the detection result of the relative position detection unit. In an embodiment in which the first imaging unit 13 or the distance measurement unit 16 is provided in front of the control target 2 (see FIG. 1), if the face included in the image or the like is facing forward, it can be determined that the user 1 and the control target 2 are facing each other directly. In this case, face recognition and recognition of face direction are performed. For face recognition, for example, a method using a Haar feature-based cascade classifier or a method using HOG (Histogram of Oriented Gradients) features can be used. Furthermore, recognition of face direction utilizes the fact that the relationship between feature points such as the eyes, nose, and mouth changes depending on the face direction.

[0114] Another method for determining whether the user 1 and the control target 2 are facing each other directly may be to use a combination of the detection result of the relative position detection unit described above and user movement information. Specifically, in the first operation mode and the second operation mode, the first control unit 10 recognizes whether the user 1 is raising his right arm 1AR or his left arm 1AL based on the output of the movement detection unit 3. At this time, the first control unit 10 also recognizes the head of the user 1 based on the output (the image described above, etc.) of the relative position detection unit, and determines whether the arm 1A is raised on the left or right side of the Y axis, with the Y axis being the direction downward from the head (sub-scanning line direction).

[0115] If it is recognized based on the user movement information that the user 1 is raising his / her right arm 1AR and that the arm 1A on the left side (usually on the origin side in the main scanning direction (X direction)) in the image, etc. is raised, it can be determined that the user 1 and the control target 2 are in a positional relationship of facing each other. Furthermore, if it is recognized based on the user movement information that the user 1 is raising his / her right arm 1AR and that the right arm 1A in the image, etc. is raised, it can be determined that the user 1 is facing away from the front of the control target 2. Similarly, if the user 1 is raising his / her left arm 1AL and the right arm 1A is raised in the image, etc., it can be determined that the user 1 is facing away from the front of the control target 2, and if the user 1 is raising his / her left arm 1AL and the left arm 1A in the image, etc. is raised, it can be determined that the user 1 is facing away from the front of the control target 2.

[0116] Then, when it is determined that the user 1 and the control object 2 are facing each other, the first control unit 10 controls the movement of the control object 2 so that the left and right are reversed (hereinafter, this may be referred to as "left-right reversal control"). That is, in left-right reversal control, for example, when the user 1 raises his right arm 1AR, the control object 2 is controlled to raise his left arm 2AL. When left-right reversal control is performed, the control object 2 appears to the user 1 as a mirror image. This makes the user 1 recognize that "this robot is like my other self," and an even deeper relationship can be built between the user 1 and the control object 2. Note that, in order to more easily achieve this effect, it is preferable that the first control unit 10 notify the user 1 via the first notification unit 14, such as "I'm going to become like a mirror from now on," before operating the control object 2 in the first operation mode or the second operation mode.

[0117] In this way, in the first variant, the relative positional relationship between the control target 2 and the user 1 is detected by a relative position detection unit (imaging unit (first imaging unit 13) or distance measurement unit 16). That is, the positional relationship between the user 1 and the control target 2 is detected based on image data or distance measurement data (3D-ToF image). Then, when the control unit (first control unit 10) determines based on the detection result of the relative position detection unit that the control target 2 and the user 1 are facing each other directly, it reverses the operation of the control target 2 in the first operation mode or the second operation mode left and right.

[0118] A modified example (second modified example) of the first embodiment will be described below with reference to FIGS. 1 and 2. In the first embodiment described above, the control target 2 is a robot, but in the second modified example, the control target 2 is provided with at least a display unit that displays an avatar. The avatar here can also be said to be the avatar of the robot described in the first embodiment. The second modified example may have the same configuration as the first embodiment (see FIGS. 1 and 2), but may also have a configuration in which, for example, the arms 2A and / or legs 2L are removed from the control target 2 (robot) shown in FIG. 1.

[0119] The avatar can be displayed on a second notification unit 15 serving as a display unit (display) arranged on the front of the main body 2BD of the control target 2 (see FIG. 1). By including a display unit that displays the avatar in the control target 2, it becomes possible to improve the effect of rehabilitation by utilizing the exercise support system S1 even if the control target 2 does not have movable parts such as arms 2A and legs 2L.

[0120] Here, avatar morphology information and the like are stored in the first storage unit 11. The avatar morphology can be any character, such as a human, animal, imaginary creature, or robot. The first control unit 10 processes the morphology information acquired from the first storage unit 11 based on the user movement information or movement pattern described above to generate motion data. Then, image data is generated based on the motion data and displayed on the second notification unit 15. Note that the user movement information is three-dimensional information in X, Y, and Z, and motion data may be generated by mapping this information to a predefined articulated model of a character (inverse kinematics). Of course, machine learning or deep learning techniques may be used to generate the motion data.

[0121] The second modification may be combined with the first modification. In this combination, the exercise support system S1 includes a relative position detection unit that detects the relative positional relationship between the control target 2 and the user 1. The control target 2 includes a display unit (second notification unit 15) that displays an avatar. When the control unit (first control unit 10) determines that the control target 2 and the user 1 are facing each other based on the detection result of the relative position detection unit, it displays the avatar in the first operation mode or the second operation mode on the display unit, flipping it horizontally. This method also allows the user 1 to recognize that "this robot is like my other self." Here, the relative position detection unit includes an imaging unit (first imaging unit 13) or a distance measurement unit 16.

[0122] (Second embodiment) A second embodiment of the present invention will be described below with reference to the drawings. The usage of the exercise support system S1 in the second embodiment is the same as that in the first embodiment, so a description thereof will be omitted (see FIG. 1). FIG. 5 is a configuration diagram showing the configuration of the exercise support system S1 according to the second embodiment of the present invention.

[0123] As shown in FIG. 5, the exercise support system S1 includes a control target 2, a motion detection unit 3, a server 30, a second imaging unit 40, a second environmental sensor 41, a third notification unit 42, and an information terminal 43. The control target 2, the server 30, the second imaging unit 40, the second environmental sensor 41, the third notification unit 42, and the information terminal 43 are connected to a network 50, and the control target 2, the server 30, and the information terminal 43 exchange information with each other via the network 50. The outputs of the second environmental sensor 41 and the second imaging unit 40 are transmitted to the control target 2 via the network 50. Image data generated by the control target 2 may be transmitted to the third notification unit 42. Although FIG. 5 shows one control target 2 connected to the server 30 and the information terminal 43, multiple control targets 2 may be connected.

[0124] The configurations of the control target 2 and the motion detection unit 3 are basically the same as those in the first embodiment. However, in the second embodiment, the first communication unit 12 includes a first communication module (not shown) that complies with a wireless communication standard such as LTE (Long Term Evolution), LTE-M (Long Term Evolution - Machine, LTE Cat. M1), 4G, or 5G, and a second communication module (not shown) that complies with a short-range wireless standard such as BLE. Of course, the first communication module may be one that complies with the WiFi (Wireless Fidelity) standard, and the control target 2 may be connected to the network 50 via a wireless router or the like.

[0125] In the second embodiment, the first communication unit 12 acquires user movement information from the movement detection unit 3 in accordance with the BLE data communication protocol, connects to the network 50 in accordance with a communication protocol such as LTE-M, and exchanges information with the server 30, the information terminal 43, etc.

[0126] The server 30 is a known computer system and includes a server control unit 30a and a server storage unit 30b. The server control unit 30a includes a CPU and a storage unit (not shown) and controls the components of the server 30. The server storage unit 30b includes a large-capacity storage device configured with ROM, RAM, and a redundant array of independent disks (RAID) or the like. This large-capacity storage stores, for each of multiple users 1, their unique identifiers (user IDs), rehabilitation plans (including the above-mentioned movement patterns, target patterns, rehabilitation goal indicators, target exercise amounts, etc.), reference biometric indicators, and personal information such as gender, age, name, and nickname (hereinafter, collectively referred to as "user information"). The user information is managed as a user database. Also, avatar morphology information, etc., described in the second modification, is stored in the server storage unit 30b and transmitted to the control target 2 in response to a request from the first control unit 10.

[0127] The user database is provided with a rehabilitation history field that stores the rehabilitation history of each user 1. The first control unit 10 transmits the user ID of the user 1 undergoing rehabilitation, along with the movement pattern used in the first movement mode, the target pattern, user movement information measured in the first movement mode and the second movement mode, and measured biological information to the server 30, and the server control unit 30a stores this in the rehabilitation history field.

[0128] The information stored in the rehabilitation history field is subject to analysis based on instructions from a doctor or other person. When a doctor or other person inputs a predetermined command into the information terminal 43, the command is sent to the server 30, and the server control unit 30a executes various statistical and analytical processes by referring to the rehabilitation history field. The results of these processes are sent to the information terminal 43, allowing the doctor or other person to check the effects of rehabilitation for each user 1 and to formulate a new rehabilitation plan.

[0129] The information terminal 43 is, for example, a mobile information terminal such as a smartphone or tablet terminal, or a PC (Personal Computer), and is equipped with a display unit, an input unit, etc. (not shown). The information terminal 43 is used, for example, by a doctor or the like. The doctor or the like operates the information terminal 43 to send an identifier (user ID) of the user 1 undergoing rehabilitation to the server 30. The server control unit 30a searches the user database based on the user ID. Then, it extracts user information corresponding to the user ID and sends it to the control target 2. The user information is passed to the first control unit 10 via the first communication unit 12. The first control unit 10 stores the received user information in the first storage unit 11 (the non-volatile memory described above). The first control unit 10 then controls the control target 2 by referring to the user information.

[0130] That is, based on the movement pattern, target pattern, and reference biometric indicators included in the user information, the controlled object 2 operates in the first and second operation modes described above. The controlled object 2 also uses personal information such as the name and nickname of the user 1 when communicating with the user 1.

[0131] Furthermore, the doctor or the like can operate the information terminal 43 to send predetermined control commands to the control target 2. The control commands include, for example, a command to make the first notification unit 14 generate a specific sound, and a command to make the control target 2 perform special movements such as moving forward, backward, rolling forward, rolling backward, standing on one leg, and squatting. This allows the doctor or the like to present various exercise menus to the user 1.

[0132] The second imaging unit 40 includes an image sensor configured with a CMOS or CCD. The second imaging unit 40 corresponds to, for example, a web camera placed in a room where the user 1 receives rehabilitation. The image captured by the second imaging unit 40 is transmitted to the control target 2 via a network 50. The first control unit 10 extracts the user 1 and the control target 2 from the image and detects the relative positional relationship between them. That is, the second imaging unit 40 functions as the relative position detection unit described in the first embodiment, and the first control unit 10 determines whether the two are facing each other based on the output of the relative position detection unit.

[0133] Specifically, the first control unit 10 first performs face recognition on the acquired image. Next, it recognizes the orientation of the user 1's face. Here, the orientation of the face is estimated by analyzing the positions and orientations of feature points such as the eyes, nose, and mouth. On the other hand, for the control target 2, for example, the area where two LEDs included in the first notification unit 14 are located is recognized as the face of the control target 2. Then, the direction and distance of the line segment connecting the two LEDs are measured, and the orientation of the control target 2 is estimated. Based on these estimation results, it is determined whether the user 1 and the control target 2 are facing each other. Then, if it is determined that the two are facing each other, it may perform the left-right reversal control described as the first modified example.

[0134] The second environmental sensor 41 may include, for example, a temperature and humidity sensor that measures the temperature (air temperature) and humidity of the space where rehabilitation is performed, a microphone that measures sound, an illuminance sensor that detects illuminance, a human presence sensor that detects infrared rays emitted by people, etc., and an odor sensor that detects odors (which may be a gas sensor that detects a predetermined gas) (all of which are not shown). The first control unit 10 receives the output of the second environmental sensor 41 via the network 50, and can select a movement pattern or a target pattern by referring to, for example, the temperature and humidity.

[0135] The third notification unit 42 is configured with, for example, an LCD (Liquid Crystal Display) or an OLED (Organic Light Emitting Diode), and displays text information and an avatar. The third notification unit 42 may also include audio equipment such as a speaker or an amplifier. The first control unit 10 may transmit image data based on the motion data described in the second modified example to the third notification unit 42 via the network 50, and an avatar may be displayed on the third notification unit 42. At this time, audio data may be transmitted in a similar manner, and audio may be played on the third notification unit 42 instead of the first notification unit 14 of the control target 2.

[0136] As described above, in the first embodiment, the controlled object 2 is mainly used as a partner (personal assistant) of the user 1, and there is basically a one-to-one correspondence between the user 1 and the controlled object 2. In contrast, in the second embodiment, one controlled object 2 can be shared by multiple users 1. That is, the second embodiment is an extension of the first embodiment, from the viewpoint of eliminating the limit on the number of users 1 who use the controlled object 2.

[0137] (Third embodiment) A third embodiment of the present invention will be described below with reference to Figs. 1 and 5. As described above, in the first and second embodiments, the controlled object 2 is controlled to imitate the movements of the user 1 in the second operation mode. The third embodiment extends the second operation mode described above, and a predetermined movement (first additional movement) is added to the imitated movement in the second operation mode. In the following description, the controlled object 2 is, for example, a robot (including a robot toy).

[0138] First, the first control unit 10 determines whether the user 1 is performing a predetermined movement based on the output of the movement detection unit 3. Here, the predetermined movement is, for example, stepping in a fixed position. The first control unit 10 references the outputs of the right leg movement detection unit 3LR and the left leg movement detection unit 3LL to determine whether the right leg 1LR and the left leg 1LL of the user 1 are moving up and down alternately and periodically. If this condition is met, the first control unit 10 determines that the user 1 is stepping.

[0139] At this stage, the controlled object 2 is already simply imitating the movement of the user 1 and performing a stepping motion. While causing the controlled object 2 to perform a stepping motion, the first control unit 10 also controls the right leg drive unit 2LRa and the left leg drive unit 2LLa to tilt the main body 2BD forward relative to the leg 2L. This causes the center of gravity of the controlled object 2 to move forward, and the controlled object 2 begins a walking motion and moves forward. Of course, the main body 2BD may also be tilted backward relative to the leg 2L, in which case the center of gravity of the controlled object 2 moves backward, and the controlled object 2 moves backward. This center of gravity movement motion is performed regardless of the output of the motion detection unit 3.

[0140] At this time, the period or displacement of the stepping motion is determined (imitative motion) based on the output of the motion detection unit 3. When the user 1 performs a stepping motion without changing his / her standing position, the output of the motion detection unit 3 does not include information that can be used to move (forward or backward) the controlled object 2 (i.e., when the user motion information is averaged over a predetermined period, the information in the planar direction (X-axis and Y-axis directions) does not change substantially), and the controlled object 2 does not move simply by performing the imitation motion.

[0141] In the third embodiment, the control object 2 is caused to move its center of gravity regardless of the output of the motion detection unit 3, and further the legs 2L of the control object 2 are displaced based on information on the up and down direction (Z-axis direction) among the output of the motion detection unit 3, thereby moving the control object 2. At this time, the displacement amount (movement amount, movement speed) of the control object 2 is determined substantially based on the output of the motion detection unit 3. That is, the control unit (first control unit 10) causes the control object 2 to perform a stepping motion as an imitation motion, and also moves the control object 2 in a predetermined direction (for example, forward) by a displacement amount based on the output of the motion detection unit 3 as a first additional motion.

[0142] The third embodiment can be effectively utilized in situations where a goal to be achieved is given to the user 1 during rehabilitation. That is, the doctor or the like informs the user 1 that the control target 2 (robot) will move forward when the user 1 steps, and further instructs the user 1 to move the control target 2 to a goal set, for example, 1 meter ahead from its current position. That is, the doctor or the like presents the user 1 with a rehabilitation task. Note that conditions for determining whether or not the goal has been reached (hereinafter, sometimes referred to as "goal attainment conditions"), such as position information of the goal or distance information to the goal, are transmitted and set to the first control unit 10, for example, by the doctor or the like operating the information terminal 43. The first control unit 10 determines whether or not the control target 2 has reached the goal by comparing the output of the first inertial sensor 18 with the goal attainment conditions. Note that the above-mentioned target momentum may be used as the goal attainment condition.

[0143] When the controlled object 2 performs a stepping motion, the motion cycle and displacement amount of the leg 2L (for example, the cycle and displacement amount of the forward and backward swing motion) are determined based on the output of the motion detection unit 3, and therefore the movement speed of the controlled object 2 changes depending on the situation in which the user 1 steps. For example, the movement amount of the controlled object 2 increases when the user 1 shortens the stepping cycle (moves the leg 1L faster) or increases the displacement amount of the leg 1L when stepping (moves the leg 1L more widely).

[0144] As described above, the third embodiment includes a motion detection unit 3 that detects the motion of at least a part of the body of the user 1, a control target 2, and a control unit (first control unit 10) that controls the control target 2. The control unit causes the control target 2 to perform an imitation motion that imitates the motion of the user 1 based on the output of the motion detection unit 3, and simultaneously controls the control target 2 (here, by shifting the center of gravity) regardless of the output of the motion detection unit 3 to cause the control target 2 to perform a predetermined first additional motion (forward or backward). This allows the user 1 to try to make the control target 2 reach the goal as if it were a game. In other words, the user 1 is motivated to actively participate in rehabilitation, which can improve the effectiveness of rehabilitation.

[0145] The relationship between the output of the motion detection unit 3 and the stepping period and displacement amount (movement speed) of the control target 2 can be determined as appropriate. In other words, if the user 1 has a physical disability, a doctor or the like can issue a predetermined command from outside (information terminal 43) and change the correspondence relationship between the output of the motion detection unit 3 and the displacement amount (movement amount, movement speed) of the control target 2 based on the degree of the individual disability of the user 1.

[0146] As a result, for example, if the user 1 has a disability that makes it difficult to move the leg 1L significantly, the displacement amount of the leg 2L of the controlled object 2 can be increased in response to small movements of the leg 1L of the user 1, resulting in a larger movement amount of the controlled object 2. Conversely, if the user 1 does not have a disability in the leg 1L, the displacement amount of the leg 2L of the controlled object 2 can be reduced in response to large movements of the leg 1L of the user 1, resulting in a smaller movement amount of the controlled object 2. Note that the movement amount here refers to, for example, the movement amount per one forward and backward reciprocation of the leg 2L of the controlled object 2, and if the period for displacing the leg 2L is the same, the movement speed of the controlled object 2 can be increased by increasing the movement amount.

[0147] Furthermore, it is also extremely effective for rehabilitation if multiple users 1 each have the control target 2 perform the imitation and first additional actions, and compete to see which of their controlled control targets 2 reaches the goal first. By introducing the concept of competition into rehabilitation in this way, users 1 can be strongly motivated to achieve their goals. In addition, in competition, it is very important to be able to adjust the relationship between the output of the motion detection unit 3 and the movement amount (movement speed) of the control target 2, as mentioned above. In other words, by imposing conditions that are advantageous to users 1 with disabilities or disadvantageous conditions (handicaps) on users 1 without disabilities, even users 1 with disabilities can participate in the "competition" to move the control target 2 without any disadvantage.

[0148] In the above example, the control target 2 is a robot, but the control target 2 may also be, for example, a display unit (third notification unit 42) that displays an avatar. In this case, the user 1 may be made to move the left and right arms 1A up and down, and an avatar with wings may imitate this movement, and further, a scene in which the avatar flies in the air may be reproduced as a first additional action. Alternatively, the user 1 may be made to move the arms 1A from back to front, and the avatar may imitate this movement, and further, a scene in which the avatar swims may be reproduced as a first additional action.

[0149] Here, if the first additional motion is the forward movement of the robot, the motion that causes the imitation motion (the robot's stepping) is the motion of the leg 1L of the user 1. If the first additional motion is the motion of the avatar flying in the air, the motion that causes the imitation motion (the up and down movement of the avatar's wings, etc.) is the motion of the arm 1A of the user 1. If the first additional motion is the motion of the avatar swimming, the motion that causes the imitation motion is the motion of the arm 1A of the user 1. Thus, in the third embodiment, a motion related to the motion of the user 1 that caused the imitation motion is added as the first additional motion. In other words, the first additional motion can be said to be a motion that can be generated as a result (secondarily derived) of the motion of the user 1 that caused the imitation motion. This gives the user 1 the feeling that he or she is controlling the control target 2, making it possible to actively participate in rehabilitation.

[0150] If the control target 2 is an avatar, the target exercise amount can be effectively used as the goal attainment condition described above. In this case, the third notification unit 42 displays the goal at a position corresponding to the target exercise amount, and the avatar at a position corresponding to the accumulated exercise amount of the user 1 calculated based on the output of the motion detection unit 3 (for the target exercise amount and the accumulated exercise amount, see the description of the rehabilitation completion determination (ST08) in the first embodiment).

[0151] Now, when the control target 2 is performing the above-mentioned imitation movement and the first additional movement, if a specific movement of a body part different from the body part corresponding to the movement of the user 1 that caused the first additional movement (a specific movement different from the movement that caused the first additional movement) is detected, the first control unit 10 may change the content of the first additional movement. That is, if the control unit (first control unit 10) detects a specific movement of the user 1 based on the output of the movement detection unit 3, it may change the content of the first additional movement.

[0152] For example, when the first control unit 10 detects, based on the output of the motion detection unit 3, that the position of the arm 1A of the user 1 has been displaced directly upward, it changes the movement direction of the control target 2. Specifically, when the first control unit 10 detects, for example, that the user 1 has raised his right arm 1AR directly upward, it controls the control target 2 to turn right (or left). Conversely, when it detects that the user 1 has raised his left arm 1AL directly upward, it controls the control target 2 to turn left (or right). By changing the displacement amount of the legs 2L of the control target 2 between the left and right, it is possible to adjust the turning direction and turning radius. This allows the user 1 to imagine the control target 2 as a vehicle such as a car, and to get the feeling that he or she is operating the turn signal.

[0153] In this way, during the rehabilitation process, the user 1 notices that when his / her own movement meets certain conditions, the control target 2 performs a special action (the first additional action is changed). Then, by trying to manipulate the control target 2 more proactively, the user 1 can increase the amount of exercise he / she does.

[0154] Furthermore, the control unit (first control unit 10) may derive a rating for the user 1, and when the rating satisfies a predetermined condition, have the control target 2 perform a predetermined second additional action. That is, the first control unit 10 functions as a rating derivation unit. Specifically, for example, the accumulated amount of exercise of the user 1 is derived as a rating, and when the accumulated amount of exercise exceeds a predetermined value, the first control unit 10 causes the control target 2 to perform a special action as the second additional action. Here, the special action corresponds to, for example, a forward roll, a backward roll, standing on one leg, etc. Here, the second additional action is performed regardless of the output of the motion detection unit 3. Note that the rating may be derived by a second control unit 3a (see FIG. 3) provided in the motion detection unit 3. In this case, the second control unit 3a functions as a rating derivation unit, and the rating is transmitted to the first control unit 10 via the second communication unit 3c.

[0155] Note that the second additional action may be executed when the control target 2 reaches the goal by linking the score with the above-mentioned goal attainment condition (i.e., by using the achievement status up to the goal attainment condition as the score). For example, if the goal attainment condition is to move the control target 2 by a predetermined distance, the first control unit 10 may refer to the output of the first inertial sensor 18 and calculate the score in the process of reaching the goal attainment condition.

[0156] When executing the second additional action, the first control unit 10 interrupts the imitation action and the first additional action, and then executes the second additional action. After executing the second additional action, the first control unit 10 returns to the imitation action and the first additional action. This provides a so-called surprise when the user 1 completes the rehabilitation task, making it possible to improve the motivation of the user 1 to engage in rehabilitation.

[0157] Furthermore, the control unit (first control unit 10) may cause the control target 2 to perform the second additional action based on the output of the first environmental sensor 21 provided in the control target 2 or the second environmental sensor 41 provided outside the control target 2. As described above, the first environmental sensor 21 and the second environmental sensor 41 (hereinafter, these may be collectively referred to as "environmental sensors") are configured, for example, with a human presence sensor and a sound sensor (microphone). In the third embodiment, an image sensor (first imaging unit 13 or second imaging unit 40) is also included in the environmental sensor. By using a human presence sensor, sound sensor, or image sensor as the environmental sensor, it is possible to detect the environment (ambience of the place) around the user 1 with a simple configuration.

[0158] The first control unit 10 estimates the situation in the place where the user 1 is undergoing rehabilitation based on the output of the environmental sensor. For example, if a large number of people are detected by a human presence sensor and applause is detected by a sound sensor, the first control unit 10 determines that the atmosphere in the place where the user 1 is undergoing rehabilitation is upbeat. Here, for example, if high sound pressure is repeatedly detected in a short period of time, the frequency distribution when the high sound pressure is detected is biased toward the high frequency region, and the sound pressure changes rapidly, it can be determined that applause is occurring in the place. Also, if an image acquired by an image sensor includes a large number of people and it is detected that these people are clapping, it may be determined that the atmosphere in the place is upbeat.

[0159] When the first control unit 10 determines that the user 1's rehabilitation has created an uplifting atmosphere, that is, when people around the user 1 are encouraging the user 1's efforts, the first control unit 10 causes the control target 2 to execute the second additional action. As a result, the user 1 is encouraged by the atmosphere of the place visualized by the second additional action, and his / her motivation to work on rehabilitation is improved.

[0160] The pattern of the second additional action is stored as a motion pattern in the first storage unit 11 or the server storage unit 30b. A plurality of patterns of the second additional action are predefined, allowing the doctor or other medical professional to select a desired pattern. When the doctor or other medical professional inputs a pattern selection command into an input unit (not shown) of the information terminal 43, the command is transmitted to the first control unit 10 of the control target 2 via the network 50. The first control unit 10, upon receiving the command, extracts a motion pattern corresponding to the selected second additional action from the first storage unit 11. Alternatively, the first control unit 10 communicates with the server control unit 30a via the network 50 and extracts a motion pattern corresponding to the selected second additional action from the server storage unit 30b. The first control unit 10 then controls the control target 2 based on the extracted motion pattern.

[0161] That is, the exercise support system S1 of the third embodiment includes an information terminal 43, and the control unit (first control unit 10) changes the content of the second additional action based on instructions input to the information terminal 43. This changes the content of the second additional action that is executed when the user 1 accomplishes a rehabilitation task or when the atmosphere in the rehabilitation setting becomes uplifting, etc., so that the user 1 does not get bored with the action of the control target 2 and maintains motivation to engage in rehabilitation.

[0162] In FIG. 1, the user 1 wears the motion detection units 3 on both arms 1A and both legs 1L (however, when adding the first additional motion described above, the motion detection units 3 may be worn on either both arms 1A or both legs 1L, for example). Here, if the user 1 has a disability (meaning a situation in which the user is essentially unable to move) in one of the arms 1A (or one of the legs 1L), and the disabled part is, for example, the right arm 1AR, the right arm motion detection unit 3AR that should be worn on the right arm 1AR can be worn on the right leg 1LR, thereby making it possible to make the control target 2 perform the above-mentioned imitation / first additional motion. In this case, when the user 1 moves the left arm 1AL and right leg 1LR alternately, the control target 2 performs a stepping motion and further executes the above-mentioned first additional motion.

[0163] However, for example, a user 1 with disabilities in both legs 1L and the left arm 1AL, or a user 1 with disabilities in both legs 1L and both arms 1A, cannot make the controlled object 2 perform the imitation and first additional movements. Therefore, in the third embodiment, a mode is provided in which movement information of a body part of the user 1 other than the part where the movement detection unit 3 is attached is generated based on the output of the movement detection unit 3, and the imitation movement and the first additional movement are performed based on this generated movement information (hereinafter, this mode may be referred to as the "third movement mode").

[0164] The third operation mode is activated when a doctor or the like operates the information terminal 43 to send a predetermined command to the first control unit 10. In the third operation mode, a motion vector (first motion vector) is first detected based on the output of one of the motion detection units 3. Then, the first motion vector is regarded as a waveform and a motion vector (second motion vector) with an opposite phase, shifted by 180°, is generated. In this way, in the third operation mode, based on the first motion vector based on the movement of, for example, the left arm 1AL (or right arm 1AR) of the user 1, the first control unit 10 generates the second motion vector, which is a motion vector with an opposite phase to the first motion vector.

[0165] This second motion vector is used as motion information of a body part of the user 1 different from the part where the motion detection unit 3 is attached. Then, the right arm 2AR of the control target 2 is controlled based on the first motion vector (mainly the z-axis component), and the left arm 2AL is controlled based on the second motion vector (mainly the z-axis component).

[0166] Furthermore, for example, if user 1 wears the motion detection unit 3 around his neck, the first control unit 10 generates a second motion vector (motion information) based on a first motion vector derived from the user 1's neck movement, the second motion vector being a motion vector with an opposite phase to the first motion vector. Then, for example, the right leg 2LR of the control target 2 is controlled based on the first motion vector, and the left leg 2LL is controlled based on the second motion vector (mainly the z-axis component). In other words, even if user 1 can only move his neck, the above-described stepping motion and first additional motion can be executed, allowing user 1 to participate in a competition. This can increase the motivation of user 1 to engage in rehabilitation, even if he has a disability in some part of his body.

[0167] A modified example of the third embodiment will be described below. In the above-described third embodiment, the imitation movement (stomping) and the first additional movement (walking) are executed prior to the second additional movement. In the modified example, the first additional movement is not executed, and the second additional movement is executed following the imitation movement. In the modified example, the second additional movement is also executed regardless of the output of the movement detection unit 3.

[0168] Furthermore, when the score for the user 1 derived by the score derivation unit (the first control unit 10 or the second control unit 3a) satisfies a predetermined condition, the controlled object 2 may be made to perform a predetermined second additional action. Also, the score may be linked to a goal attainment condition, or the second additional action may be performed based on the output of an environmental sensor.

[0169] As described above, the exercise support system S1 of the modified example includes a movement detection unit 3 that detects the movement of at least a part of the body of the user 1, a control target 2, and a control unit (first control unit 10) that controls the control target 2, and the first control unit 10 causes the control target 2 to perform an imitation action that imitates the movement of the user 1 based on the output of the movement detection unit 3, and also causes the control target 2 to perform a predetermined second additional action at a timing different from the execution of the imitation action and independent of the output of the movement detection unit 3. A further modified example may include a score derivation unit (first control unit 10 or second control unit 3a) that derives a score for the user 1, and the control unit (first control unit 10) may be configured to cause the control target 2 to perform the predetermined second additional action when the score derived by the score derivation unit satisfies a predetermined condition.

[0170] (Fourth embodiment) FIG. 6 is a configuration diagram showing the configuration of an exercise support system S1 according to a fourth embodiment of the present invention. In the first to third embodiments described above, one control target 2 performs an imitation movement based on the output of a motion detection unit 3 including a right arm motion detection unit 3AR, a left arm motion detection unit 3AL, a right leg motion detection unit 3LR, and a left leg motion detection unit 3LL worn by a user 1. In other words, one control target 2 is exclusively used by one user 1. Note that in the second embodiment, one control target 2 can be shared by multiple users 1, but sharing is only possible by dividing the time period during which the control target 2 is used after one user 1 has exclusive use of the control target 2.

[0171] On the other hand, the fourth embodiment assumes a case in which multiple users 1 simultaneously use a single control target 2 to undergo rehabilitation. As shown in FIG. 6, there are multiple users 1, and each user 1 is equipped with a motion detection unit 3 (at least one of a right arm motion detection unit 3AR, a left arm motion detection unit 3AL, a right leg motion detection unit 3LR, and a left leg motion detection unit 3LL). Note that the configuration of each motion detection unit 3 is the same as that described using FIGS. 3(A) and 3(B), and the configuration of the control target 2 is the same as that described using FIG. 2 or FIG. 5, so their description will be omitted here. The description will continue below using FIG. 5 in addition to FIG. 6.

[0172] In the fourth embodiment, for each user 1, user movement information corresponding to the part of the body to which the movement detection unit 3 is attached is transmitted to the control target 2 (first control unit 10). Then, the first control unit 10, which has received the user movement information, controls the control target 2 to perform an imitation movement based on the movements of the multiple users 1 detected by the respective movement detection units 3. That is, the exercise support system S1 of the fourth embodiment includes, for each of the multiple users 1, a movement detection unit 3 that detects the movement of at least a part of the body of the user 1, the control target 2, and a control unit (first control unit 10) that controls the control target 2. There is only one control target 2, and the first control unit 10 controls the control target 2 so as to imitate the movements of the multiple users 1 detected by the respective movement detection units 3. This allows multiple users 1 to simultaneously perform rehabilitation using a single control target 2.

[0173] In such an embodiment, when multiple users 1 simultaneously move the same body part (for example, the right arm 1AR), the movements of each user 1 may not match. In this case, the first control unit 10 acquires multiple pieces of user movement information corresponding to the movements of the multiple users 1, but these pieces of user movement information will conflict when processed by the first control unit 10 (i.e., when performing an imitation movement). Examples of cases of conflict include the following. The following combinations are also possible.

[0174] (C1) One user 1 raises his right arm 1AR, and another user 1 lowers his right arm 1AR. That is, the directions of the movements compete with each other. (C2) One user 1 raises (lowers) his / her right arm 1AR directly above (directly below), while another user 1 raises his / her right arm 1AR to shoulder height (lowers it to knee height). In other words, the displacements of the movements compete with each other. (C3) One user 1 quickly raises his right arm 1AR, while another user 1 slowly raises his right arm 1AR. In other words, the speed of the movements competes.

[0175] When there is conflict between the user motion information output from the motion detection units 3, the first control unit 10 may calculate the average values ​​(Mean) for each of the X, Y, and Z axes for the user motion information output from, for example, the right arm motion detection units 3AR worn by each user 1, and control the control target 2 based on these average values. Of course, it is also possible to derive the median for each of the X, Y, and Z axes and use this as the average value. The outputs of the left arm motion detection unit 3AL, right leg motion detection unit 3LR, and left leg motion detection unit 3LL are processed in the same manner. In other words, when there is conflict between the user motions 1 detected by the motion detection units 3 worn by multiple different users 1, the control unit (first control unit 10) controls the control target 2 to reproduce the average motion of the multiple users 1.

[0176] Here, as in (C1), if the direction of movement of each user 1 is not consistent (for example, there is a large variation in the output of each right arm movement detection unit 3AR), resulting in a small calculated average value, the movement (displacement) of the control target 2 will be small. However, during the rehabilitation process, the users 1 realize that cooperation between each user will result in a larger movement of the control target 2. Each user 1 will then try to have the control target 2 imitate their coordinated movement. Furthermore, each user 1 is expected not only to cooperate, but also to make an effort to increase their own movement and increase the displacement of the imitated movement of the control target 2.

[0177] This makes it possible to increase the overall amount of exercise of each user 1 participating in rehabilitation. If the average value is calculated to be small, the first control unit 10 may issue a notification via the notification unit (first notification unit 14 or third notification unit 42) such as, "Make sure everyone moves in unison, and move more for me." This motivates each user 1 to move in unison.

[0178] Of course, even in the case of (C2) competing movement displacement amounts or (C3) competing movement speeds, the average movement of each user 1 described above may be reproduced by the control target 2. This can motivate each user 1 to coordinate their movements with each other.

[0179] On the other hand, the control target 2 may be controlled to imitate the movements of one user 1 selected from multiple users 1. That is, when the movements of users 1 detected by the movement detection units 3 worn by multiple different users 1 compete with each other, the control unit (first control unit 10) controls the control target 2 to imitate the movements of one user 1 selected from the multiple users 1. This allows the selected user 1 to make an effort to have the control target 2 imitate his or her movements, thereby improving the effect of rehabilitation.

[0180] Furthermore, the control unit (first control unit 10) may derive scores for multiple users 1 based on the output of the motion detection unit 3, and may cause the control target 2 to perform a predetermined third additional motion if the scores satisfy a predetermined condition. In this way, in the fourth embodiment, the first control unit 10 also functions as a score deriving unit. Specifically, for example, the accumulated amounts of exercise of multiple users 1 may be added up as a score, and if this sum exceeds a predetermined value, the control target 2 may be caused to perform a special motion as the third additional motion. Examples of the special motion include a forward roll, a backward roll, and standing on one leg. Note that the third additional motion is performed regardless of the output of the motion detection unit 3. Of course, as in the third embodiment, the score may be derived by a second control unit 3a (see FIG. 3) provided in the motion detection unit 3. In this case, the second control unit 3a functions as a score deriving unit.

[0181] When executing the third additional action, the first control unit 10 interrupts the imitation action and then executes the third additional action. Then, after executing the third additional action, the first control unit 10 returns to the imitation action again. This provides a surprise as an outcome for all of the multiple users 1 participating in rehabilitation, and makes it possible to increase the motivation of each of the multiple users 1 to engage in rehabilitation.

[0182] Furthermore, although different from the above (C1) to (C3), a situation that should be taken into consideration when making the control target 2 execute an imitation motion is a case where one user 1 lifts the right leg 1LR and another user 1 lifts the left leg 1LL. In such a situation, if the control target 2 (here, a robot) executes an imitation motion, it will be structurally unable to maintain an upright state.

[0183] Here, the first control unit 10 may continue to perform the imitation action and control the control target 2 to intentionally cause it to fall over. Then, the first notification unit 14 may issue a speech such as "Be careful not to fall over" to alert the user 1. This may encourage the user 1 to not only cooperate but also actively check the movements of other users 1, stimulating cognitive functions such as attention. Of course, if the control target 2 is structurally unable to maintain an upright position, the imitation action may be interrupted and only an alert may be issued.

[0184] (Fifth embodiment) In the above-described fourth embodiment, it is assumed that multiple users 1 simultaneously use a single control object 2 to perform rehabilitation, and the control object 2 is controlled to simultaneously imitate the movements of the multiple users 1 as long as the movements of the multiple users 1 do not conflict with each other. In contrast, the fifth embodiment is similar in that multiple users 1 perform rehabilitation using a single control object 2, but differs in that the control object 2 is controlled to imitate the movements of one user selected from the multiple users 1. In other words, the fifth embodiment can also be said to be an embodiment in which the right to operate the control object 2 (hereinafter sometimes referred to as the "operation right") is shared by multiple users 1 on a time-sharing basis.

[0185] As shown in FIG. 6, there are multiple users 1 in the fifth embodiment, such as a first user 1a, a second user 1b, and a third user 1c, who use the exercise support system S1. Of course, the number of users 1 may be two or more (N users, where N is a natural number equal to or greater than two). Each user 1 is fitted with a motion detection unit 3 (at least one of a right arm motion detection unit 3AR, a left arm motion detection unit 3AL, a right leg motion detection unit 3LR, and a left leg motion detection unit 3LL). The configuration of each motion detection unit 3 is the same as that described using FIGS. 3(A) and 3(B), and the configuration of the control target 2 is the same as that described using FIG. 2 or 5, so a description thereof will be omitted here.

[0186] In the fifth embodiment as well, the user movement information of each user 1 is transmitted from the movement detection unit 3 to the control target 2 (first control unit 10). Then, the first control unit 10, which has received the user movement information, controls the control target 2 based on the user movement information detected by the movement detection unit 3 worn by any of the users 1, and performs the imitation action.

[0187] FIG. 7 is a flowchart showing the operation of the exercise support system S1 according to the fifth embodiment of the present invention. The following description will continue with reference to FIG. 7 in conjunction with FIGS. 6 and 5. First, the first control unit 10 presents a "movement model" to multiple users 1 (ST101). Specifically, as in the first embodiment, a predetermined movement pattern is extracted from the first storage unit 11, and the arms 2A, legs 2L, etc. of the control target 2 are controlled based on this movement pattern. That is, the control target 2 operates in the first movement mode already described. At this time, it is preferable to motivate the users 1 by uttering something like "Come on, imitate me and move like this" via the first notification unit 14, etc.

[0188] The first control unit 10 acquires user movement information from the movement detection unit 3 worn by each user 1 when the control target 2 is operating in the first operation mode. Then, based on each user movement information, it selects from the multiple users 1 a user 1 (here, the first user 1a) whose movements the control target 2 will imitate (ST102). Then, it controls the control target 2 based on the user movement information output from the movement detection unit 3 worn by the selected user 1. That is, the control target 2 imitates the movements of the selected user 1 (ST103). The criteria for selection will be described later.

[0189] As described above, the exercise support system S1 of the fifth embodiment includes a movement detection unit 3 worn by each of multiple users 1 and detecting the movement of at least a part of the body of the user 1, a control target 2, and a control unit (first control unit 10) that controls the control target 2, and the control unit controls the control target 2 based on the output of the movement detection unit 3 worn by the first user 1a so as to imitate the movements of a first user 1a selected from the multiple users 1. This allows the selected first user 1a to make an effort to have the control target 2 imitate his or her movements, thereby improving the effect of rehabilitation.

[0190] Next, the first control unit 10 uses a timer (not shown) to determine whether a first period (exclusive period) has elapsed (ST104). Here, the exclusive period refers to a time allocated to the selected user 1. During the exclusive period, the selected user 1 can have the control target 2 imitate his / her movements (i.e., monopolize the operation right of the control target 2). The exclusive period may be determined as appropriate, and the same period may be uniformly set for each user 1 (e.g., 30 seconds). Alternatively, the exclusive period may be set individually for each user 1 based on the rehabilitation goal index or target exercise amount stored in the server storage unit 30b. Furthermore, if the exercise load is determined to be excessive based on the measured biological information described in the first embodiment, the exclusive period of the user 1 who has exclusive operation right may be immediately terminated.

[0191] If the exclusive period has not elapsed (No in ST104), the first control unit 10 returns the process to ST104 and waits until the exclusive period has elapsed. If the exclusive period has elapsed (Yes in ST104), it determines whether the second period (rehabilitation duration) has ended (ST105). Here, the rehabilitation duration is set to, for example, the rehabilitation duration (20 minutes). The multiple users 1 continue rehabilitation during the rehabilitation duration. If the rehabilitation duration has ended (Yes in ST105), the process ends. On the other hand, if the rehabilitation duration has not ended (No in ST105), the process returns to ST101.

[0192] In the fifth embodiment, the process of ST101 and the process of ST102 are repeatedly executed until the rehabilitation duration ends. During the process of repeatedly executing these processes, the user 1 whose movements are imitated by the control target 2 is sequentially switched (this can also be said to mean that the users 1 are replaced, or that the right to operate the control target 2 is switched to another user 1). That is, in the exercise support system S1 of the fifth embodiment, the control unit (first control unit 10) controls the control target 2 based on the output of the movement detection unit 3 worn by the second user 1b after a predetermined period has elapsed, so that the control target 2 imitates the movements of the second user 1b, who is different from the first user 1a. This provides multiple users 1 with an opportunity to operate the control target 2, improving the motivation of each user 1 and enabling improved rehabilitation effects.

[0193] In addition, during the course of rehabilitation, the first control unit 10 may notify information about the selected (or selected) user 1 via a notification unit (first notification unit 14 or third notification unit 42). A doctor or the like operates the information terminal 43 to set the user 1 wearing each movement detection unit 3 (at least one of the right arm movement detection unit 3AR, left arm movement detection unit 3AL, right leg movement detection unit 3LR, and left leg movement detection unit 3LL). This associates the identifier (ID) of each movement detection unit 3 with the user ID that identifies the user 1. This correspondence is stored in the first storage unit 11 of the control target 2 or the server storage unit 30b.

[0194] The first control unit 10 accesses the first storage unit 11 and identifies the user ID based on the identifier (ID) transmitted from the movement detection unit 3 together with the user movement information. Then, the first control unit 10 accesses the server storage unit 30b via the network 50 and acquires user information associated with the user ID (see the above-mentioned explanation regarding the mass storage of the server 30). The user information includes the name or nickname of the user 1, and the first control unit 10 acquires the name, etc. of the user 1 (i.e., the selected user 1) whose movements are to be imitated. Then, when performing an imitation action based on the output of the movement detection unit 3 attached to the user 1, the first control unit 10 utters the name, etc. of the user 1 via the first notification unit 14 or the third notification unit 42, such as, for example, "I'm going to imitate you now," or "I'm imitating you now."

[0195] As described above, the exercise support system S1 of the fifth embodiment includes a notification unit (first notification unit 14 or third notification unit 42), and the control unit (first control unit 10) notifies multiple users 1 of information about a specific user 1 whose movements are imitated by the control target 2 via the notification unit. As a result, the user 1 who has been notified of the name or nickname will hope to hear their own name or nickname uttered by the control target 2, and the competitive spirit of the other users 1 will be aroused, making it possible to encourage each user 1 to participate more actively in rehabilitation.

[0196] Below, a specific example will be given to explain the criteria for selecting, from a plurality of users 1, a specific user 1 whose movements the control target 2 will imitate. <Example 1> While presenting the above-mentioned "movement model" (ST101), the first control unit 10 receives user movement information from the movement detection unit 3 worn by each user 1. Then, the first control unit 10 compares the movement of the control target 2 with the user movement information, and extracts the movement detection unit 3 that outputs user movement information that most closely resembles the movement of the control target 2. Furthermore, the first control unit 10 accesses the server 30 and extracts the user ID associated with the user movement information (the identifier (ID) of the movement detection unit 3) from the server storage unit 30b.

[0197] When presenting a "movement example," the control object 2 is controlled based on a movement pattern. As described above, the movement pattern includes items (parameters) related to the movement speed, movement cycle (rhythm), and displacement range (displacement angle) of each part of the control object 2 (e.g., arm 2A). Furthermore, the user movement information based on the movement of each user 1 includes three-axis acceleration information and / or three-axis angular velocity information based on the movement of the user 1's arm 1A and leg 1L (see Figure 1). All of this can be treated as vector information (three-dimensional movement vectors).

[0198] The first control unit 10 acquires a motion vector based on the movement pattern and a motion vector based on the user movement information at a predetermined timing, and determines the similarity between the movement of the control target 2 and the movement of each user 1, for example, using the cosine similarity of the two vectors (= inner product of the two vectors / magnitude of the two vectors (L2 norm)).The first control unit 10 then selects a user 1 wearing a movement detection unit 3 that outputs user movement information that is most similar to the movement pattern (i.e., the movement of the control target 2), and imitates the movement of the selected user 1.Of course, the difference in magnitude between the two vectors may be taken into consideration when calculating the similarity, and the similarity may be calculated to be greater the smaller the absolute value of the difference.

[0199] In the flowchart of FIG. 7, it is preferable that the first user 1a is selected when the processing of ST102 is executed for the first time (i.e., the first iteration of the processing loop), and that a second user 1b different from the first user 1a is selected when the processing of ST102 is executed for the second time. That is, when the processing of ST102 is executed for the second time, the first user 1a, who has already been selected, may be excluded from the selection candidates. Furthermore, when the processing of ST102 is executed for the third time, the first user 1a and the second user 1b may be excluded from the selection candidates. However, since one of the purposes of the fifth embodiment is to stimulate the competitive spirit of each user 1 and encourage them to participate more actively in rehabilitation, for example, all users 1 may be selected in the processing of ST102 from the third iteration onward.

[0200] As described above, in the exercise support system S1 of the fifth embodiment, the control unit (first control unit 10) causes the control target 2 to perform a predetermined movement (exemplary movement), and selects a first user 1a (or a second user 1b other than the first user 1a) based on the movements of the multiple users 1 while the control target 2 is performing the predetermined movement. This allows each user 1 to desire to be imitated by the control target 2, thereby making it possible to have each user 1 participate more actively in rehabilitation.

[0201] <Example 2> Instead of performing the process of presenting a "movement model" (ST101), the first control unit 10 may output predetermined instruction information (movement instructions) to multiple users 1, for example, via the first notification unit 14 (see FIG. 5) or the third notification unit 42 (see FIGS. 5 and 6). That is, in Example 2, the control target 2 does not present a "movement model" to the users 1. The movement instructions are preferably information that identifies the body part to be moved and further specifies the specific content of the movement, such as "Raise your right arm straight up" or "Step to this rhythm, one, two." Here, the first storage unit 11 (or the server storage unit 30b) stores movement instructions and movement patterns in association with each other. That is, when giving movement instructions to the user 1, the first control unit 10 extracts a movement pattern associated with the instruction. Then, a motion vector is derived from the movement pattern. Of course, time-series motion vectors corresponding to the movement patterns may be derived in advance and stored in the server storage unit 30b, etc.

[0202] After issuing a movement instruction, the first control unit 10 derives a motion vector from the user movement information at a predetermined timing (multiple timings), and calculates the cosine similarity between the motion vector based on the movement pattern and the motion vector based on the user movement information, as in Example 1. Then, the first control unit 10 selects the user 1 with the highest similarity (correlation) between the movement instruction for the user 1 and the output of the movement detection unit 3. Also in Example 2, when the processing of ST102 is executed for the first time, the first user 1a is selected as the user 1, and when the processing of ST102 is executed for the second time, a second user 1b different from the first user 1a is selected.

[0203] As described above, the exercise support system S1 of the fifth embodiment includes a notification unit (first notification unit 14 or third notification unit 42), and a control unit (first control unit 10) outputs instruction information instructing a plurality of users 1 to move via the notification unit, and selects the user 1 having the highest correlation between the instruction information and the output of the movement detection unit as the first user 1a or the second user 1b. This allows each user 1 to obtain the right to operate the control target 2 by making a movement in accordance with the instruction information, thereby enabling each user 1 to more actively participate in rehabilitation.

[0204] <Example 3> For example, after outputting the instruction information in Example 2, the first control unit 10 may select a user 1 with a relatively larger or smaller displacement from among the multiple users 1, and control the control target 2 based on the movement of the selected user 1. By selecting a user 1 with a large displacement as the first user 1a or the second user 1b, the user 1 with a small movement will try to make larger body movements, improving the effectiveness of rehabilitation. Furthermore, by having the control target 2 imitate the movement of a user 1 with a small displacement, even a user 1 who cannot make large movements of the part of the body to which the movement detection unit 3 is attached can maintain their motivation to participate in rehabilitation. Whether the imitated movement is based on a large or small displacement movement may be explicitly included in the instruction information or may be specified by a doctor or other person operating the information terminal 43.

[0205] Furthermore, the first control unit 10 may select, from among multiple users 1, a user 1 that has made relatively faster or slower movements, and control the control target 2 based on the movement of the selected user 1. By having the control target 2 imitate faster movements (agile movements), users 1 that move slowly will try to move their bodies faster, improving the effectiveness of rehabilitation. Furthermore, by having the control target 2 imitate slow movements, even users 1 who cannot move the part of the body to which the movement detection unit 3 is attached quickly can maintain their motivation to participate in rehabilitation. Whether the imitated movement is based on a faster or slower movement may be explicitly included in the instruction information, or may be specified by a doctor or the like operating the information terminal 43.

[0206] As described above, in the exercise support system S1 of the fifth embodiment, the control unit (first control unit 10) selects as the first user 1a or the second user 1b the user 1 whose slower or faster movement, or whose smaller or larger displacement is detected by the movement detection unit 3. This makes it possible to allow each user 1 to participate in rehabilitation fairly even if the degree of disability of each user 1 is different.

[0207] The above describes the criteria for selecting (changing or switching) a user 1 whose movements the control target 2 will imitate. Alternatively, the control unit (first control unit 10) may randomly select the first user 1a, the second user 1b, or the Nth user 1 (1n) from among multiple users 1. This allows all users 1 participating in rehabilitation to participate fairly in the operation of the control target 2. Note that if a random lottery is employed, in which one user is simply selected randomly from a group of users 1, a bias in the selection (winning) of users 1 may occur. To avoid this, a user 1 who has been selected once may be excluded from the next lottery, and one user may be randomly selected from the group of users 1 from which the winner was excluded. After all users 1 have been selected once, a random lottery may be held again for all users 1. This further ensures fairness. Of course, the above-mentioned <Example 1>, <Example 2>, and <Example 3> may be combined with the random lottery.

[0208] Although the exercise support system S1 according to the present invention has been described in detail above based on specific embodiments, these embodiments are merely examples, and the present invention is not limited to these embodiments. For example, in the fifth embodiment, a second additional action (provision of a surprise) may be performed based on the output of the second environment sensor 41 described in the third embodiment.

[0209] Furthermore, in the fifth embodiment, if it is determined that the selected user 1 is performing a stepping motion based on the output of the motion detection unit 3 worn by the user 1, the first additional motion (e.g., moving forward or backward) described in the third embodiment may be added.

[0210] In the second to fifth embodiments, the division of processing between the first control unit 10 of the controlled object 2 and the server control unit 30a of the server 30 may be determined as appropriate. [Industrial Applicability]

[0211] The exercise support system S1 and exercise support method of the present invention can motivate users to actively participate in rehabilitation and improve the effectiveness of rehabilitation, and therefore can be widely used in health insurance medical institutions such as hospital rehabilitation departments, orthopedic clinics, elderly care facilities, and specialized rehabilitation facilities, as well as in home rehabilitation. [Explanation of symbols]

[0212] 1 User 1a 1st user 1b 2nd user 1c 3rd user 2. Control Object 3. Motion detection section 10 First control section 11 1st memory section 30 servers 30a Server control unit 30b Server storage section S1 Exercise Support System

Claims

1. a motion detection unit attached to each of a plurality of users and configured to detect motion of at least a part of the body of each of the users; A control object; a control unit that controls the control target; Equipped with an exercise support system characterized in that the control unit controls the controlled object based on an output of the movement detection unit worn by the first user so as to imitate the movement of a first user selected from the plurality of users.

2. 2. The exercise support system according to claim 1, wherein the control unit controls the controlled object based on the output of the movement detection unit worn by the second user so as to imitate the movement of a second user different from the first user after a predetermined period of time has elapsed.

3. The exercise support system according to claim 2 , wherein the control unit randomly selects the second user from among a plurality of the users.

4. The exercise support system according to claim 1, characterized in that the control unit causes the control object to perform a predetermined movement, and selects the first user based on movements of the plurality of users while the control object is performing the predetermined movement.

5. The exercise support system according to claim 2, characterized in that the control unit causes the controlled object to perform a predetermined movement and selects the second user based on movements of the plurality of users while the controlled object is performing the predetermined movement.

6. Equipped with a notification unit, 3. The exercise support system according to claim 2, wherein the control unit outputs instruction information instructing a movement to the plurality of users via the notification unit, and selects the user having the highest correlation between the instruction information and the output of the movement detection unit as the second user.

7. The exercise support system according to claim 1, characterized in that the control unit selects as the first user the user whose slower or faster movement, or whose smaller or larger displacement is detected by the movement detection unit.

8. The exercise support system according to claim 2, wherein the control unit selects as the second user the user whose slower or faster movement, or whose smaller or larger displacement is detected by the movement detection unit.

9. Equipped with a notification unit, The exercise support system according to any one of claims 1 to 8, characterized in that the control unit notifies the plurality of users of information regarding the specific user whose movement the control target imitates via the notification unit.

10. a motion detection unit attached to each of a plurality of users detects a motion of at least a part of the body of the user; An exercise support method, comprising: controlling a control target based on a movement of at least a part of a body of a first user selected from a plurality of users, so as to imitate the movement of the first user.

11. The exercise support method according to claim 10, characterized in that, when a predetermined period of time has elapsed, the control object is controlled based on the movement of at least a part of the body of a second user different from the first user so as to imitate the movement of the second user.

Images