State monitoring system for person to be assisted

Abstract

Provided is a state monitoring system for a person to be assisted, capable of: quickly ascertaining, in an initial stage, a deterioration tendency of the state of a person to be assisted such as an infant; and taking measures with a temporal margin.　A state monitoring system 1 for a person to be assisted comprises: a state prediction model creation means (control unit 34) for performing machine learning by using teacher data which is outputted from one or more measurement means (a solar radiation meter 18, a temperature sensor 22, a humidity sensor 24, a clinical thermometer 26, a pulse meter 28, and a strain sensor 30) and which includes environmental information that changes every moment and biological information of a person to be assisted as a monitoring subject, and creating a state prediction model in which at least one of the environment information and the degree of the state based on the biological information are associated with each other; an evaluation means (control unit 34) for inputting the current environmental information to the created state prediction model to predict and evaluate the state of the person to be assisted; and a notification means (control unit 34) for determining the tendency of the evaluation by the evaluation means and notifying the outside when it is determined that there is a deterioration tendency.

Description

Assisted person status monitoring system

[0001] The present invention relates to a system for monitoring the condition of a person being assisted, such as an infant or small child, who requires assistance (attention) from another person.

[0002] Patent Literature 1 discloses an infant comfort monitoring system that includes one or more sensors operable to acquire information indicative of an infant's physical condition and a controller in communication with the sensors. The controller determines whether the infant is experiencing distress based on the information from the sensors, and operates a vehicle system to improve the condition based on the determination result.

[0003] Japanese Patent Application Laid-Open No. 2020-142080

[0004] The monitoring system described in Patent Document 1 determines the condition of an infant based on information about the infant at the time an abnormality occurs, such as when the infant is in pain, and therefore is unable to predict the impact that will occur if, for example, the temperature, humidity, or amount of sunlight in the environment in which the infant is placed continues.Furthermore, because the system does not monitor the history of changes in the infant's condition, it is not possible to grasp if the infant's condition is currently normal and if it tends to worsen, and therefore it is not possible to detect the condition until it worsens, leaving no time to take measures.

[0005] Considering the space in which infants are placed, for example, the seat (riding area) of a stroller carrying an infant is generally partitioned into a semi-private compartment, with the front open to facilitate easy entry and exit. This makes it easy for heat generated by the infant's biological activity to build up, and the stroller is directly exposed to environmental changes such as rising temperatures due to sunlight and rising humidity during rainy weather. This makes the infant's condition prone to change. Furthermore, to protect infants in strollers from sudden rain or cold winds, measures such as covering the entire seat with a transparent rain cover are sometimes taken, but when an infant is placed in such a nearly enclosed space, rapid changes in condition are likely to occur, making it difficult for the above-mentioned monitoring system to respond.

[0006] The present invention has been made in consideration of the above-mentioned points, and its purpose is to provide a condition monitoring system for assisted persons that can quickly identify any tendency for the condition of assisted persons, such as infants and young children, to deteriorate at an early stage, allowing measures to be taken with plenty of time to spare.

[0007] In order to achieve the above-mentioned object, the assisted person condition monitoring system (1) of the present invention is characterized by comprising: a condition prediction model creation means (52) that performs machine learning using training data (58) including constantly changing environmental information (60) and biometric information (62) of the assisted person being monitored, each output from one or more measuring means (18, 22, 24, 26, 28, 30), and creates a condition prediction model that corresponds at least one of the environmental information (60) with the degree of condition based on the biometric information (62); an evaluation means (54) that inputs the current environmental information (60) into the created condition prediction model to predict and evaluate the condition of the assisted person (3); and a notification means (56) that judges the trend of the evaluation by the evaluation means (54) and notifies the outside if it is judged that there is a worsening trend.

[0008] The assisted person condition monitoring system of the present invention allows the assistant who receives the notification to reliably recognize that the baby's condition is deteriorating, which allows them to quickly take measures such as removing the baby from the seat, thereby preventing the baby's condition from worsening.

[0009] Furthermore, in the above-mentioned assisted person condition monitoring system (1), the measuring means for acquiring environmental information (60) may include a temperature sensor (22) for measuring temperature and a humidity sensor (24) for measuring humidity, and the measuring means for acquiring biological information (62) may include a thermometer (26) for measuring body temperature, a pulsometer (28) for measuring pulse rate, and a respirometer for measuring respiratory rate. This makes it possible to reliably grasp changes in environmental information and changes in biological information, and to grasp signs of changes in the infant's condition with high accuracy.

[0010] Furthermore, in the above-described assisted person status monitoring system (1), the pulse rate monitor (28) may also function as a respiration monitor. This simplifies the configuration. There is a certain correlation between the pulse rate and the respiration rate, so there is no practical problem in estimating the respiration rate from the pulse rate measured by the pulse rate monitor.

[0011] Furthermore, in the above-described assisted person condition monitoring system (1), the measuring means for acquiring environmental information (60) may further include a solar radiation meter (18) for measuring solar radiation, and the measuring means for acquiring biological information (62) may further include a strain sensor (30) for detecting the movements of the assisted person (3). Assisted people, such as infants, often move when they feel too hot or uncomfortable, so capturing this with a strain sensor makes it possible to grasp changes in their condition more accurately.

[0012] Furthermore, in the above-mentioned assisted person status monitoring system (1), data on the amount of solar radiation may be acquired by acquiring solar radiation forecast values ​​from an external organization via an information and communication network, instead of using the solar radiation meter (18). This simplifies the configuration and reduces costs.

[0013] According to the present invention, it is possible to quickly grasp the tendency for the condition of an assisted person, such as an infant, to deteriorate at an early stage, allowing measures to be taken with ample time to respond.

[0014] FIG. 1 is a perspective view of a stroller equipped with a system for monitoring the status of an assisted person according to one embodiment of the present invention. FIG. 2 is a schematic cross-sectional view of the stroller shown in FIG. 1. FIG. 3 is a control block diagram of the system for monitoring the status of an assisted person. FIG. 4 is a block diagram showing the functions of a control unit. FIG. 5 is a diagram showing the contents of training data. FIG. 6 is an image diagram of data on environmental information and biometric information. FIG. 7 is an image diagram showing data on environmental information and biometric information changing from moment to moment.

[0015] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

[0016] Figures 1 and 2 show a stroller 2 for carrying an infant or toddler 3 who is an assisted person and requires assistance when getting on and off, and is an example of an implementation target of the assisted person status monitoring system 1 (see Figure 3) of this embodiment.

[0017] The stroller 2 has a typical configuration, with an opening 4a at the front for getting in and out, a seat 4 that accommodates and covers the infant 3 being assisted, and a movable support 8 that has a handle 6 for pushing and supports the seat 4. The support 8, which stably supports the seat 4, is made up of rods or pipes joined together and is movable by wheels 10 with locking levers. A smartphone holder 14 is attached to one of a pair of shafts 8a extending diagonally upward from the support 8, for holding a smartphone 12 used by the person assisting the infant 3.

[0018] The seat 4 is surrounded by a seat cover 16 made of a flexible fabric that blocks sunlight, and the opening 4a is left open. On rainy, windy, or dusty days, the opening 4a is covered with a transparent or semi-transparent cover (not shown) via a fastener such as a zipper.

[0019] An actinometer 18 serving as a measuring means is disposed on the outer upper surface of the seat cover 16. As shown in Figure 2, a sensor board 20 is fixed to the seat cover 16 inside the seat portion 4, and a temperature sensor 22 serving as a measuring means for measuring the temperature inside the seat portion 4 and a humidity sensor 24 serving as a measuring means for measuring the humidity inside the seat portion 4 are disposed on the sensor board 20. A non-contact thermometer 26 serving as a measuring means for measuring the body temperature of the infant 3 is disposed on the upper part inside the seat portion 4. A pulse meter 28 for measuring the pulse rate of the infant 3 is attached to the right hand of the infant 3. Furthermore, a plurality of strain sensors 30 serving as a measuring means for detecting the movements of the infant 3 are disposed inside a mat 29 on which the infant 3 sits.

[0020] A storage box 32 is provided on the underside of the seat 4, and a control unit 34 and a DC power supply 36 serving as a battery are housed in the storage box 32. Each of the above-mentioned measuring means is electrically connected to the DC power supply 36 by wires (not shown).

[0021] As shown in Figure 3, the assisted person condition monitoring system 1 includes a control unit 34 and various measuring means (solar radiation meter 18, temperature sensor 22, humidity sensor 24, thermometer 26, pulse meter 28, and strain sensor 30). The control unit 34 is a microcomputer including a CPU 38, ROM 40, RAM 42, a memory unit 46 storing a program 44, a communication I / F 48, and an external I / F 50. The various measuring means (solar radiation meter 18, temperature sensor 22, humidity sensor 24, thermometer 26, pulse meter 28, and strain sensor 30) are electrically connected via the external I / F 50.

[0022] As shown in Figure 4, the control unit 34 combines a condition prediction model creation means 52 that performs machine learning using training data including constantly changing environmental information and biological information of the monitored assisted person (infant 3), each output from one or more measurement means, to create a condition prediction model that matches at least one piece of environmental information with the degree of condition based on the biological information; an evaluation means 54 that inputs current environmental information into the created condition prediction model to predict and evaluate the condition of the assisted person (infant 3); and a notification means 56 that judges the trend of the evaluation by the evaluation means 54 and notifies the outside if it is determined that there is a worsening trend.

[0023] The control unit 14, functioning as the state prediction model creation means 52, performs machine learning using the training data 58 shown in FIG. 5 to create a prediction model. The training data 58 includes environmental information 60 and biological information 62 of the infant 3. The environmental information 60 includes the temperature measured by the temperature sensor 22, the humidity measured by the humidity sensor 24, and the amount of solar radiation measured by the solar radiation meter 18. The biological information 62 includes the body temperature measured by the thermometer 26, the pulse rate measured by the pulse meter 28, the respiratory rate estimated from the pulse rate, and the movements measured by the strain sensor 30. It is said that the respiratory rate is, on average, one-quarter of the pulse rate. In other words, the pulse meter 28 also functions as a respirometer. The control unit 34 loads the program 44 stored in the memory unit 46 into the RAM 42, and the CPU 38 interprets and executes the loaded program 44 to control each unit. Well-known machine learning techniques, such as logistic regression, can be used.

[0024] FIG. 6 is a conceptual diagram of environmental information 60 and biological information 62. As shown in FIG. 7, the environmental information 60 and biological information 62 measured by the above-mentioned measuring means change from moment to moment, and the data accumulates. That is, the amount of teacher data 58 increases over time, and the control unit 14, functioning as the state prediction model creation means 52, performs machine learning using the teacher data 58, including past data (history), to create a prediction model. In FIG. 7, the symbol t indicates the time axis.

[0025] The control unit 34, functioning as the evaluation means 54, inputs current environmental information into the created state prediction model to predict and evaluate the state of the infant 3. In this case, thresholds at which the infant 3 feels discomfort or pain are set in stages in advance, and the evaluation means 54 grasps the trend. The control unit 34, functioning as the notification means 56, notifies the outside when it determines that the trend grasped by the evaluation means 54 is approaching the threshold. For example, the notification is made to the smartphone 12 via the communication I / F 48 by sound or message display.

[0026] The assistant (the person pushing the stroller 2) who receives the notification can recognize that the condition of the infant 3 is deteriorating. This allows them to quickly take measures such as removing the infant 3 from the seat 4, and prevent the condition from worsening in advance. Even if the opening 4a is covered with a cover and the seat 4 is almost sealed during rainy weather, for example, the tendency for the condition of the infant 3 to deteriorate due to a sudden increase in humidity or carbon dioxide concentration can be detected at an early stage, and the assistant can take prompt action.

[0027] While the present invention has been described above as an embodiment, it is not limited to the above embodiment and various modifications are possible within the scope of the claims and the technical concepts described in the specification and drawings. For example, while the above embodiment illustrates implementation with a stroller 2, it can also be implemented in a wheelchair carrying a person requiring assistance, such as an elderly person requiring care. The stroller 2 or wheelchair may be equipped with an electric assist function. Furthermore, while the above embodiment illustrates an example in which the actinometer 18 is directly installed on the stroller 2 to acquire data on the amount of solar radiation, the actinometer 18 may be replaced by a communication I / F 48 that acquires solar radiation forecast values ​​from an external organization via an information and communication network such as the Internet. Furthermore, a pulse oximeter may be used as a measuring device for acquiring biological information.

Claims

1. Performing machine learning using teacher data output from one or more measurement means respectively, including environmental information that changes moment by moment and biometric information of the assisted person who is the monitoring target, and creating a state prediction model that associates at least one of the environmental information with the degree of the state based on the biometric information, a state prediction model creation means; An evaluation means for inputting the current environmental information into the created state prediction model to predict and evaluate the state of the assisted person; A notification means for judging the tendency of the evaluation by the evaluation means and notifying the outside when it is judged that there is a deterioration tendency; A state monitoring system for an assisted person, characterized by comprising the above.

2. The measurement means for acquiring the environmental information includes a temperature sensor for measuring temperature and a humidity sensor for measuring humidity, and the measurement means for acquiring the biometric information includes a thermometer for measuring body temperature, a pulse meter for measuring pulse rate, and a respirometer for measuring respiratory rate. The state monitoring system for an assisted person according to Claim 1, characterized by including.

3. The state monitoring system for an assisted person according to Claim 2, characterized in that the pulse meter also serves as the respirometer.

4. The measurement means for acquiring the environmental information further includes a solar radiation meter for measuring solar radiation amount, and the measurement means for acquiring the biometric information further includes a strain sensor for detecting the movement of the assisted person. The state monitoring system for an assisted person according to Claim 2 or 3, characterized by including.

5. The state monitoring system for an assisted person according to Claim 4, characterized in that the acquisition of the solar radiation amount data is made by acquiring the solar radiation forecast value of an external institution via an information communication network instead of the solar radiation meter.