Translation. Region: Russian Federal
Source: Novosibirsk State University – Novosibirsk State University –
A fourth-year student developed a hardware and software system for monitoring vital signs and motor activity for people with disabilities Faculty of Information Technology, Novosibirsk State University Mikhail Evdokimov. It allows remote monitoring of the user’s pulse rate, blood oxygen saturation, body temperature, as well as his movement in space. The prototype of the complex has already been assembled, testing is underway.
— The increase in the number of elderly people and patients suffering from various lifestyle-related diseases makes it critically important to develop systems that facilitate monitoring of the health of these people outside of hospitals, allowing them to stay at home or at work. Having various health limitations, these people need constant monitoring of their physical condition. Often, for various reasons, relatives cannot provide them with constant supervision, and negative changes occur when patients are left alone and are unable to seek help in a timely manner, which can sometimes lead to tragic consequences. Existing monitoring tools rarely combine autonomy and mobility, compactness and low cost, so we decided to create a monitoring system that would track the main indicators of the user’s physical condition and, if they deviate from the norm, notify the medical workers under whose supervision the user is, — said Mikhail Evdokimov.
The young researcher is confident that his project, which he is working on as part of his final qualifying work under the scientific supervision of the adviser to the rector of NSU, Professor Alexander Shafarenko, will help the elderly, including those with some forms of dementia, by simplifying health monitoring by transmitting the dynamics of indicators to medical organizations for the timely provision of medical care.
An important element of the system is a wristband that reads the pulse, body temperature and relative coordinates of the user. Other elements are a microcontroller with a magnetic sensor (one or more), as well as a central microcontroller, where all information from the sensors and the bracelet is sent.
The device’s bracelet looks very similar to a regular smart watch, and if the size of the device decreases during further development of the project, it will resemble fitness bracelets. But if the “smart watch” partially or completely implements health monitoring functions, then due to the limited functionality, they are not able to recognize abnormal conditions, and even more so, they do not have the functions of transmitting alarm information to medical workers. In addition, the “smart watch” device is tied to a specific manufacturer and is closed, which means that it does not allow modification by the medical service provider. These shortcomings have been eliminated by the developers of this project.
According to Mikhail Evdokimov, the vital signs monitoring system should be open and independent. Therefore, he studied and analyzed the design and functionality of several modifications of “smart watches” in search of successful ideas and the formation of requirements for his project. The young researcher came to the conclusion that a solution that meets all his requirements is currently absent, and the category of “smart watches” is only indirectly related to health monitoring and is not suitable for use as a component of the monitoring system he is creating. Therefore, his own “smart bracelet” was assembled and programmed, which has the set of functions necessary for the monitoring system. Using the C language, the program code was written for the operation of the built-in accelerometer, pulse sensor, data transmission via communication modules, power management and analysis of the collected data. The ESP-NOW protocol was chosen for communication between the system nodes. It was developed specifically for transmitting information between microcontrollers based on the ESP-32 processor and is a more efficient version of classic Wi-Fi. The interaction between the nodes has a client-server architecture: the bracelet and magnetic sensors send their readings to the central microcontroller with a static IP address.
This system works as follows: a magnetic sensor is installed on the doors of the apartment. It transmits information to the central microcontroller about whether the door is open or closed. The central microcontroller uses this data for analysis.
A magnetic sensor can be installed, for example, on the toilet door. If it has not been opened for a long time, this is an alarm signal. Perhaps the person is unconscious or in a helpless state, and urgent help is needed. A similar sensor installed in the kitchen can warn about the same thing – if a person does not come here for a long time, even to drink water, the system will issue a warning. Situations are envisaged when the user keeps the door open and does not close it behind him, or opens it, but does not enter the room. For this purpose, a comprehensive solution for the task of tracking movement has been developed, in which several sensors work together.
In such cases, an accelerometer is connected to track the user’s movements. The joint work of its coordinate system and magnetic sensors completely solves this problem. It is adjusted by a technical specialist during the initial calibration.
To track the user’s movements in space, the “smart bracelet” has a built-in accelerometer. Its readings are represented by three relative coordinates x, y, z. By analyzing the changes in coordinates, the device determines whether the user is moving or at rest. A fall of a person wearing the “smart bracelet” is recorded by a sharp change in the accelerometer sensor readings. At first, in order to distinguish a fall from movement in the direction of one of the coordinate axes, the difference was estimated not for each coordinate separately, but for the arithmetic mean of their changes. Later, Mikhail Evdokimov replaced this formula with a more accurate one, which uses the root of the sum of the squares of the change in coordinates. And as a threshold separating a fall from uniform movement, an experimentally calculated value was chosen, which is the average between the value of the formula when walking and when falling. When such a state of the user is recorded, the information is sent to the central microcontroller, and ultimately to the medical institution, from where the user’s condition is monitored.
And then, after the initial setup of the system, the user lives his normal life. He does not need to be in a hospital under 24-hour supervision. While at home, he is under the close attention of the monitoring system, which will notify medical workers about deviations of the user’s vital signs from the norm. It is important that this system is reliable and works autonomously.
— The advantage and key feature of my development is its focus on the autonomy of the system, which has its own power source and is highly energy efficient. And unlike smart watches, this is an open system that can be supplemented with other sensors if necessary. It has the ability to improve the analysis algorithms and transmit data to medical personnel. On the currently available hardware, the system can operate without recharging the source for about a week. When using more specialized and advanced hardware, it is expected that the equipment’s operating time without recharging will increase to one month. This is very important, since many elderly people often forget to charge their mobile phones, smart watches and other devices. With this approach, our solution will have minimal dependence on user activity, external power supply, communication channels and third-party platforms. Since all nodes have sufficient autonomy, system maintenance will need to be performed no more than once every six months. To achieve this, we decided to use high-capacity batteries at the hardware level, and at the software level, we created energy-efficient software code, said Mikhail Evdokimov.
An important task of the project was to organize energy-efficient operation of the programmable “smart bracelet” by changing the operating modes of the processor and peripherals depending on the situation. The emphasis was not on hardware, but on software reduction of energy consumption. Mikhail Evdokimov studied software capabilities for managing the energy consumption modes of the ESP-32 microprocessor, which is located on the bracelet. Then the student designed a system that regulates the sleep modes, which the processor can go into depending on the situation.
— In case of a long-term loss of connection with the central microcontroller, the bracelet goes into modem sleep mode, in which the power supply of the communication modules is turned off. After a certain period of time, the device wakes up to check the connection and, in case of failure, goes back to sleep. To minimize the loss of information, the collected sensor readings are partially stored in the bracelet’s memory and sent to the central microcontroller immediately after the connection is restored. As a result, energy consumption is reduced several times, — explained Mikhail Evdokimov.
The user can find out that the watch has been removed from the readings of the infrared heart rate sensor. In this case, since there is no more data to send, the system first goes into modem sleep mode, and after a few seconds, if the user has not returned, the light sleep mode is activated. The watch periodically wakes up to check whether it is on the user’s hand, and either resumes its work or goes back to sleep. If there are no significant changes in the accelerometer readings, for example, when the user is sleeping, the main processor goes into deep sleep mode. At this time, the Ultra Low Power coprocessor is engaged in data processing. It can wake the main processor from sleep by a timer or in the case of active user movement when he or she wakes up. Mikhail Evdokimov clarified that the implementation of this architecture required working with the assembly language.
To assemble the device units, the student purchased ready-made internal components and parts, and made the body of the “smart bracelet” on a 3D printer. The developed prototype of the vital signs and motor activity monitoring system has already confirmed its operability and feasibility of energy autonomy requirements in laboratory conditions, and also showed high potential for further development due to the openness and expandability of the system. The development of such solutions can improve the quality of life of people whose health requires special attention.
Please note: This information is raw content directly from the source of the information. It is exactly what the source states and does not reflect the position of MIL-OSI or its clients.