Keywords

1 Introduction

1.1 Research Background

According to a survey report by the Chinese Sleep Research Association in 2018, nearly 80% of Chinese people between the ages of 10 and 45 fell asleep poorly, and 13% of them suffered from physical and mental discomfort due to insufficient sleep. At the same time, sleep problems are getting younger, and young people are becoming the main population of insomnia [1].

Sleep is an important physiological process of the human body. It plays an important role in replenishing body energy, enhancing immunity, promoting bone development, and slowing down human aging. It is extremely significant to protect people’s physical and mental health. The human sleep phase is divided into non-rapid eye movement (NREM) and rapid eye movement (REM). During the non-rapid eye movement phase, the deep sleep phase occurs periodically. The length of the deep sleep phase largely determines the human body’s sleep quality. Sleep problems such as insomnia, irregular sleep and daytime drowsiness can damage human health over time. Experts on sleep health have stated that irregular schedules are harmful to health, and can easily induce physical and mental diseases such as cardiovascular disease, metabolic diseases, cancer, and depression, and may be life-threatening in severe cases [2].

Traditional healthcare mainly relies on manual or large-scale medical equipment, which costs consumers a lot of time and money, therefore many people are unwilling to get healthcare service. Nowadays, smart home products have entered millions of households with their advantages of portability, intelligence, and affordability, enabling people to enjoy the comfort and convenience of life given by technology. Combining the Internet of Things (IoT) technology with healthcare products, and designing smart healthcare products to replace traditional healthcare methods are the current hotspots and industry trends.

Therefore, an intelligent sleep care product combined with the IoT technology can well fill the market demand and meet people’s expectations for sleep health. The product monitors the sleep quality of the human body in a reasonable and effective way, presents visual analysis results to the user, and gives effective suggestions based on the analysis results, to help users improve sleep quality and achieve the efficacy of sleep care.

1.2 Related Work

The historical achievements on sleep monitoring methods include Polysomnography (PSG), Portable Monitoring (PM), Actigraphy, Micro-movement Sensitive Mattress Sleep Monitoring System (MSMSMS), and Electro Mechanical Film (EMFi) sensors (see Fig. 1).

Fig. 1.
figure 1

Sleep monitoring methods

Full size image
  1. (1)

    Polysomnography (PSG) is a standard test for the diagnosis of sleep-related breathing disorders and narcolepsy, positive airway pressure titration, and assessment of abnormal sleep. Monitoring parameters include electroencephalogram, electrooculogram, mandibular electromyography, oral and nasal airflow and respiratory movement, electrocardiogram, blood oxygen, snoring, limb movement, body position, etc., which has the advantages of accuracy, reliability, intuitiveness and detailedness. At present, it’s the standard physiological diagnosis method in hospital for patients with sleep disorders [3]. However, polysomnography requires the installation of detection electrodes of various instruments on the subject, which is tedious and complicated. At the same time, the cost is high, the learning cost is high, the portability is insufficient, and it is inconvenient to use outside the hospital.

  2. (2)

    Portable Monitoring (PM) is also called Home Sleep Testing (HST). Compared with polysomnography, the equipment is more compact, and it monitors at least 3 parameters: airflow, breathing effort, and Pulse oxygen saturation, but it is still necessary to wear sensors on more than 3 parts of the subject, and patients have a high failure rate when using alone [4].

  3. (3)

    Actigraphy is a wearable device that records physical activity over a period of time, such as a smart bracelet, etc., using the static and movement data collected by the acceleration sensor to estimate the amount of body movement, awake time and other parameters. The actigraphy is convenient to wear and use, and can be used to evaluate some sleep disorders, such as circadian disorder, insomnia, sleep apnea, excessive sleep, sleep movement disorders, etc. [5], which is suitable for assisting polysomnography. Its accuracy and professional is not as good as polysomnography.

  4. (4)

    Micro-movement Sensitive Mattress Sleep Monitoring System (MSMSMS), developed by Academician Yu Mengsun, uses pressure sensors to record any small changes in a person’s bed, including even heartbeat signals [6]. The sleep monitoring process of this system tends to be natural, without the need for medical staff to pre-process, electrodes, and wiring. It only requires the subject to lie on a special mattress to be monitored, and the monitoring results are more reliable. In this system, the heart rate is calculated from the pulse wave signal recorded by the finger-tip photoelectric volume pulse wave detection recorder, which requires the subject to wear the detector at all times during night sleep, which brings certain inconvenience to the subject.

  5. (5)

    Electro Mechanical Film (EMFi) sensor, first invented by Finnish company VTT in 1980, has a long strip shape, light weight and thin shape, and can be isolated from external interference [7]. The sensitivity of EMFi sensors is an order of magnitude better than that of ordinary piezoelectric materials. Due to the high resistance and the hole-like structure of organic materials, the charge can last for a long time, which is convenient for the timely collection of electrical signals [8]. The Finnish company Emfit uses self-developed EMFi sensors for sleep monitoring and has designed an Emfit QS health monitor that can monitor multiple sleep data such as heart rate, breathing, and turning over during sleep. The sleep monitor is currently the most convenient mat-type sleep monitoring product. Users do not need to wear any equipment and can lay on a smart mat based on an EMFi sensor to get non-invasive sleep quality monitoring. It gets rid of the problem that the original smart products need to be replaced frequently, and there is no need to learn how to use it, which significantly improves the convenience of monitoring.

In view of the many excellent characteristics of the EMFi sensor, this paper will use the EMFi sensor to design and develop intelligent sleep assistance system. The electrical signals collected by the EMFi sensor will be used to analyze and calculate physiological signals including respiratory frequency, heart rate, and body movement by the sleep analysis algorithm. The user’s sleep quality can be visualized on the mobile phone, and the system can help the user to harvest a healthy and comfortable sleep experience and develop good sleep habits through sound and light, relatives and friends supervision, and sleep reward program, etc.

2 Investigation and Analysis

2.1 Questionnaire

According to a survey report by the Chinese Sleep Research Association in 2018, nearly 80% of Chinese people between the ages of 10 and 45 fell asleep poorly, and 13% of them suffered from physical and mental discomfort due to insufficient sleep. At the same time, sleep problems are getting younger, and young people are becoming the main population of insomnia. This paper conducts a questionnaire on sleep quality and sleep-related products. The results are shown in Fig. 2.

Fig. 2.
figure 2

Information visualization of survey results

Full size image

Of the respondents in this survey, 54% slept less than 7 h a day, and 49% rated their sleep quality as average or poor. Among the main reasons that affect people’s sleep quality, using electronic devices before bedtime ranks first, accounting for 64%, and study/work stress accounts for 61%, followed by insufficient exercise, strong light/noise, and so on. Among related sleep products, the use rate of smart bracelets was the highest, accounting for 27%, followed by smart sleep aid lamps, accounting for 18%. When asked opinions on a smart product that can monitor and improve sleep quality, 89% of the respondents said they would like to try to accept it, and 30% of them are looking forward to the arrival of this product very much. For the beneficial functions of smart sleep products, 73% of respondents chose monitoring sleep quality, 59% chose adjusting indoor temperature during sleep, 58% chose turning off music and indoor lights after falling asleep, and 52% chose playing music or the radio to adjust the mood. Regarding the concerns about smart sleep products, 77% of the respondents are worried that there may be radiation, and long-term use will be harmful to health, and 47% of them are worried about inconvenience. Finally, regarding the price expectations of smart products, 46% of the respondents chose 200–500 yuan, and 27% chose 500–1000 yuan.

The above questionnaire analysis results show the respondents’ overall sleep status, the opinions and acceptance of sleep-related smart products, the expectations of the functions, deficiencies and prices of sleep-related smart products, and the related product forms still need to be further explored.

2.2 User Interview

User interview is a face-to-face exchange between the designer and the interviewee. It can provide insight into special phenomena, specific situations, specific problems, common habits, special situations, and consumer preferences. It helps designers to understand consumers’ perceptions, opinions, consumption motivations and behaviors of products or services [9]. In order to better understand the sleep status of young people, this paper conducted user interviews to conduct in-depth questions and investigations on the questions in the questionnaire.

The experiences and feelings of one of the interviewees are highly representative. She is an undergraduate student, and she usually suffers from insomnia, mostly because the exam is approaching and the deadline for submitting homework is imminent. She needs to stay up late to study and work. In daily life, the mobile phone is not left for a long time, it is easy to play the mobile phone and watch videos for too long. At night, the spirit is in a state of excitement. When she realized that she should sleep, she did not have a strong sleepiness. In most cases, she usually goes to bed at 11 or 12 in the evening, depending on the situation in the morning. She wakes up at about 7 o’clock in the morning or when there is something. She can sleep late at about 10 o’clock.

She has not used any sleep-related smart devices, and has accepted the products of wristbands, but also admitted that she does not have the habit of wearing a watch. It feels a bit uncomfortable to wear things while slee**, which may affect the quality of sleep. For products such as bedside lamps, it is considered useful to have different light responses in different situations, such as eye protection lights when reading at night, and sleep assistance lights at night. For the new products of a smart mat, she expects practicality and convenience. At the same time, she wants to ensure that there is no electronic radiation that may harm physical health.

For the post-90s generation who started on the health road, she expressed understanding and support. Because of insomnia and stresses of study and work, acne appeared on her face, she said “the body began to automatically fight against bad habits and required to be taken good care of”. When asked about self-regulation, she thought it was important to develop a regular sleep routine, but she did know it was easy to know and hard to implement.

She said, “Insomnia is a fierce self-resistance. There is a flood of thoughts at night, a variety of beautiful visions of life in the mind, conflicts with various realistic dilemmas and inadequate abilities, huge pressure and confusion swee** through the body. It is impossible to sleep peacefully”. At the same time, she also firmly believes that it is necessary to work hard to cope with sleep problems, reasonably plan future careers, and maintain a relaxed attitude, and everything will gradually get better.

2.3 Competitive Products Analysis

This paper collects 8 typical sleep-related competitive products and evaluates them in four aspects. The evaluation of each competitive product is shown in Table 1.

Table 1. The evaluation of competitive products
Full size table

Based on the above product evaluation table, the two products with the highest scores stand out-Mijia bedside lamp and RestOn smart sleep device, which have the functions of assisting sleep light and monitoring sleep data. Therefore, these two functions can be used as the focus of subsequent design work.

The Mijia bedside lamp has a unique and effective function, which can help to fall asleep, wake up, and adjust the light mode. The price is average and it is fashionable. It does not collect enough sleep data.

The RestOn smart sleep assistance device has a large monitoring area, and the measured data is more accurate and reliable. It does not need to be worn, is easy and comfortable, has dedicated functions, and is expensive.

Combined with the competition product analysis, we can infer the two functions that the user needs most-sleep monitoring and light sleep assistance, while ensuring good practicability.

3 Sleep Data Collection and Analysis

3.1 Sleep Data Collection

The intelligent sleep assistance product designed in this paper uses the Arduino platform to carry out the design of intelligent hardware. The EMFi sensor is used to collect sleep information.

The EMFi sensor is thin and light, and is sensitive to changes in its surface force. It can convert pressure changes into output electrical signals, but the output electrical signals are very weak [10]. The voltage fluctuation range of human body movement is 0–20 mV, while the voltage fluctuation range corresponding to breathing is only in the range of 0–0.3 mV. The change of electrical signals cannot be obtained directly through the Arduino analog input, and it is easy to be interfered by the power frequency of 50 Hz mains. Therefore, more sophisticated data collection and processing circuits need to be designed. By consulting the literature, the overall circuit layout of “EMFi sensor → charge amplifier → filter → AD converter → Arduino development board” was determined [11]. The overall circuit is shown in Fig. 3.

Fig. 3.
figure 3

Circuit layout

Full size image

After some circuit study, we decided to use LM324 quad operational amplifier to design the amplifier circuit, MAX292 eighth-order low-pass filter to design the low-pass filter circuit, and 24-bit AD converter ADS1256 to design the high-precision AD acquisition circuit. After reading and researching the relevant data manual, complete the design of the overall circuit, and cooperate with the EMFi sensor for debugging. We selected the appropriate DC power supply, clock signal, resistance, and capacitance, to complete the design of Arduino program and high-precision data acquisition. The actual effect of the detection circuit is good. it can capture human body movements in time, roughly observe the signal fluctuation caused by breathing, and observe the signal fluctuation caused by heartbeat while holding the breath.

In order to obtain the values ​​of breathing frequency and heart rate from the data analysis of electrical signals, it was decided to use MATLAB for data analysis. The electrical signals in the time domain were subjected to fast Fourier transform (FFT) to obtain the frequency domain distribution map of the electrical signals. (see Eq. 1)

$$ \hat{f}\left( \xi \right) = \int_{ - \infty }^{\infty } {f\left( x \right)e^{ - 2\pi ix\xi } dx} $$
(1)

The crest distribution can accurately obtain the values ​​of respiratory frequency and heart rate. The analysis of each electrical signal and frequency is shown in Fig. 4.

Fig. 4.
figure 4

Electrical signals and frequency

Full size image

3.2 Sleep Data Analysis

According to research, the basis for judging good sleep quality includes 10–20 min of falling asleep, deep sleep, less waking up at night, getting up quickly, and being awake during work during the day and so on. The length of deep sleep largely determines our sleep quality. The deeper people sleep during the night, the higher the sleep quality is.

The determination of the degree of sleep can be processed based on the data collected by the mat, including heart rate, breathing rate, and body movement. To simplify the user’s understanding, we divide the sleep phase into the awake phase, light sleep phase and deep sleep phase. Among them, the user does not fall asleep during the awake phase, and the user may be in a light sleep phase or a deep sleep phase after falling asleep. The condition for the judgment is that when the body movement frequency exceeds 3 times per minute, we think that the user is still awake; when the body movement frequency is less than 3 times per minute, we think that the user starts to fall asleep, and continue to judge whether the user is in the light sleep stage or in the deep sleep phase. When the user’s heart rate falls to 80% before falling asleep, we determine that the user has entered the deep sleep phase, otherwise the user is in the light sleep phase [12]. The sleep phase flow chart is shown in Fig. 5.

Fig. 5.
figure 5

The sleep phase flow chart

Full size image

In addition, the mat can also determine whether the user is lying down and ready to sleep. This can be determined based on the heart rate value. When the user is sitting on the bed for reading or playing with a mobile phone, his heartbeat is difficult to be captured or in a huge fluctuation range because the human heart is far from the mat. When the user is lying in bed and ready to sleep, the back of the human body is close to the mat, and the sensor can accurately capture a stable heartbeat value, so as to make a judgment that the user has been lying down.

4 Intelligent Sleep Assistance System

4.1 Product Form and Interaction Process

After completing the collection and analysis of sleep data, this paper began to establish the brand image of smart sleep assistance products-LotusLeaf sleep assistance equipment, and product form-smart mat combined with smart bedside lamps. The inspiration of LotusLeaf’s brand is derived from the ancient Chinese poetry, “Little lotus leaf only showed sharp pointed corners, and a dragonfly stood on its head”, which depicts a pleasant and peaceful pastoral scene. The natural beauty can bring great relaxation and stress relief.

The specific usage scenarios and interaction processes are as follows:

When the user goes to bed, he can choose to turn on or off the bedside lamp. If it is turned on, he releases soft sleep aids. At the same time, he can choose to play white noise, music or audio programs on the mobile application. The mat will record the sleep data of the user’s body movement, breathing, heart rate from the moment the user goes to bed. It can be determined that the user is in one of the awake phase, light sleep phase, and deep sleep phase based on the body movement and heart rate.

After the user switches from the awake phase to the light sleep phase or the deep sleep phase, the bedside lamp will slowly turn off, and the sleep assistance audio will slowly disappear.

When the user wakes up at night to go to the toilet, the bedside lamp will automatically turn on, and it will automatically turn off after going to bed.

According to the sleep plan and alarm time set by the user on the mobile application, combined with the user’s sleep stage obtained from the mat sensing analysis, in the light sleep moment near the alarm time, the alarm clock will sound and the bedside lamp will shine as the sun rises, awaken users, and welcome a new day.

After the user wakes up, all devices stop monitoring, and the related sleep data will be transmitted to the mobile application via Wi-Fi. The mobile application will analyze and obtain the sleep quality assessment results. The user can view the sleep scores, sleep data, sleep map, sleep analysis information at night, and get suggestions on adjusting personal habits.

During the long-term use, the user can set sleep goals (such as the length of sleep, the time to fall asleep, the time to get up, etc.) according to his own situation and official recommendations. The virtual image lotus leaf will grow with the quality of the user’s sleep. When establishing good sleep habits, the lotus will bloom, and users can get reward medals and physical prizes. At the same time, they can add relatives and friends’ accounts to supervise themselves, and develop good sleep habits. They can also link their elders’ mobile phones with theirs to regularly know the elders’ sleep quality.

4.2 Product Design

After thinking about some complicated or simple design concepts, we decided to stick to the original intention of the product design, to design a non-invasive, comfortable and easy-to-use smart product. The most user-related experience is the feeling of the user slee** on the mat: the thinner and lighter, the more comfortable. Therefore, based on the thinness of the EMFi sensor, we decided to design the product as a thin and light mat covering the EMFi sensor. The wire was connected to the cylinder that wrapped the circuit of the development board. The wire was powered on, and the desk lamp interface was reserved on the cylinder. Choose whether to use a matching table lamp, and the table lamp base and development board circuit are combined in a cylinder, the overall design is simpler. In order to ensure the maximum monitoring perimeter under the smallest area, and combined with the natural shape of the lotus leaf, the final shape of the mat is changed to a circular shape, which is convenient for monitoring user data.

The product design of the LotusLeaf sleep assistance device includes a lotus leaf mat and a bedside lamp, which are simple and elegant in shape. The mat is made of skin-friendly linen material, and a thin EMFi sensor is embedded inside, which can ensure that the user does not feel its presence when slee**; bedside lamp is made of flame-retardant PC material, withstands high strength, and feels soft and comfortable. It integrates all communication modules, speakers, data processing modules and control units, and embeds LED strips with chips in round tubes to switch multi-color. The appearance of the product is shown in Fig. 6.

Fig. 6.
figure 6

The LotusLeaf product rendering

Full size image

4.3 Mobile App Design

LotusLeaf App is a tool to assist the smart mat. It displays and processes data from the smart mat. It aims to let users understand their sleep status and establish good sleep habits. With the main intention and visual element of the lotus leaf, we will be able to analyze and visualize the data collected by the sleep assistance device.

Comprehensive data such as heart rate, breathing frequency, body movement frequency, sitting position in bed, and time can be used to obtain the analysis results of total sleep time, deep sleep time, bed time, falling asleep time, number of waking at night, and sleep pattern. These analysis results finally calculate the user’s total sleep score and rating at night. When the rating is poor, the homepage animation will show the dry lotus with bitterness and sentimentality, so as to remind users to pay attention to improving sleep quality; when the rating is medium, the homepage animation will show the lotus leaves exposed in early summer a cute corner, with a tender green hope, indicates that the user’s sleep quality has improved; when the rating is good, the homepage animation will show a flourishing scene like the lotus leaves shining in the sky, and a beautiful lotus flower embellishes it, which means to the user full affirmation and praise of sleep quality. These animations will have a certain visual impact on the user and give the user intuitive and easy-to-understand visual feedback. At the same time, the user can jump to the analysis of sleep details by clicking the animation. The user can make self-adjustment according to his specific sleep problems and official suggestions, gradually improve his sleep quality, develop good slee** habits, and achieve the ultimate goal of maintaining good health. The application UI design is shown in Fig. 7.

Fig. 7.
figure 7

The UI of LotusLeaf mobile application

Full size image

The main color of the application is blue with high lightness. High lightness conveys a light and lively sensory experience, and blue symbolizes peaceful sleep. The application is divided into three major sections: General, Analysis, and Me.

The General page, which is also the home page, will display comprehensive sleep scores and information such as breathing, heartbeat, and sleep duration, accompanied by “normal”, “low”, and “high”. The user can click to enter the heart rate page, breathing page, and sleep page. Each page has its own data curve and shows sleep-related knowledge.

The Analysis page displays the results of various analysis of sleep data all night. Above is the fluctuation image of the awakening stage, light sleep stage and deep sleep stage with time, and the image visually shows the sleep situation all night. The four columns below show the length of sleep (total sleep and deep sleep), the state of falling asleep (time to go to bed and time to fall asleep), the judgment of waking at night and sleep patterns, which are divided into “normal”, “insufficient”, “late”, “good”. Advice and guidance are given on slee** conditions in the four columns.

The Me page is a collection of user information and settings, including sleep plans, relatives and friends, sleep reminders, FAQs, settings, etc. Users can refer to the application guidance to set a sleep plan, and obtain rewards after completing the corresponding goals. The user can also check the slee** status of relatives and friends, supervise and communicate with each other, and develop good slee** habits.

The LotusLeaf application selects the Android system for software development work, uses the ESP8266 Wi-Fi communication module to communicate with Arduino, and uses the Gizwits IoT platform as a background data transmission hub.

5 Conclusion

This paper takes sleep problems and sleep assistance demands as the starting point, and researches and designs a smart product that can collect sleep data, visualize sleep quality, and assist users to develop good sleep habits. It meets the needs of young people for smart sleep assistance products. Through objective and rational design method, this paper carries out scheme design, the appearance model, the intelligent hardware and the mobile application.

In terms of product form and function, it has carried out in-depth exploration and self-innovation, meeting the functions of sleep monitoring and light assisting sleep, and it can more intelligently detect the user’s sleep at night and give sleep assistance by the light change. The shape of the LotusLeaf mat is beneficial for full contact with the human body and sleep data collection, which is non-invasive. The design of the base can reduce the thickness of the mat, which is convenient for the product to be carried apart. In terms of intelligent hardware, combined with the Arduino platform, EMFi sensor, and ESP8266 module, it realizes the functions of sleep data collection and intelligent interconnection. In terms of the application, the growth and wilt of lotus leaves are used to visualize the user’s sleep quality. The combination of sleep information generates a night sleep map. Sleep quality analysis can help users understand their sleep status and improve their sleep quality.

Sleep healthcare is a social issue that attracts much attention. The design and development of smart home devices based on the IoT and new sensing devices is also a major industry hotspot. This paper’s smart sleep assistance system is a design case based on this background. There are also some deficiencies that the entire system is not fully developed for real use. With regard to the design of emotional care, this paper stays at the level of sleep suggestion feedback and sleep quality visualization. It does not dig deep into the emotional needs behind sleep problems. In the future study, we will use cutting-edge technologies to learn the user’s sleep characteristics, analyze the user’s emotional needs, achieve more effective physical and mental care, and design a more intimate smart sleep assistance system.