Roles and features of smart control and sensing applications for sleep quality improvement: a scoping review

Abstract

Objectives

This scoping review synthesises evidence on the measures and characteristics of the components of combined smart control and sensing technologies, and their impact on sleep quality.

Design

Scoping review following Joanna Briggs Institute (JBI) methodology and reports using the Preferred Reporting Items for Systematic Reviews and Meta-analyses extension for Scoping Reviews (PRISMA-ScR).

Data sources

A comprehensive literature search was conducted in PubMed, Web of Science, EMBASE, Ovid, China National Knowledge Infrastructure, Wanfang Data and VIP Information from the inception of the databases to November 2024 following the PRISMA-ScR statement and updated in June 2025.

Eligibility criteria

This review included peer-reviewed studies evaluating smart home products integrating smart control and sensing technologies to improve sleep quality, with outcomes focused on sleep duration, efficiency or satisfaction.

Data extraction and synthesis

Two independent reviewers screened the title, abstracts and full texts of the selected studies based on the inclusion criteria. Data extraction was performed by two independent reviewers. The data were summarised in tabular format and a narrative summary.

Results

All original studies (N=13) investigated the role and features of these technologies. Seven types of sensors and five smart control methods were identified. These were: biosignal, environmental, chemical sensors, contact and motion sensors, imaging and vision sensors, integrated smart sensors and specialised sensors, along with audio-based, pressure-based, temperature-based, vibration-based and physician-guided control methods. These technologies improved sleep-related health metrics including total sleep time, sleep onset latency, sleep efficiency, deep sleep percentage and subjective sleep quality.

Conclusion

The findings highlight the potential of these technologies for improving sleep, emphasising the role and usability. Future research and product development can build on these insights to design sleep improvement products to innovative, personalised smart home solutions for better sleep.

Ethics and dissemination

As a review, ethical approval is not required. The results from this study will be presented at international conferences and disseminated through peer-reviewed publications. Patients and the public will be involved in the dissemination plans.

Registration details

The Open Science Framework (https://doi.org/10.17605/OSF.IO/FC236).

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• The major strength of the proposed study is that it synthesised a large amount of evidence from different methodologies to provide a completer and more detailed picture of existing smart control measures and sensing technology applications for sleep improvement.

• Another strength of this study is that it highlights the potential of these technologies for improving sleep, emphasising their role and usability and will integrate insights to design sleep improvement products to innovative, personalised smart home solutions for better sleep.

• Another limitation of this review is that it just identifies and maps the available evidence rather than provides a synthesised, clinically meaningful conclusion.

Introduction

Sleep is an important component of the body’s physiological functions.¹ Sleep quality not only affects quality of life but is also closely related to the occurrence and development of various physical and psychological disorders. High-quality sleep helps people recover from physical exhaustion, is beneficial for the maintenance of psychological health and well-being, and improves the overall quality of life.^{2 3} According to the National Sleep Foundation,⁴ sleep quality is a multidimensional concept that includes both subjective satisfaction and objective indicators such as sleep latency, sleep duration, number of awakenings, wake after sleep onset and sleep efficiency. Common sleep disorders—such as insomnia, obstructive sleep apnoea (OSA) and circadian rhythm disturbances—are major contributors to poor sleep quality. Although there is consensus that maintaining good sleep quality—including adequate duration, regular timing and minimal disturbance—is essential for health, sleep problems are prevalent in all age groups.^{5 6} The Sleep Research Report of China in 2023 showed that more than 50% of citizens under the age of 45 have sleep problems.⁷ Even among those aged 25 years and younger, more than 40% experience problems such as difficulty falling asleep or sleep deprivation.7,9 Improving sleep quality has become an important social issue of public concern, and immediate intervention strategies are needed to mitigate sleep problems and enhance sleep quality across the population.¹⁰

Existing intervention studies have mainly focused on improving sleep quality at the individual level, and strategies can be divided into two categories: pharmacotherapy11,14 and mind–body intervention therapies incorporating cognitive–behavioural therapy,15,17 sleep hygiene education,^{18 19} relaxation training and meditation^{20 21} and acupuncture.²² In terms of efficacy, pharmacotherapy relies on the dosage and frequency of medication, making it difficult to balance long-term effects with medication side effects. Although mind–body interventions have demonstrated effectiveness in improving sleep quality, their common drawbacks are also significant, primarily including the slow onset of effects and large differences between individuals. For example, cognitive–behavioural therapy for insomnia treatment has long-term durability, but its treatment efficacy is moderated by patient characteristics such as age or comorbidities, and it typically requires several weeks or longer to produce significant improvements.²³

Over the long term, sleep quality is also greatly influenced by environmental factors, including bedding comfort,^{24 25} sleep position,²⁶ light,^{27 28} sound,²⁹ etc. A familiar, quiet and comfortable bedroom environment is conducive to falling asleep and maintaining sleep.³⁰ In recent years, the rapid development of sensing, chip and artificial intelligence (AI) technologies has laid the foundation for daily sleep health monitoring and improvement. This progress has driven the transition from traditional sleep diagnostics to the concept of ‘active sleep health’, which has become an important development direction.³¹ Ren et al propose a fine-grained sleep monitoring system based on smartphones that simultaneously detects breathing rate and sleep events such as snoring, coughing, turning over and getting up, providing people with non-invasive, continuously fine-grained sleep monitoring.³² Ulfah et al emphasised the importance of circadian regulation and introduced emerging approaches such as precision chronotherapy and AI-driven circadian health monitoring, with key interventions including light exposure control, meal timing optimisation and sleep hygiene enhancement.³³ Ulfah et al emphasised the importance of circadian regulation and introduced emerging approaches such as precision chronotherapy and AI-driven circadian health monitoring, with key interventions including light exposure control, meal timing optimisation and sleep hygiene enhancement. Notably, our previous sleep studies have also demonstrated that bedding and sleep position play important roles in sleep. Sleep posture is closely related to the distribution of body pressure and perceived comfort and that if body pressure can be reasonably coordinated in different postures, it can significantly improve an individual’s perceived comfort and enhance their quality of sleep.³⁴ Therefore, continuous monitoring and dynamic adjustment of pressure or other data on changes in different sleep positions and real-time regulation may be an important point of breakthrough for improving sleep quality.^{35 36}

Smart control refers to the use of automatic adjustments and technologies such as AI³⁷ and machine learning (ML)³⁸ to achieve precise control of environments and systems. Sensing is the pioneer stage of smart control.^{39 40} Sensing technologies for sleep monitoring have made significant progress in recent years, including monitoring of heart rate, respiration and brain waves.⁴¹ The combination of smart control and sensing technology is widely used in small home appliances,^{42 43} automotive seating^{44 45} and other areas, and has greatly improved the functionality and user experience of these products. The combination of smart control and sensing applications allows for precise monitoring, real-time adjustments and targeted optimisation, which has great potential to improve sleep quality. However, existing literature reviews in this domain often focus exclusively on either ‘sensing’ or ‘control’ aspects of sleep improvement. These include reviews of advanced technologies for certain types of sensors;^{46 47} applications of specific sensors for monitoring sleep activity;^{48 49} sensor components for specific needs;50,58 the usefulness of certain types of controls for people with sleep disorders;⁵⁹ and the technological components of certain types of controls.^{51 60} For example, Kato et al developed a sleep-mask-type sensor that employs a microelectromechanical system differential pressure sensor chip to detect minute vibrations on the dorsal surface of the nose, enabling the measurement of pulse and respiratory waves.⁵⁷ To address the challenge of cumbersome wiring in conventional polysomnography (PSG) devices, Wang designed a flexible, integrated multimodal sensing patch based on hydrogel for unconstrained sleep monitoring.⁵⁸ This patch allows for simultaneous sensing of temperature, pressure and non-contact proximity, with no crosstalk between signals due to distinct sensing mechanisms. Zhang focused on developing a novel smart electrically heated sleeping bag.⁶⁰ By incorporating a proportional–integral–derivative heating control system into a traditional sleeping bag, it can maintain human feet temperature within the thermal neutral range, thereby aiding sleep. Overall, there has been no systematic summary of the effects of the application of existing technologies on sleep quality improvement.

In addition, our initial literature scan revealed substantial heterogeneity in study design, outcome measures and interventions for studies using smart control and sensing technologies to improve sleep quality.⁶¹ In contrast to systematic reviews, scoping reviews are used to map evidence in a specific area to answer broader review questions, especially ‘when a set of literature has not been comprehensively synthesised, or exhibits size, complexity or heterogeneity that precludes a more precise systematic review⁶²’. Therefore, we conducted a scoping review to systematically identify and map existing evidence on the combined use of intelligent control and sensing technologies to improve sleep quality. We analysed the role and usability of these approaches. This information can inform future systematic reviews and is essential for guiding future studies on sleep improvement product innovations in smart homes.

Methods

Study design

The purpose of this study was to conduct a scoping review of the existing literature on the features of smart home product designs based on smart control or sensing technologies to improve sleep quality. The structure of this article follows the reporting guidelines for scoping reviews (the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR)).⁶³ The protocol was registered prospectively with the Open Science Framework (Registration DOI: (https://doi.org/10.17605/OSF.IO/FC236 (accessed on 8 January 2025)).

Search strategy

Comprehensive literature searches were conducted on PubMed, Web of Science, EMBASE, Ovid, China National Knowledge Infrastructure, Wanfang Data and VIP Information covering the years from the inception of the databases to November 2024, and updated in June 2025. The core search terms were “smart control”, “sensing technology” and “sleep quality”. We also used an exploratory collection of synonyms and related terms for the core vocabulary using the MeSH (Medical Subject Headings) tree in PubMed to ensure comprehensive literature retrieval.

The identified core keywords and index terms were used for an initial preliminary search followed by a comprehensive systematic search across the databases. The reference lists of all the identified reports and articles were also searched for additional studies. The final search strategy for PubMed can be found in online supplemental table S1. All search strategies can be found in online supplemental table S2.

Data management and study selection

The bibliographic records of all the studies retrieved were first imported into EndNote (V.21.0) and duplicates removed. To identify potentially eligible studies, two trained reviewers independently screened the titles and abstracts of each study based on predefined inclusion and exclusion criteria. After the initial screening, we compared the selections made by the two reviewers and discussed and resolved any discrepancies. Studies that passed this stage were independently reviewed in full by the same two reviewers to determine their eligibility. The reviewers met regularly to discuss progress, clarify the reasons for disagreements and reach a consensus. The PRISMA flow diagram was used to document the study selection steps, reasons for exclusion and final number of studies included (figure 1).

Figure 1. PRISMA flow diagram. CNKI, China National Knowledge Infrastructure; PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses.

Eligibility criteria

Population: The purpose of this review was to explore the efficacy of smart home products combining smart control and sensing technologies for sleep quality improvement. Because these smart home products have a wide audience, we did not limit the inclusion of any information about factors such as the age and physical characteristics of the study subjects.

Concept: Sensing technology captures and collects data such as temperature, humidity, light, sound and human physiological signals on the environment or user status. Based on these data, smart control systems automatically adjust device parameters such as brightness, air conditioning temperature or sound volume to optimise the user experience of environmental comfort and safety. These control functions are typically driven by real-time or adaptive feedback mechanisms, enabling the system to respond dynamically to changes in the environment or user behaviour. Common control strategies include light-based dimming, thermal regulation or auditory masking, which are particularly relevant in sleep-related contexts. In this review, studies were included only if they incorporated both a sensing technology component and a corresponding smart control mechanism aimed at improving sleep quality through environmental or behavioural adjustment. Studies using sensing-only approaches for monitoring, diagnosis or sleep staging, without any form of responsive control or feedback intervention, were excluded. Such technologies enable more efficient and adaptive personalised services and are widely used in smart homes. The outcomes focused on improving sleep-related parameters, such as sleep duration, sleep efficiency and overall satisfaction with sleep.

Context: There were no restrictions on the type or duration of these protocols, nor were there limitations based on cultural factors, geographical location, racial or sex-specific interests, or specific details of the setting in which the research was conducted.

Study design: Only peer-reviewed full-text manuscripts and development or protocol papers were included.

Data management and study selection

We developed the data extraction form based on the PRISMA guidelines and previous research, aiming to comprehensively capture information of interest. This included basic study information (such as author names, publication year and study location), study characteristics (such as study design, research objectives and sample characteristics), participant characteristics (such as participant age, health status and disease description), product design features (such as the types of smart control methods or sensing technologies used) and outcome measures (such as the primary outcome measures and the time points for outcome assessment). To ensure clarity and consistency in classification, study designs were categorised as follows: single-arm pre–post studies assessed outcomes before and after the intervention within the same participant group; pilot studies evaluated feasibility or preliminary effects, typically involving small sample sizes; development studies centred on system design, prototyping or technical validation without formal outcome evaluation; randomised controlled trials (RCTs) compared outcomes between intervention and control groups using parallel-group allocation; and crossover RCTs exposed each participant to multiple interventions in a randomised sequence.

Patient and public involvement

Patients and/or the public were not involved in the design, conduct, assessment of the burden, reporting or dissemination plans.

Results

Included studies

A total of 2272 articles were retrieved from the database searches, and 1036 duplicates were removed, leaving 1236 articles for screening. This process is summarised in figure 1. Subsequently, 1138 articles were removed after applying the inclusion criteria, and the full texts of the remaining 98 articles were screened. Finally, 13 studies met the inclusion criteria: 6 single-arm pre–post studies, 3 pilot studies, 2 development studies, 1 RCT and 1 crossover RCT. Study designs were classified according to predefined methodological criteria. Four trials were conducted in China; two in the USA; two in Italy; and one each in Singapore, Japan, Korea, Spain and Belgium. The features of these 13 studies are summarised in table 1. Although this review did not focus on outcomes because most of the included studies were pilot studies, brief descriptions of the outcomes are provided in the table.

Table 1. Characteristics of included studies.

Author and year	Origin	Study design	Population and sample size	Target group	Sensor application options	Smart control methods	Action mechanisms	Sleep improvement indicators
Zhang et al, 201573	Singapore	Single-arm, pre–post study	Adults (N=28): male (n=16); female (n=12)	Healthy people	Single-channel EEG. (Sleeping mask)	The appropriate audio stimulation. (Stereo earphones)	Based on the online EEG computing algorithm, the user’s sleep state is detected and appropriate audio stimulation is provided to shorten the time to sleep onset.	Sleep onset latency
Zhao et al, 202074	China	Pilot study	Adults (N=2)	Healthy people	Light intensity sensor; Temperature and humidity sensor; Facial expression recognition sensor.	The device automatically plays corresponding music.	Regulating heart rate and breathing, improving the user’s mood and promoting sleep quality.	Proportion of deep sleep; Subjective sleep quality
Wei et al, 202375	China	Single-arm, pre–post study	Adults (N=20): male (n=10); female (n=10)	Healthy people	Air pressure sensor; EEG	Adjusting the pressure in three body zones in real time using the auto-air mattress.	Keeping the spine in normal physiological curvature through comfortable pressure support, which in turn improves sleep quality.	Total sleep time; Sleep efficiency; Proportion of deep sleep; Subjective sleep quality
Verhaert et al, 201376	Belgium	Single-arm, pre–post study	Adults (N=30): male (n=16); female (n=14)	Healthy people	Linear displacement sensor/linear potentiometers. (Mattress indentation measurements)	New bed configuration. (Active mechanical adjustment of mattress firmness in eight comfort zones)	Spinal alignment is significantly improved, thus positively affecting sleep.	Sleep efficiency; Subjective sleep quality
Van der Loos et al, 200364	USA	Pilot study	Not mentioned	People who snore or suffer from OSA	Force sensitive resistors; Resistive temperature devices; Microphone/acoustic sensor; Potentiometers/displacement sensor.	Automatic adjustment of the bed frame.	Diagnosis and alleviation of mild sleep disorders.	Duration and frequency of snoring
Tang et al, 202372	Japan	Single-arm, pre–post study	Older people (N=18): male (n=11); female (n=7)	Older people	A telehealth monitoring system centred around an electronic water metre; A sleep monitoring device: heart rate, respiratory rate, body movement, the status measured when the user is lying on the bed.	The telehealth system.	Improving people’s health, predicting health status and the risks of lifestyle-related diseases, and addressing the shortage of doctors, nurses and home caregivers.	Total sleep time; Sleep onset latency; Proportion of deep sleep
Liao et al, 202465	China	Development study	Not mentioned	People who snore	An IoT snore tracker contains a detection microphone/acoustic sensor.	The acoustic-controlled pillow with IoT features provides soothing sounds.	Reducing the effects of intrusive ambient noise, which leads to relaxation and sleep induction.	Sleep onset latency
Kim et al, 201966	Korea	Development study	Not mentioned	Older people or healthcare users	The smart mat system contains gyro sensors; the human body sensing pressure sensor.	The telehealth system intelligently sets the temperature at the optimal level for sleeping.	Improving people’s health and addressing the shortage of doctors, nurses and home caregivers.	Temperature control
Hu et al, 202167	Italy	Single-arm, pre–post study	Older people (N=19)	Older people	An environmental sensor package: passive infrared motion sensors; contact-based door sensors; pressure-based bed and chair sensors; and proximity-based toilet sensors.	A remote healthcare service system.	Improving elderly patients’ quality of life and alleviating pressure on the healthcare system.	Sleep continuity; Nap variables
He et al, 202268	China	Pilot study	Adults (N=1)	People who snore	A novel low-cost flexible patch with MEMS microphone and accelerometer.	Snoring suppression using a small vibration motor.	Snore suppression.	Duration of snoring
Ferrer-Lluis et al, 202169	Spain	Single-arm, pre–post study	Adults (N=9)	People who snore or suffer from OSA	An Android smartphone containing an accelerometry sensor; Oximetry.	The vibration.	Positional therapy using smartphones helps reduce supine position, improves ODI severity and is effective for pOSA patients.	Percentage of time spent in supine position
Donati et al, 202170	Italy	Randomised controlled trial	Newborns (N=45)	Newborns with CHD	Biomedical sensor: ECG; heart rate; oxygen saturation; body temperature; body weight.	Telemedicine is born.	Improving quality of life for the whole family and reducing hospital admissions.	Subjective sleep quality
Bogan et al, 201771	USA	Crossover randomised controlled trial	Adults (N=65)	People who suffer from OSA	CPAP device/air pressure sensor.	The adjustable SensAwake Pressure.	Reducing sleep-disordered breathing events, effectively controlling OSA and improving sleep quality and insomnia.	Subjective sleep quality

CHD, congenital heart disease; CPAP, continuous positive airway pressure; ECG, electrocardiogram; EEG, electroencephalogram; IoT, Internet of things; MEMS, microelectromechanical system; ODI, oxygen desaturation index; OSA, obstructive sleep apnoea.

Composition of participants and timing of interventions

As shown in table 1, a total of 237 participants were included in the examined studies, with sample sizes ranging from 1 to 65 per study. Three studies did not report the characteristics of the study population;64,66 five studies recruited participants with sleep problems;67,71 three studies required the inclusion of special populations, such as the elderly or children;^{67 70 72} and the other four studies did not restrict the population.73,76 Regarding the timing of interventions, two studies reported continuous monitoring for up to 1 year or more,^{67 72} and three studies ranged from 0.5 to 6 months.^{68 70 71}

Main approaches and technical characteristics of smart control and sensor applications

An overview and summary of the sensor application and smart control components are presented in online supplemental table S3. Regarding the technologies used in the 13 included studies, 7 main sensor applications were identified, including biosignal, environmental, chemical sensors, contact and motion sensors, imaging and vision sensors, integrated smart sensors and specialised sensors; and 5 primary smart control methods were identified: audio-based, pressure-based, temperature-based, vibration-based and physician-guided control methods. Approximately three-quarters of the included studies (n=11) incorporated at least two types of sensor application. One study incorporated a combination of the two control methods, whereas the remainder used single-category control methods.

Sensor applications components

Biosignal sensors (n=6)

Biosignal sensors were used in six studies to monitor physiological signals in the human body to form intuitive sleep datasets that were used to track long-term changes in sleep, related trends and core tools for sleep quality and health assessment.^{64 69 70 72 73 75} These include electroencephalogram (EEG),^{73 75} electrocardiogram (ECG),^{70 75} electromyogram (EMG),⁷⁵ oximetry^{69 70} and respiratory sensors.^{64 70 72} The data measured by the respiratory sensors do not require secondary conversion and can directly constitute a sleep data set. In contrast, EEG, ECG and EMG data can typically be recorded by PSG equipment or specific single-channel devices and later reconverted according to specialised techniques (eg, support vector machines) or standards (eg, the American Academy of Sleep Medicine standards) to form intuitive sleep study data.⁷³

Environmental sensors (n=6)

Studies used temperature,^{64 66 70 74} humidity,⁷⁴ acoustic^{64 65 68} and light sensors⁷⁴ to monitor humidity, light and noise in the sleep environment in real-time. These data are commonly used to construct sleep environment data, and a microcomputer calculates the difference between them and the environmental parameters required for sleep quality,⁷⁴ to activate the sleep-aid system. In addition to recording audio directly, acoustic sensors can also process based on acoustic properties, rules or algorithms. Taking snoring as an example, the time-domain and frequency-domain signals of its audio are highly regular and very different from other sounds (eg, music, conversation and footsteps) and are often used as a basis for detection of snoring.^{64 68} Furthermore, it is also possible to extract sound features by employing mel-frequency cepstral coefficients and leveraging a classification model to effectively differentiate snoring from other sounds, thus enhancing the accuracy of snore detection.⁶⁵

Contact and motion sensors (n=7)

Contact and motion sensors were used in seven studies to provide indirect indicators of sleep behaviour and physiological states by capturing body movements, posture and contact pressure.^{64 66 67 69 72 75 76} Specifically, they included body movement,^{66 67 72 76} pressure^{64 67 75} and posture sensors.⁶⁴ Their advantages over non-contact sensors are high accuracy and stable data. The most widely used of these is the thin-film pressure sensor, which provides data on changes in sleep position,⁶⁴ sleep duration⁶⁶ and their trends over time,⁷² and can provide medical caregivers with clear information about the patient’s health status trajectory during sleep improvement measures.⁶⁷ Air pressure sensors in auto-air mattresses detect the presence or absence of people, as well as the sleeper’s body type and sleep posture, through changes in the pressure value of each air cell. Corresponding to the control level, the pressure of each air cell is adjusted at a relatively slow rate, thus providing comfortable pressure support that can improve sleep efficiency.⁷⁵

Chemical sensors (n=1)

One study used chemical sensors to dynamically monitor carbon dioxide concentration and sweat composition.⁷¹ In this study, chemical sensors were placed in a continuous positive airway pressure (CPAP) device to dynamically monitor respiratory patterns and carbon dioxide concentrations and avoid sleep disruption or respiratory depression due to conditions such as hyperventilation.

Imaging and vision sensors (n=3)

Imaging and vision sensors refer to technologies that monitor sleeping position, breathing and activity in a non-contact manner through vision-related technologies such as cameras,⁶⁴ infrared,⁶⁷ depth⁷⁴ and thermal imaging. These sensors are particularly suitable for sleep monitoring in children, older adults and other sensitive individuals.

Integrated smart sensors (n=5)

In addition to single-attribute types of sensors, integrated smart sensors combine the monitoring capabilities of pressure, temperature and biosignals to achieve all-round tracking of sleep-related data.^{64 66 70 72 74} Smart mattresses, wearables and Internet of things with integrated sensors not only act as the data collection end of the telehealth system but also transmit data to a remote server, such that the combination of hardware and software enables the integration of data collection, analysis and feedback.

Specialised sensors (n=5)

Five studies provided customised sensors for groups with special needs such as snoring, teeth grinding and pressure distribution imbalance.^{64 65 68 70 71} These sensors can accurately detect snoring^{65 68} or abnormal breathing patterns⁶⁴ and drive appropriate devices to provide timely interventions that can effectively improve sleep quality.

Smart control components

Audio-based control (n=3)

Three studies used audio controllers to provide specific sound stimuli such as soothing music or nature sounds to promote relaxation, relieve anxiety and reduce stress, thereby improving sleep quality.^{65 73 74} These control methods are particularly helpful for individuals who have difficulty sleeping because of anxiety or environmental noise.

Pressure-based control (n=3)

Pressure-based control refers to the improvement of body support and comfort by dynamically adjusting the pressure distribution of the bed or mattress, particularly for users who are bedridden for long periods or have specific comfort needs. Three included studies used mechanical or pneumatic pressure adjustments for pressure control.^{71 75 76} The difference is that mechanical adjustment adjusts the postures of the bedboards of different partitions, whereas air pressure adjustment adjusts the air pressure of each cell within the air mattress bed. Both provide a better quality configuration solution to provide comfortable pressure support for the human body, thus improving sleep efficiency.^{75 76} In addition, pressure-based control also involves the air acting inside the body; for example, when CPAP is used for the treatment of OSA, it is possible to control the air pressure to avoid discomfort or awakening of the patient and to improve their compliance.⁷¹

Temperature-based control (n=1)

Temperature-based control refers to the automatic adjustment of bedroom or bed temperature to optimise the sleep environment based on an individual’s sleep needs. Kim and Jeong⁶⁶ automatically adjusted smart temperature-controlled mats and air conditioners according to user needs and environmental temperature changes, thus providing the most suitable temperature conditions for sleep.

Vibration-based control (n=3)

One study used low-frequency vibrations to treat specific sleep problems or provide a soothing effect.^{64 68 69} Vibrating mattresses or pillows sense the user’s sleeping position changes or snoring episodes and activate slight vibrations to induce the user to change positions and reduce snoring. In addition, vibration-based control can provide a massage or soothing effect through vibration to relieve anxiety or tension, thus helping the user relax and fall asleep.

Physician-guided control (n=4)

Physician-guided control combines sensed data collection and processing with the expertise of healthcare professionals to provide personalised care solutions.^{66 67 70 72} It ensures accurate advice, decision-making and adaptive interventions, particularly in complex or chronic care scenarios. In the included literature, this type of control is also transitional to a semiautomated mode; however, it can be combined with unsupervised learning, large-model training or other techniques to achieve automated control.

Categories of oriented sleep quality outcomes

Table 2 summarises the outcome indicators related to sleep quality in the included studies. Based on the outcome indicators used for sleep quality improvement in the 13 included studies, 7 main types of sleep quality improvement indicators were identified: total sleep time (TST), sleep onset latency (SOL), sleep efficiency (SE), proportion of deep sleep, subjective sleep quality, symptom relief and other indirect improvements.

Table 2. Included studies, study designs and sleep quality improvement effects (N=13).

Author and year	TST	SOL	SE	Proportion of deep sleep	Subjective sleep quality	Symptom relief	Other indirect improvement
Single-arm pre–post studies
Zhang et al, 201573		Significantly reduced the SOL for people with difficulty in falling asleep (p=1.22e−04, Wilcoxon signed-rank test).
Wei et al, 202375	Significantly increased TST for female participants (p=0.035).		Significantly improved SE for female participants (B=2.666, sig=0.033).	Significantly decreased non-rapid eye movement N2 proportion (B=2.666, sig=0.033), and increased non-rapid eye movement N3 proportion (B=2.8, sig=0.078). Decreased ANS activity during N3 sleep.	Significantly improved the self-rated sleep quality for female participants (p=0.025).
Verhaert et al, 201376			Spinal alignment was significantly improved in the active support condition.		The group with active control of bed attributes (7.00±0.87) scored higher on subjective perceived sleep quality than the reference night group without control of bed attributes (5.67±1.41).
Tang et al, 202372	Increased total sleep duration.	Shorter sleep latency.		Average heart rate and respiratory rate stabilised, and the number of body movements decreased.			Offered help, suggestions and care plans based on disease risk prediction (including sleep disorders such as OSA).
Hu et al, 202167			Long sleep continuity; Frequent nap variables.				Better adherence, incorporating medically advised interventions into the care system, and improving overall quality of life.
Ferrer-Lluis et al, 202169						The percentage of time subjects spent supine before and after postural therapy (45.6% → 2%).
Pilot study
Zhao et al, 202074				The users’ deep sleep time and deep sleep frequency increased significantly.	Sleep score increased and the effect was more obvious (81 → 92; 92 → 98).
Van der Loos et al, 200364						Gently encourages a person to roll over to alleviate snoring and OSA.
He et al, 202268						The average snoring time (135 → 15 min).
Development study
Liao et al, 202465		Effectively promoted restful sleep and improved nightly rest in noisy environments.
Kim et al, 201966							Set the temperature at the optimal sleep condition and monitored the sleep condition based on the telemedicine system.
RCT
Donati et al, 202170	Sleeping hours (6 hours to almost 8 hours)				PGWBI: 3.4 → 3.9. PSQI: 1.9 → 2.3. PedsQL: 1.3 → 0.7.
Crossover RCT
Bogan et al, 201771					FOSQ-10. SensAwake ON compared with baseline (↑, 4.63±0.99). SensAwake OFF compared with baseline (↑, 6.47±0.90).

ANS, autonomic nervous system; FQSQ-10, Short Functional Outcomes of Sleep Questionnaire; OSA, obstructive sleep apnoea; PedsQL, Paediatric Quality of Life Inventory; PGWBI, Psychological General Well-Being Index; PSQI, Pittsburgh Sleep Quality Index; RCT, randomised controlled trial; SE, sleep efficiency; SOL, sleep onset latency; TST, total sleep time.

Wei et al demonstrated that a smart adjustable zoned air mattress based on biosignal sensors and contact and motion sensors with a pressure-based control method performed better in terms of TST (↑, p=0.035), SE (↑, B=2.666, sig=0.033), percentage of deep sleep (N2↓, B=2.666, sig=0.033, N3↑, B=2.8, sig=0.078) and subjective sleep quality (↑, p=0.025).⁷⁵ Zhang et al⁷³ and Liao et al⁶⁵ showed that the audio-based smart control method has significant effects on SOL (↓, p=1.22e−04), especially for people who have difficulty falling asleep. Subjective sleep quality is an intuitive indicator for evaluating sensor-driven control methods, including the composite satisfaction score,74,76 the Psychological General Well-Being Index, the Pittsburgh Sleep Quality Index, the Paediatric Quality of Life Inventory⁷⁰ and the Short Functional Outcomes of Sleep Questionnaire,⁷¹ which are used flexibly depending on the focus. The studies of Ferrer-Lluis et al,⁶⁹ Van der Loos et al⁶⁴ and He et al⁶⁸ showed that vibration-based control is a powerful measure for symptomatic relief, specifically snoring symptom relief. The studies by Tang et al,⁷² Hu et al⁶⁷ and Kim and Jeong⁶⁶all involved the physician-guided approach, and the improvement indicators were not quantified.

Discussion

Main findings

This scoping review identified and mapped the available evidence on the combination of smart control and sensor technology applications to improve sleep quality. 13 full-text studies from eight countries were included in this review. Seven different sensors and their characteristics were identified based on their active components: biosignal, environmental, chemical sensors, contact and motion sensors, imaging and vision sensors, integrated smart sensors and specialised sensors. In addition, five categories of smart control methods and their characteristics were summarised based on their active components, including audio-based, pressure-based, temperature-based, vibration-based and physician-guided control methods, although of course there is perhaps some crossover between these categories. Based on these results, seven types of sleep quality improvement indicators were categorised: TST, SOL, SE, proportion of deep sleep, subjective sleep quality, symptom relief and other indirect improvements. However, studies on the selection of appropriate smart control methods for improving sleep based on sensor-conducted measurements of sleep status are limited.

The findings of this review indicate that the increasing number of studies on sleep quality improvement using sensor applications and smart control, especially in the form of increasingly humanised technological implementations, is inextricably linked to advances in technological development. The categorisation of sensor applications and smart controls in sleep improvement processes by active components is a prerequisite for identifying which components or combinations of components influence the effectiveness and usability of sleep improvement solutions. This is important for establishing a foundation for future sleep product innovations in the smart home arena. Sensor technology provides accurate real-time data support for smart control, which can dynamically sense a user’s physiological state and environment to help assess sleep stages, sleep quality and the presence of potential sleep disorders. Currently, these sensing technologies can be used in combination with or integrated into instruments such as PSGs or consumer products such as smartphones and smartwatches.⁷⁷ Correspondingly, smart control strategies use sensor data to achieve environmental regulations or personalised interventions.^{76 78} In addition, AI and ML make it possible to learn from an individual’s sleep patterns⁷⁸ and even dynamically adjust sleep intervention strategies based on an individual’s physiological changes, to achieve improvements in sleep quality. While sensing technologies alone enable the observation and recording of sleep-related parameters, their role is primarily diagnostic or retrospective in nature. In contrast, the integration of sensing with smart control mechanisms transforms passive monitoring into an interactive system capable of immediate adaptation. Rather than simply detecting sleep problems, combined systems aim to prevent or alleviate disturbances in real time—shifting the focus from measurement to regulation. This transition fundamentally changes the system’s function from descriptive to prescriptive. Several studies in this review demonstrated how such integration allows timely adjustments to lighting, sound or temperature based on sleep phase or detected restlessness, which may contribute to improved sleep continuity and comfort. This effect is further supported by recent advances in sensing modalities—such as non-contact motion detection and multimodal capacitive sensors—which not only reduce user burden but also generate more reliable and actionable data streams for control decision-making.

During the exclusion process of some of the studies, we found that most of the studies focused on the sleep monitoring stage. Although these studies collected a large amount of data on the sleep cycle, rest rhythms, disease potential, etc, they lacked integration with responsive control mechanisms capable of addressing the issues identified. For instance, although well-established wearable devices such as the Oura ring and actigraphy-based monitors are widely applied in sleep research, their use is primarily limited to passive tracking rather than driving intervention. As our review specifically focused on systems that actively link sensing and control to improve sleep outcomes, such studies fell outside our inclusion criteria. This disconnect from an actionable feedback loop—enabling ‘monitoring-improvement-re-monitoring’—was not fully anticipated during our initial review planning. Unfortunately, some of the included studies reported limited information on the characteristics of sleep interventions, which may pose additional challenges for researchers in synthesising the available evidence. It is recommended that future researchers actively structure an all-links thinking that is not limited to the disciplinary domains of clinical medicine, medical technology and mechanical engineering, and that more linked studies are needed to focus on this issue.

Overall, the evidence on the effect of different combinations of sensing and control measures for sleep improvement is limited. We were unable to draw definitive conclusions regarding the effectiveness of the different measures based on the results of the included studies because of the lack of well-established methodologies for sleep quality assessment and advanced data analysis and processing techniques that are in line with the recommendations of the scoping review regarding the extraction, analysis and presentation of results. Specifically, six single-arm pre–post studies and two RCTs provided significant supporting evidence, while the pilot studies had insufficient capacity to test the intervention hypotheses because of the limited experimental description of the measures studied and the small sample sizes, which may have resulted in imprecise effect size estimates and should, therefore, be interpreted with caution. The remaining developmental studies could not test the effectiveness of interventions, although they often provided supporting evidence that contributed to a better understanding of the effectiveness of the interventions. Moreover, in some studies, the quantitative relationship between the effectiveness of sleep quality improvement and intervention measures was ambiguous, mostly because the assessment methods were limited to subjective perceptions⁷⁶ and did not incorporate appropriate objective monitoring methods.^{69 70} In addition to temporal indicators that are used to measure better sleep quality, such as sleeping longer, falling asleep faster, sleeping more efficiently and having a higher proportion of deeper sleep,6870 73,75 vital sign indicators, such as heart rate, respiratory rate and body movement, are also included or integrated into sleep monitoring devices placed next to the bed or under the mattress.⁷² Telehealth systems are also effective for sleep quality improvement and have the potential to improve quality of life and promote health.^{66 67 72} Future high-quality intervention studies can be conducted based on the aforementioned research content and outcome indicators to verify their effectiveness.

Usability is also a critical consideration for sleep improvement measures. If the usability experience of the measures is positive, participants are more likely to use a sleep aid or receive sleep intervention as required and will be more compliant. Overall, the findings were generally positive concerning the usability of sleep improvement measures, particularly bed/mattress^{75 76 79 80} and telehealth systems,^{70 72} which did not interfere with users’ daily lives or compromise their privacy,⁶⁷ resulting in better adherence. However, for sleep disorders such as snoring and OSA, treatment outcomes tend to be prioritised over usability and comfort, and even though studies have been conducted on CPAP for pressure reduction and pressure relief modulation, adherence has not improved significantly.⁷¹ Of course, there are also portable devices for adjunctive treatments that are worth investigating if they are downgraded from ‘solution’ to ‘relief’.⁶⁸ Previous studies have shown that soothing audio stimulation (eg, soft music) can improve sleep quality.^{65 81 82} In addition, consumer products are promising mHealth tools that can be combined with specialised devices such as pulse oximetry and CPAP to improve adherence.⁶⁹ In summary, the complexity of smart controls and the comfort of wearing sensors are important aspects that may affect a user’s long-term use.

The potential of smart home products, which combine smart control and sensor technology, is mainly reflected in the optimisation of environmental factors such as temperature and humidity,⁸³ personalised adjustment of devices⁸⁴ and forecasting of potential risks,⁸⁵ which not only helps improve sleep comfort but also contributes to long-term health management.⁸⁶ However, after our preliminary study, we found that smart home products, that is, the combined application of smart control and sensing technologies, still have some challenges, mainly in terms of insufficient personalisation of interventions, untimely control responses and poor user compliance, which can make sleep improvement much less effective. Some control methods rely on predefined rules and do not consider user preferences, behavioural patterns or long-term data, making it difficult to provide precise interventions for different users, especially when their sleep needs are complex or abnormal. Evaluations of effect also lack scientific and harmonised conclusions. In addition, data can be redundant, conflicting or noisy when coordinated across multiple sensors, which can lead to the sluggish execution of control methods. The sleep improvement effect may also be somewhat biased compared with the results of previous studies owing to an insufficient number of participants.⁷⁵ Future studies could tailor sleep improvement measures to specific subpopulations such as women, older adults and patients with chronic diseases such as coronary heart disease. Furthermore, most of the included studies were conducted in developed countries, where the sleep status of people is also affected by their lifestyle, living environment, etc. There is a need for more independent studies that target developing countries, which account for a larger proportion of the world’s population.

Although there is a wide variety of existing sensing and control methods, new technologies from other fields such as virtual reality, digital twins and large language modelling could also be used to design new technological solutions for sleep improvement and rehabilitation. In addition, many existing studies are still in the laboratory with an insufficient duration of control measure interventions. Therefore, more attention needs to be paid to the effect of smart home systems in the home or long-term care. Privacy protection mechanisms should also be strengthened to address data protection issues that are often easily overlooked.

Strengths and limitations

This scoping review synthesised a large amount of evidence from different methodologies to provide a completer and more detailed picture of existing smart control measures and sensing technology applications for sleep improvement. However, this study had several limitations. The exploratory and descriptive nature of our research questions resulted in significant heterogeneity across study designs, intervention components, target populations and outcome measures, which limited our ability to comprehensively evaluate the effect of the interventions. However, because assessing the methodological quality of the included studies is not mandatory for scoping reviews, no such assessment was performed in this review. The primary purpose of this scoping review was to identify and map the available evidence rather than to provide a synthesised, clinically meaningful conclusion.⁶¹ Additionally, this review exclusively incorporated peer-reviewed articles; future research will further explore significant insights from other types of grey literature (eg, technical reports, conference proceedings) concerning the application of smart control and sensor technologies in improving sleep quality, so as to offer a more holistic overview of the evidence landscape in this field.

Conclusions

This scoping review investigated the roles and features of sensors and smart control strategies on metrics related to improving sleep quality. Seven major sensor application approaches and five primary smart control methods were identified, including biosignal, environmental, chemical sensors, contact and motion sensors, imaging and vision sensors, integrated smart sensors and specialised sensors, as well as audio-based, pressure-based, temperature-based, vibration-based and physician-guided control methods. Utilisation of these sensors and smart control strategies has improved sleep-related health metrics such as TST, SOL, SE, percentage of deep sleep and subjective sleep quality. The findings of this review highlight the great potential of smart control and sensing technologies for sleep improvement and focus on the effect and usability of their methodological measures. In the future, researchers and project designers in the smart home field can use this as a basis for the heuristic development and design of sleep improvement products to further support and enrich the existing evidence.

Data availability statement

Data are available upon reasonable request. All data relevant to the study are included in the article or uploaded as supplementary information.