The Development of Non-contact Heart Rate Monitoring Devices Using Radar Technology

Advancements in healthcare technology have ushered in an era where vital signs can be measured without the need for wires, electrodes, or even physical contact. Among the most promising innovations is the use of radar technology for non-contact heart rate monitoring. These devices leverage the ability of radio waves to detect minute mechanical movements on the body surface caused by cardiac activity. By analyzing the phase and frequency changes in reflected signals, radar systems can extract heart rate data with remarkable accuracy. This approach offers a hygienic, comfortable, and continuous monitoring solution suitable for hospitals, sleep labs, elderly care facilities, and even home environments. Unlike traditional electrocardiogram (ECG) patches or chest straps, radar-based monitors do not irritate the skin, pose no risk of allergic reaction, and can function through clothing and bedding. As the global population ages and the demand for remote healthcare grows, radar heart rate monitoring is poised to become a standard tool in both clinical and consumer health settings.

The journey from a laboratory curiosity to a viable medical device has been driven by advances in radar hardware miniaturization, signal processing algorithms, and machine learning. Today, several commercial products have received regulatory clearance for medical use, and ongoing research continues to refine their accuracy and expand their applications. This article explores the underlying principles, practical advantages, challenges, and future directions of radar-based heart rate monitoring, providing a comprehensive overview for clinicians, engineers, and informed consumers alike.

Understanding Radar-Based Heart Rate Monitoring

Radar technology in healthcare operates on a simple yet powerful principle: transmitted electromagnetic waves are reflected by moving objects, and the characteristics of the reflected wave encode information about that movement. In the context of heart rate monitoring, the moving object is the chest wall, which displaces by roughly 0.5–1.5 millimeters with each heartbeat. This minuscule mechanical motion is the target that radar systems detect, isolate, and convert into a beat-to-beat heart rate signal.

Most modern radar heart rate monitors operate in the microwave frequency range, typically between 2.4 GHz and 10 GHz, or at ultra-wideband (UWB) frequencies. These frequencies penetrate clothing and light bedding while being safe for continuous human exposure at the low power levels used (well below FCC limits). The choice of frequency band influences penetration depth, spatial resolution, and sensitivity to motion artifacts. For instance, higher frequencies provide better resolution but are more attenuated by obstacles, whereas lower frequencies penetrate deeper but may capture more extraneous movement.

Doppler Radar vs. FMCW vs. UWB

Three primary radar architectures are employed for vital sign monitoring: continuous-wave (CW) Doppler radar, frequency-modulated continuous-wave (FMCW) radar, and ultra-wideband (UWB) impulse radar. Each has distinct strengths and trade-offs.

  • Continuous-Wave Doppler Radar transmits a single-frequency signal and extracts heartbeat information from the phase modulation of the reflected wave. It is simple, low-cost, and power-efficient, making it suitable for compact consumer devices. However, it cannot directly measure the distance to the subject, and it struggles to separate heartbeat from respiration and body movements in cluttered environments.
  • Frequency-Modulated Continuous-Wave (FMCW) Radar sweeps its frequency over time, allowing the device to measure both the range and the velocity of targets. This range resolution helps isolate chest motion from other moving objects in the room (e.g., a ceiling fan or a pet). FMCW systems are more complex but offer better rejection of stationary clutter and can simultaneously monitor multiple people if their distances differ.
  • Ultra-Wideband (UWB) Impulse Radar emits very short pulses (picoseconds to nanoseconds) across a broad spectrum. The reflected pulse train is analyzed in the time domain, providing high range resolution (centimeters) and the ability to detect very small displacements. UWB is particularly effective at rejecting multipath interference and can even track heart rate through thick walls or mattresses. It is often used in through-the-wall sensing and automotive occupancy detection, and an increasing number of medical prototypes leverage UWB for contactless monitoring.

Regardless of the radar type, the raw signal must be processed to extract a clean heart rate waveform. The small amplitude of chest displacement (0.5–1.5 mm) means the phase change in the reflected signal is on the order of a few degrees, requiring sensitive analog front ends and high-resolution analog-to-digital converters. The heartbeat signal is typically superimposed on a much larger respiratory signal (chest displacement 4–12 mm). Sophisticated filtering techniques, including adaptive notch filters and wavelet denoising, are employed to separate the two signals before peak detection or frequency-domain analysis yields the heart rate.

Advantages of Radar-Based Devices

Radar-based heart rate monitors offer compelling benefits over contact-based alternatives such as ECG patches, chest straps, oximeters, and photoplethysmography (PPG) watches. These advantages make them especially valuable in scenarios where patient comfort, hygiene, and long-term wear are priorities.

  • Non-invasive and Contactless: No electrodes, gels, or straps touch the skin. This eliminates irritation, allergic reactions, and the need for daily cleaning or replacement of adhesive patches.
  • Continuous Monitoring: Radar can operate 24/7 without disturbing the subject. It is ideal for sleep monitoring, neonatal intensive care (where fragile skin cannot tolerate adhesives), and long-term observation of patients with chronic conditions such as atrial fibrillation or heart failure.
  • Hygienic: Because there is no physical contact, the risk of cross-contamination or hospital-acquired infections is reduced. This is critical in burn units, isolation wards, and during infectious disease outbreaks like influenza or COVID-19.
  • Versatile Installation: A single radar module can be embedded in a wall, ceiling, bed frame, or chair. It does not require the subject to wear or hold anything, making it unobtrusive and suitable for patients with dementia or cognitive impairments who may remove wearable sensors.
  • Privacy-Preserving: Unlike camera-based non-contact methods (e.g., remote PPG using a video camera), radar does not capture any visual information. The system only records motion signatures, addressing privacy concerns that can arise in hospital rooms or homes.

These advantages have driven adoption in specialized settings. For example, radar-based monitors are now used in some hospitals to track heart rate and respiration in patients under MRI (where leads and cables can pose RF safety hazards). They are also embedded in smart mattresses for sleep-stage analysis and in automotive cabins to detect driver drowsiness or medical emergencies.

Applications and Clinical Studies

The versatility of radar heart rate monitoring has led to a wide range of applications spanning acute care, chronic disease management, wellness, and safety. Below we examine several key use cases supported by recent research and product development.

Hospital and Clinical Settings

In hospitals, continuous monitoring of vital signs is essential for early detection of deterioration. However, traditional wired monitors restrict patient mobility and can cause skin breakdown with prolonged electrode use. Radar systems mounted above the bed can track heart rate, respiratory rate, and even body movements indicative of seizures or agitation. A study published in IEEE Transactions on Biomedical Engineering demonstrated that a 2.45 GHz Doppler radar achieved a median heart rate error of less than 2 beats per minute compared to a reference ECG in supine patients during quiet breathing. Another clinical trial at a major academic medical center validated a UWB radar system against standard telemetry monitors in a cardiac step-down unit, showing >90% agreement in heart rate measurements even when patients were sitting up or reading.

Radar is particularly advantageous in neonatal intensive care units (NICUs). Preterm infants have sensitive skin that can be damaged by adhesive electrodes, and the small size of their chests makes wearable sensors challenging. Several NICU pilots using low-power 5.8 GHz CW radar have reported accurate heart rate detection with zero skin injury and fewer false alarms compared to commercial respiration monitors. These systems can also detect apnea events where an infant stops breathing, prompting immediate staff attention.

Sleep Monitoring and Sleep Apnea Detection

Sleep studies traditionally require the patient to spend a night in a specialized lab connected to electroencephalography (EEG), ECG, pulse oximetry, and respiratory belts. This environment is unnatural and can disrupt sleep architecture. Radar-based home sleep monitors offer a solution by unobtrusively measuring heart rate, breathing patterns, and movement throughout the night from a device placed beside the bed or under a mattress. Advanced signal processing can estimate sleep stages (light, deep, REM) by analyzing heart rate variability (HRV) and respiratory regularity. Some systems have been validated against polysomnography for heart rate and apnea-hypopnea index, receiving Class II FDA clearance for screening purposes. For patients with obstructive sleep apnea, radar can flag periods of respiratory effort without airflow, prompting further diagnostic evaluation.

Geriatric Care and Fall Detection

Elderly individuals living alone face risks of falls and sudden medical events. Radar systems installed in a home or assisted living facility can monitor heart rate and detect falls by analyzing the Doppler signature of a sudden, rapid downward motion. When a fall is detected and accompanied by an abnormal heart rate (e.g., bradycardia or tachycardia), the system can automatically alert caregivers or emergency services. Unlike wearable fall pendants, radar does not require the person to remember to wear or activate it. Pilot deployments in senior communities have shown high sensitivity and specificity for fall detection while providing continuous heart rate monitoring that helps identify early signs of infection or dehydration (which often manifest as elevated resting heart rate).

Fitness and Wellness

Although most fitness trackers use PPG (optical) sensors that require skin contact, radar will soon enter the consumer wellness space. Automotive manufacturers, for example, are embedding radar in driver seats to monitor the driver’s heart rate and detect drowsiness (microsleep events) before an accident occurs. In smart homes, a radar module in the living room ceiling could track the heart rate of a person watching TV and integrate with a health dashboard or emergency response system. Several companies have announced plans for radar-based smart scales and bathroom mirrors that can capture heart rate without the user touching any electrodes.

Disaster Response and Military Medicine

Radar’s ability to detect human vital signs through rubble or walls has applications in search and rescue. Portable radar systems can locate survivors under collapsed buildings by sensing their heartbeat and breathing, even when the person is unconscious. Military triage also benefits: a medic can assess the heart rates of casualties from a distance without exposing themselves to enemy fire or secondary hazards.

Challenges and Innovations

Despite its many strengths, radar-based heart rate monitoring is not without limitations. The primary challenges include susceptibility to motion artifacts, interference from other moving objects, limited range in certain conditions, and difficulties in separating multiple people in a room. However, ongoing innovations in hardware and software are rapidly overcoming these barriers.

Motion Artifacts

Body movements such as turning in bed, arm gestures, or even deep sighs produce chest displacements that are orders of magnitude larger than the heartbeat signal. During such movements, the heart rate estimate becomes unreliable or completely lost. Traditional signal processing uses adaptive filters to reject motion periods, but this reduces the duty cycle of available data. Recent machine learning approaches, particularly convolutional neural networks (CNNs) and recurrent neural networks (RNNs), can detect motion intervals in the raw radar data and either exclude them or reconstruct the heartbeat signal from the residual. Some systems now employ a multi-input architecture that uses both radar and an additional low-cost inertial sensor (e.g., a small accelerometer on the bed frame) to cancel body motion. As algorithms improve, the time during which a clean heart rate can be extracted increases, moving toward truly continuous monitoring.

Separation of Multiple Subjects

In clinical settings where two patients share a room or in a smart home with multiple residents, the radar system must attribute each heartbeat to the correct person. FMCW and UWB radars can spatially separate subjects by their range from the sensor, provided they are at least 30–50 cm apart in distance. However, if two people are close together (e.g., a couple in the same bed), separation becomes challenging. Advanced beamforming arrays and multiple-input multiple-output (MIMO) radar techniques create virtual arrays that can estimate the angle of arrival of reflected signals, allowing separation by angle as well as range. Some research prototypes have demonstrated reliable tracking of two people in a bed with heart rate errors under 3 BPM using a 4×4 MIMO radar at 77 GHz.

Signal Interference and Environmental Clutter

Radars operating in the ISM (industrial, scientific, and medical) bands share frequencies with Wi-Fi, Bluetooth, and other wireless devices. Interference can cause false counts or signal dropout. Shielding, frequency hopping, and spread-spectrum techniques mitigate these effects in commercial products. Additionally, stationary clutter (furniture, walls) can create large reflections that desensitize the receiver. Modern radar chips include digital cancellation loops that subtract the static background, leaving only dynamic motion (heartbeat and respiration).

Regulatory and Privacy Considerations

Medical-grade radar monitors must pass stringent safety standards for radio frequency exposure (e.g., IEEE C95.1, FCC Part 15) and obtain regulatory clearance from agencies like the U.S. FDA or European MDR. Most devices use power densities well below 1 mW/cm², which is orders of magnitude lower than levels known to cause thermal effects. Privacy is another concern: because radar can detect presence through walls, there is potential for surreptitious monitoring. Manufacturers are responding by designing devices that only operate when within a defined range (e.g., in the same room) and by processing data locally on the device rather than transmitting raw signals to the cloud. Some products also include an indicator light that activates when radar is transmitting.

Future Prospects and Integration

The trajectory of radar heart rate monitoring is toward miniaturization, lower power consumption, and seamless integration into everyday environments. Several trends will define the next five to ten years.

Integration with Smart Home Ecosystems

Major technology companies are exploring radar as a sensor modality for smart speakers, thermostats, and light switches. Google’s Soli radar chip (now in the Pixel 4 phone and some Nest devices) can detect gestures and presence. Future iterations could add vital sign sensing, enabling a smart home to automatically adjust lighting and temperature based on sleep stage, or to call for help if no heartbeat is detected. Amazon has filed patents for radar-based health monitoring in its Echo devices. As these features become standard, continuous heart rate data will be available to consumers without any additional equipment, driving a shift from episodic to continuous wellness tracking.

Machine Learning for Diagnostic Insight

Current radar heart rate monitors output a simple BPM value. More advanced algorithms are being developed to extract health indicators beyond rate: heart rate variability (HRV) is a marker of autonomic nervous system function and is predictive of cardiovascular risk. Radar can also capture the morphology of the chest displacement waveform, which correlates with cardiac ejection timing and contractility. With enough training data, deep learning models may be able to detect arrhythmias such as atrial fibrillation or premature ventricular contractions from radar signals alone. Researchers at the University of Waterloo demonstrated a 77 GHz radar system that detected atrial fibrillation with 90% sensitivity and specificity in a small cohort. Larger validation studies are underway.

Wearable and On-Body Radar

While the focus has been on stationary radar, flexible, low-profile radar patches worn on the body are also in development. These devices use near-field radar to monitor heart rate through clothing without direct skin contact. Because the radar antenna is only a few centimeters from the chest, motion artifacts are reduced, and the signal strength is higher. Early prototypes have been demonstrated in a wristband form factor, capturing heart rate and blood pressure via pulse wave velocity. Such wearables address the comfort limitations of optical and electrical sensors while providing comparable or better accuracy.

Combined Radar and AI for Contactless Blood Pressure Estimation

An exciting frontier is non-contact blood pressure monitoring using radar. Blood pressure estimation typically requires two measurements: pulse transit time (PTT) or pulse arrival time (PAT) and a calibration parameter. Radar can measure the time delay between the heart’s electrical depolarization (ECG R-wave) and the arrival of the pulse wave at a peripheral site (e.g., the wrist or temple). Some systems use two radar modules — one pointed at the chest and one at the wrist — to estimate PTT. Coupled with a one-time cuff calibration, continuous blood pressure estimates have been achieved in lab studies with mean absolute errors below 5 mmHg for systolic pressure. If validated in larger trials, this could revolutionize hypertension management by allowing patients to monitor their blood pressure unobtrusively throughout the day and night.

Conclusion

Non-contact heart rate monitoring using radar technology has matured from a research concept into a commercially viable and clinically valuable tool. Its ability to track cardiac activity without wires, patches, or even awareness opens new possibilities in hospital care, home health, sleep medicine, and safety monitoring. While challenges related to motion artifacts, multi-subject separation, and privacy remain active areas of research, the pace of innovation is rapid. As radar chips become smaller, cheaper, and more power-efficient, and as machine learning algorithms grow more sophisticated, radar monitoring is likely to become a ubiquitous sensing modality — embedded in our homes, vehicles, and public spaces. The ultimate promise is a future where heart health can be continuously, non-invasively, and passively observed, leading to earlier detection of abnormalities, fewer emergency visits, and a more personalized approach to cardiovascular wellness.

For further reading on the technical principles and clinical validation of radar vital sign monitoring, refer to the IEEE publication on Doppler radar for healthcare: IEEE Transactions on Biomedical Engineering. Information about regulatory clearances can be found on the FDA Medical Devices website. The Mayo Clinic also provides an overview of non-invasive monitoring technologies at Mayo Clinic – Vital Signs Monitoring. Recent advances in UWB radar for sleep apnea screening are described in a study from Nature Scientific Reports (search for “UWB radar sleep apnea”). Lastly, the Google Soli project is documented at ATAP Google Soli.