How Pod 4 Ultra Detects and Mitigates Snoring via Machine Learning: The Intelligence Behind This Groundbreaking Feature

Many people struggle to sleep due to snoring, whether it be their own or their partners. While it’s obvious that loud snoring can be extremely disruptive to a good night’s sleep, snoring has also been linked to anything from sleep apnea to the common cold.

We’ve launched a revolutionary new feature with Pod 4 to combat the negative impacts of snoring on sleep fitness. 

Track your daily snoring patterns with snoring detection 

We leveraged Autopilot’s intelligence to detect — with 93 percent accuracy — snoring as it happens. 

Before being able to effectively stop snoring, we’d first need to identify if and when it is happening. The sensors inside the Pod can detect snoring thanks to the vibrations picked up by the sensors that span the entire width of the Cover. These sensors are located at chest level and pick up vibrations from snoring through the chest wall. These vibrations are then fed into one of Autopilot’s algorithms to analyze your (or your partner’s) snores. Through this algorithm, the data is analyzed by a neural network trained using data from more than 2,500 known snorers, ultimately allowing us to provide accurate information to users about their sleep every night. Every day our deep-learning model analyzes millions of minutes of vibration data, continuously improving its overall accuracy and intelligence. 

While many products in the market use an app to detect snoring via a microphone, unique sensors within the Pod 4 allow us to detect the vibrations from your body during snoring, rather than sound. By not relying on sound, we can accurately determine which side of the bed the snoring is coming from, making it easier to provide reliable sleep data to users in the morning. 

When users wake up in the morning, they’ll see a comprehensive report of their snoring, with details like when they snored, for how long, and even how intensely – all without using a wearable device. 

With Pod 4 Ultra, eliminate your snoring entirely

Snoring can be mitigated in a multitude of ways, from adjusting your sleep position to your side, to limiting alcohol intake before bedtime. However, the success of these fixes largely depends on the individual making behavior changes. In more severe cases, you may even be recommended to sleep with mouthguards or a CPAP machine to promote airflow in your airways while sleeping, which can be uncomfortable options. 

To make snore mitigation more effortless for the user, we determined that an adjustable Base would be the most seamless way to approach the problem. With Pod 4 Ultra, we can both detect and also reduce or stop snoring entirely using the movement of the Base. 

Pod 4 Ultra’s adjustable Base gently elevates the bed when snoring is detected – in real time and without disruption. The Base was designed to be silent and gentle in movement to avoid waking either partner when making adjustments. We achieved this by using custom motors and control systems to move the Base with a much lower minimum speed than other products in the market. The gradual angle change moves at a nearly imperceptible rate, supporting our ultimate goal of keeping users asleep throughout the night, undisturbed.

Unlike other products on the market, Pod 4 Ultra also reacts to snoring in real-time, making instant adjustments. You can review how the Pod reduced your snoring each morning in your sleep and health report.

Clinically proven to reduce snoring by up to 45%

We tested snoring detection and mitigation in a clinical study that captured over 200 nights of data while people slept on Pod 4 Ultra. The study revealed that participants saw up to a 31% decrease in snoring when Pod 4 Ultra’s head angle was raised to 6 degrees, and up to a 45% decrease in snoring when raised to 15 degrees. 

To assess how the Pod 4 Ultra impacts snoring duration, we had 39 subjects (33 male, 6 female, 38 ± 12 y) tested for 281 nights on the Pod 4 Ultra. To understand typical snoring patterns, we collected snoring data for 3-5 nights where the Base was flat the entire night. Subjects then slept for 3 nights at Base angles of 6, 10, or 15 degrees, where the head of the Base would automatically rise to one of these angles when snoring was detected. 

All snore data used for these analyses was collected via SnoreClock, our ground truth microphone-based snore detection app. The subject’s phone, with the SnoreClock app, was placed on a bedside table one foot from the person’s head (when lying in bed). SnoreClock provides the total percentage of time the person snored during the night. We used this percentage of nightly snoring to calculate the percent change in snoring for each subject as: (snore % – baseline snore %) / (baseline snore %). Then, the average percent change in snoring across all subjects was calculated.

For 76% of subjects, higher angles further reduced snoring. However, as it may take a few nights to get used to using the adjustable Base, we recommend starting at 6 degrees (or the equivalent to the Low mitigation setting in the Eight Sleep app), and then trying higher angles as you see fit. In fact, 41% of our study subjects decided to continue using the medium and high angles (i.e. 10 and 15 degrees, respectively) after the study was over.

Shop the Pod

Upgrade your sleep with Eight Sleep's cooling technology

Learn more
Pod 4

Read more

Empty image

How Taylor Fritz, pro tennis player, uses the Pod for his competitive advantage

Empty image

Dr. Andrew Huberman, Dr. Matthew Walker, and Dr. Peter Attia join the Eight Sleep Scientific Advisory Board

Empty image

Eight Sleep’s New Deep Learning Sleep Algorithm

Empty image

How Taylor Fritz, pro tennis player, uses the Pod for his competitive advantage

Empty image

Dr. Andrew Huberman, Dr. Matthew Walker, and Dr. Peter Attia join the Eight Sleep Scientific Advisory Board

Empty image

Eight Sleep’s New Deep Learning Sleep Algorithm

Empty image

How Taylor Fritz, pro tennis player, uses the Pod for his competitive advantage

Empty image

Dr. Andrew Huberman, Dr. Matthew Walker, and Dr. Peter Attia join the Eight Sleep Scientific Advisory Board

Empty image

Eight Sleep’s New Deep Learning Sleep Algorithm