The research behind the Pod

Contributed by Natasha Ragland

At Eight Sleep, every detail on the Pod is guided by one principle: improving your sleep with accurate biometric insights without compromising comfort. That’s why any significant change to the Pod goes through rigorous testing in our labs, where our R&D Clinical team ensures that each iteration of the Pod meets or exceeds our biometric accuracy criteria and delivers on its promise to improve sleep.

Every new product goes through at least four rounds of clinical testing to maintain our clinical-grade accuracy (heart rate, for example, is 99% accurate).

Transitioning from Pod 3 to Pod 4 with clinical testing

Member feedback from Pod 3 inspired improvements in Pod 4’s comfort and temperature control. We added thicker foam for softness and redesigned the thermal grid for better temperature performance (Image 1). As you can see in the image, Pod 4 uses a tubing technology in place of Pod 3’s grid. The materials on each are completely different, as well as the layout, design, and ultimately the performance they provide.

Given that the Pod sensors (which produce biometric data like heart rate) are in close contact with this thermal technology and the foam layers above it, we needed to test whether these changes altered the signal quality and biometric accuracy.

Step 1. We tested early prototypes to evaluate accuracy of heart rate and heart rate variability

In the very early stages of the process, the Clinical team receives only a few Pod prototypes, which are often built by hand. In the case of Pod 4, after we received these early prototypes, we began with one week of in-lab testing involving 30 participants. Each person laid on Pod 3 and Pod 4 for 1.5 hours each, while wearing a Polar H10 heart rate monitor for gold-standard heart rate (HR) and heart rate variability (HRV) benchmarks. This heart rate monitor collects gold-standard, medical-grade HR and HRV data, which we use to compare to our Pod algorithms. In addition to the in-lab protocol, we asked participants to rate the comfort and thermal performance of the new Pod Cover.

We then analyzed the data against strict HR and HRV pass/fail criteria. If the new Pod design failed, we ran follow-up tests to find the issue. Thankfully, Pod 4 passed on the first round of testing.

Step 2. We test prototypes in participants’ homes to capture real-world behaviors

Next, we tested Pod 4 in real homes to see how it performed in everyday conditions that can’t be replicated in a lab. In the first round of this testing stage, we placed 14 prototypes in homes to evaluate the Pod’s accuracy with 27 different sleepers.

Each participant wore the same heart rate monitor used during the in-lab testing, logged their bed entry and exit times to validate our presence detection algorithm, and completed surveys to monitor comfort and thermal experience. Each round of testing typically lasted up to seven nights over a two-week period, followed by statistical analysis to confirm Pod 4 met our strict accuracy criteria. We also wanted to make sure that we collected subjective feedback in the key areas that we set out to improve with Pod 4: comfort and thermal performance. Lastly, participants kept the prototypes in the field for 6 months to monitor the Pod’s reliability.

We went through several iterations of this in-home testing to ensure Pod 4 delivered both high precision in data accuracy and a consistently comfortable, effortless sleep experience. The main changes that occurred from this testing round were:

Foam: We discovered that some foams created a “puffy” visual as a result of friction between the fabrics. This led us to choose a foam that minimized this effect while still maintaining the highest comfort rating.
Thermal Consistency: We improved thermal performance without compromising comfort. This was achieved by finding a tube thickness that was both comfortable, flexible, and sturdy to survive the real world.
Member Experience: When you open up a new Pod, you’ll see a sticker that says “This side up”. This sticker was added after some overnight participants set up their Pod Cover upside down!

Step 3. We updated the algorithms for Pod 4, using large real-world datasets

As a result of the changes to the foam thickness, the sensor signals were altered, meaning that we needed to update the algorithms. We collected more data to train the algorithm to interpret these new signals, testing 120 people across 14 nights each, to ensure the app’s biometric data remained accurate.

We also spent multiple testing rounds to bring up Eight Sleep’s new adjustable base. To complete this testing, we had to make sure that when people raised and lowered the base, the biometric accuracy remained unchanged.

Another key upgrade in Pod 4 was the Tap-to-Control feature, which lets you adjust the Pod or Base without your phone. This Tap-to-Control feature uses an accelerometer sensor placed on the side of the Pod. This type of sensor picks up on movement. We tested this sensor to distinguish intentional taps from accidental movement. Through multiple lab sessions, we determined the most comfortable location and proper tap force required to change the temperature or move the Base.

Step 4. We worked with 322 participants across 1000+ nights of data to bring Pod 4 to market

In total, Pod 4 went through 8 rounds of clinical testing, which encompassed 1000+ nights of data and over 320 participants. However, even after all of the testing done by the Clinical team, the Product and Software teams did their final testing round before launching the product. During their testing, they ensured that the app was working as expected, the onboarding flow was optimal, and the packaging was safe for transit. Once all teams gave the final sign-off, Pod 4 was ready to launch.

We don’t stop there…

After we launch a product, the data collection continues. We follow up with members to see how the product impacts their sleep, and our results are incredible:

The result of all of this clinical testing? Better sleep through accurate data

From lab-controlled testing to at-home testing, our intensive clinical testing process ensures that every Pod we release is more than just comfortable– it delivers the biometric accuracy you can trust.

This commitment to rigorous validation is how we continue to innovate and deliver on our mission: helping you achieve better sleep and recovery every single night.

You can participate in Eight Sleep studies

If you would like to become a beta tester, you can sign up for our email list for future study recruitment by filling out this survey.

Eight Sleep Labs Enrollment Process

Each round of testing requires extensive recruitment of testers to adequately power our statistical analysis.

Recruitment:

Potential participants fill out recruitment surveys, survey responses are then filtered and assessed for inclusion and exclusion criteria based on study goals, and stratified for targeted variability in the data (i.e., have a partner, have a king-size bed, etc).

Informed consent:

After a participant passes initial recruitment screening criteria, each participant is taken through the informed consent process. The informed consent document is a walkthrough of the study details, any risks associated with testing, and to let people know that their participation is completely voluntary.

Onboarding:

After participants have consented to and enrollment is finalized, they will go through an onboarding live call. Onboarding walks participants through the in-depth instructions and protocol of the study. This process ensures participants know what to do and when. They are then ready to begin testing!