When we started building the Velmio iPhone app we were surprised to discover that Apple Watch devices don’t track sleep. Sleep patterns change significantly during pregnancy, owing partly to hormonal fluctuations and also because of physical symptoms disrupting sleep. 

Therefore, we knew that it’s super important to provide pregnant women with mechanisms for tracking their sleep throughout pregnancy. Our team really likes the way that other smartwatch devices detect sleep down to individual sleep stages (awake, REM, light, deep), so we wondered if we could create a similar experience for Apple Watch users.

Luckily we have a mathematician on our team so we were able to work some magic with mathematical algorithms and machine learning to make this happen. The result is a really cool feature that automatically detects your sleep and estimates the time spent in each sleep stage. All you have to do is wear your Apple Watch to bed and the Velmio app handles the rest.

How to use Velmio Apple Watch sleep tracking

Simply wear your watch to bed (make sure it’s charged!). In the morning, open the Velmio app and sync your data from Apple Health App. The app will run its algorithm to automatically detect a sleep event and display a popup for you to confirm your sleep times

You can make adjustments to your sleep times if needed

Our algorithm will also estimate the amount of time spent in each sleep stage (awake, REM, light, deep). Go to “Health Overview” (via the “Pregnancy” tab) to view the results

So how does Velmio sleep tracking work?

Most wearable devices that track sleep use data from the device’s accelerometer hardware (motion sensor), to determine sleep based on your movements. However, due to technical limitations with Apple’s software and hardware, it’s not yet feasible to use this method to track sleep on Apple Watch.

However, we realized that it’s possible to emulate motion data using other variables captured by your device, like steps and heart rate. Using machine learning, we developed models to recognize patterns in your data that are consistent with sleep and matched this with polysomnography (sleep study) data recorded in clinical settings, to allow our algorithm to estimate your sleep stages.

Therefore, your Velmio sleep analysis results should be interpreted as simply an estimate or approximation, based on patterns detected in your wearable data. If you are concerned about your sleep, a formal sleep study undertaken at a hospital or sleep clinic remains the only precise method of measuring sleep.

Experts speculate that Apple will roll out sleep tracking functionalities in future versions of the Apple Watch. Until then, apps like ours remain the only way to automatically track sleep using Apple Watch.