UMass app aims to make Apple Watch sleep data clinical-grade
A new app called BIDSleep could push Apple Watch sleep tracking closer to research-grade standards. Built by a team at UMass Amherst, the tool uses AI to interpret heart rate and motion data from the watch, achieving sleep stage accuracy that outperforms other consumer methods.
Apple Watch sleep tracking has a data problem
The Apple Watch already tracks sleep, but its staging data is limited. It uses heart rate and wrist motion to divide sleep into light, deep, and REM stages. The data gets synced to the iPhone’s Health app, where users can view overnight breakdowns. While useful for spotting sleep trends or encouraging better habits, the accuracy isn’t strong enough to be used in clinical settings or in formal research.
Studies show the Apple Watch does well at detecting whether you’re asleep or awake. But when it comes to identifying specific stages like deep or REM sleep, it becomes less reliable. For example, in one peer-reviewed analysis, the Apple Watch had high accuracy for light sleep but only around 50 to 60 percent accuracy for deep sleep. That might explain why some users report waking up after eight hours of sleep only to see their watch claim they had just a few minutes of deep sleep.
Part of the issue is that wearables like the Apple Watch are making guesses based on proxies, not direct brainwave activity like EEG. So while it can spot when you’re still and your heart rate slows, it has a harder time distinguishing which phase of sleep you’re actually in.
What BIDSleep adds to the mix
This is where BIDSleep becomes interesting. Developed by Professor Joyita Dutta’s team at UMass Amherst, the app runs on the Apple Watch and collects high-resolution heart rate data during sleep. That data is processed by an AI model trained to classify sleep stages. The model reached 71 percent accuracy in benchmark testing, which is higher than many wrist-based methods and closer to the gold standard used in sleep labs.
Essential reading: Top fitness trackers and health gadgets
The standout claim is that BIDSleep does a better job identifying deep sleep. That matters because deep sleep is when the brain consolidates memory, clears waste, and resets itself for the next day. It’s also one of the hardest stages for consumer wearables to get right. So improving this one metric could have outsized benefits for understanding long-term health.
Why this matters for research and users
Sleep labs typically involve a single night of testing with full EEG setups, which doesn’t always reflect a person’s regular sleep. By using a standard Apple Watch paired with a smart algorithm, researchers can collect data over many nights, in real-world conditions, without asking people to sleep with wires on their heads. That shift in how data is gathered could help uncover long-term patterns linked to neurodegenerative disease, mental health, or cardiovascular risk.
It also levels the playing field. Most universities and clinics don’t have access to large-scale sleep labs. But they do have access to students or patients already wearing Apple Watches. With BIDSleep, the watch becomes more than a lifestyle tool. It becomes a gateway for science.
This approach also raises interesting questions. If consumer hardware paired with better software can meet research standards, do we still need dedicated sleep devices for all studies? Could regulatory bodies one day accept this type of data for medical screening or diagnosis? And what role will Apple play, given that most of this is happening outside its own sleep tracking framework?
The Apple Watch still isn’t a lab-grade sleep tool on its own. But paired with research-driven apps like BIDSleep, it edges much closer. It shows that the limitations of wearable sleep tracking may have more to do with software than hardware. And for researchers, it opens up a low-cost way to study sleep at scale, over time, in real conditions.
Sources: Umass.edu, Facebook via Digitaltrends
Subscribe to our monthly newsletter! Check out our YouTube channel.
And of course, you can follow Gadgets & Wearables on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds.
