A growing number of devices are trying to gauge when we’re freaking out. But as the latest entry into the market shows, that’s a harder task than we think.
Part of the premise behind a new wearable called WellBe is pretty depressing: Many of us are so continuously stressed out—or so disconnected from our feelings—that we can’t actually tell which parts of our day are making us most anxious.
So the WellBe, now on Indiegogo, was designed to make those feelings a little more obvious. In theory, if you slap on the new wristband and sync it up with your calendar, it will tell you who and what is stressing you out most each day. Then it gives a series of simple meditations and exercises to help you better deal with those situations.
“We believe that when you know the triggers and have the solution, this is how you really reduce stress,” says Doron Libshtein, chairman and co-founder of WellBe.
One problem, however, is that it’s not yet clear that the wearable can accurately measure stress. The WellBe is designed to track heart rate variability, which can correlate with how upset you are, and it uses a custom algorithm to analyze changes in heart rhythm. But heart rate variability is notoriously difficult to measure—especially through a gadget like a simple wristband.
“If you’re off by milliseconds then that’s problematic,” says Erica Simon, a researcher in respiratory psychophysiology at the National Center for PTSD. “It can really be thrown off by things like movement.”
The WellBe only works when someone is sitting down, as an attempt to improve accuracy. (That in itself is a drawback—as someone who walks and bikes, I’m pretty sure some of the most stressful parts of my day are when I’m trying to avoid being run over by cars). But even small movements can ruin data.
“Movement isn’t just somebody walking,” says Simon. “Movement means I moved my wrist because I’m typing, or I went to take a drink from my coffee, or I’m talking on my phone…even gesturing. Any of those things can completely reduce the accuracy.”
An algorithm that works for one person might not work for another, she adds. And even if the device can accurately track heart rate variability, that doesn’t automatically mean that it knows how you feel—someone who’s excited and happy, for example, might have similar patterns to someone who’s freaking out about a missed deadline.
“It’s really tough to use heart rate variability as a measure of stress, because you can’t really disentangle the different emotions,” says Simon. It’s also not the case that people are typically either “stressed” or “relaxed”—instead, our bodies are usually in complex state of both at the same time, making stress even harder to measure.
Other wearable startups try to measure stress in different ways, like Spire, a gadget that tracks breathing patterns instead of your heart or the Neumitra, which measures electrical properties of the skin as a proxy for brain health. These methods, too, can be prone to error. For example, the temperature and humidity of the room and the medication someone is taking could skew the Neumitra’s results.
Though WellBe plans to do an independent study, it’s not clear yet that their algorithm can solve the device’s challenges. It also only gathers data for three minutes each hour, so it isn’t clear how it can necessarily catch each stress trigger. It’s possible that the app might work better if it just asked people to rate their own stress—and then offered the same relaxation techniques. Still, the rest of the app seems like it could be useful: The company offers a library of over 1,000 different meditations and other relaxation techniques, and the app tries to quickly learn which methods work best for each person.
“Our passion is to bring it to more and more people,” says Libshtein. “Especially people who never meditate, who never took the time to reduce their stress, and to help them start this kind of practice. What we’re telling them is that we can help them choose the right method for them. It’s not one mantra, or just sitting and trying not to think. There are easy ways to reduce stress.”
The lesson is that there’s no one magical way to measure stress, though it may be more possible with more streams of data. Anyway, for some people, it may not even be desirable. Stress is often a side effect of actual work getting done.