For the last few months, I’ve been wearing two fitness trackers on my wrist (in fact, for one of those months I was wearing three). And this week, with the close of Google’s Fitbit acquisition, I’ve been thinking a lot about wrist-worn wearables and the types of data they can provide.
I know there’s a lot of excitement around glasses, whether they put a digital assistant on your face or simply offer some kind of heads-up display, but in the near term and for many use cases, I think wrist-worn wearables are the device format to beat.
The Fitbit acquisition is fundamentally about bringing even more data into Google’s ecosystem — not to sell ads, but to deliver more context for its digital assistant plans. Indeed, if the Google search page was the juggernaut that propelled Google’s ad business, Google Assistant and its ability to deliver insights will be the juggernaut driving Google’s next big business opportunity. (If digital assistants do become the next big business opportunity I think it’s probably essential that we pay for them, because otherwise how could we trust the recommendations they make? But that’s a different essay.)
A wrist-worn wearable can provide health data and presence, and can even act as an adjunct to a nearby phone. For years, I’ve wished that my Fitbit would act as a trigger for turning on devices in my smart home and registering my preferences for lighting, music, and temperature. Now, thanks to Bluetooth tracking networks built by Tile along with soon-to-be-built LPWAN networks, a wrist-worn wearable could also let you track the general whereabouts of your friends, your kid, or even your spouse.
In corporate environments, bosses could track workers as they move throughout offices or warehouses for a variety of purposes, including social distancing or contact tracing. And such wearables don’t have to be passive. In a blog post by Microsoft showcasing the use of IoT in retail settings, a company called Turnpike built a communications platform using bracelets that share sales goals and inventory data with employees on the shop floor. In the example, it looked like employees could get notifications through their bracelets whenever they’re needed in a fitting room or out on the floor. Obviously the bracelets would have to be tied back to several physical and virtual services to deliver all of the information, but it’s a really cool concept.
Wrist-worn wearables aren’t the only option for this type of technology, but I think they have an advantage over many of the others. Rings and other jewelry are smaller, which makes delivering information over the device a challenge. Their batteries are smaller as well. That said, I do think wrist-worn wearables may have a slight disadvantage in heavy industries where the employee is stacking, tossing, lifting, and generally bashing about. Nodle is making a clip-on wearable for contact tracing and social distancing in industrial and warehouse environments that could meet those requirements, but again, there’s no great place to put a screen, and clipped-on items can easily fall off. And if you place them in less obvious locations you run the risk of forgetting to wear them.
So I think that for most jobs and in many other aspects of our lives, the return of a watch, bracelet, or some other connected wrist-worn strap will become the norm. The benefits are readily apparent, especially for people who may not have a phone on their person at all times. The question then becomes: How do you build devices and platforms that work for those buying the devices and those wearing them?
Right now, wearables are closely tied to our smartphones. In the case of the Apple Watch, for example, it’s so tied to the iPhone that if you don’t have one, it doesn’t really work. Such a constrictive model isn’t great because people don’t have infinite wrist space for dedicated bands tied to proprietary ecosystems. I’d hate to wear my Fitbit, a work strap, and a state-provided COVID contact-tracing band.
Instead, I’d like to see a series of reference designs comprising a set of radios, sensors, a processor, a screen, and a battery that could handle most wearables use cases. Add to this a standard OS on which companies could build apps and the ability to access all of the radios and sensors as needed. Device makers could add more to the physical hardware for additional use cases. For example, someone with a medical condition might have to spring for a fancier watch, while a kid might get the basic model. There’s plenty of room for fancier customizations on top of the basic computing platform, which gives different vendors options for innovation.
Software companies could then build to these standard designs so that whenever you get a new job or device you can just load a new app onto your wearable. In the meantime, the computing inside wearables will only get smarter. For example, Nordic Semiconductor, which makes Bluetooth modules, just announced a partnership with Edge Impulse to bring local machine learning to its chips. And at CES last week, Bosch announced a sensor that uses local ML to learn what exercises people are doing, so a user could “teach” their wearable how they exercise.
At least for the near term, the wrist is where it’s at!