MetaWatch Experiments

As the MetaWatch bluetooth watches are getting closer to shipping, I figure it’s a good time to talk about some of the UI experiments I’ve been doing with them, to give you an idea of what they might be useful for.

MetaWatch is a line of hacker-friendly wristwatches that can be paired with smartphones to enable new kinds of lightweight interactions. In the same way that glancing at a wristwatch is faster and less disruptive than pulling out a pocket watch to check the time, you can imagine how glancing at a connected watch could be more convenient and sociable than pulling your phone out of a pocket or handbag to see cloud-based information.

After my previous experiments in showing live bus times on an older bluetooth watch, the guys at Fossil got in touch with me, and over the past couple of years I’ve served as an unpaid advisor to the MetaWatch project, in the hopes of helping to make the end products as developer-friendly as possible.1

There are a lot of things to like about the MetaWatch devices:

  1. The screen of the digital version is always on, so the information on it is always a discreet glance away. You don’t need to push or swipe anything to bring it to life.
  2. The battery life is reasonable, so you can wear these things for the better part of a week without having to charge them.
  3. Since they’re designed by experienced watchmakers at Fossil, the MetaWatch devices look more like fashion watches than cookie-sized computers strapped to your wrist.

There are definitely some trade-offs, though, compared to other devices with more horsepower and flashier displays:

  1. As an Android app author, you’re basically treating the MetaWatch as a dumb terminal. You send pixels to the screen, and you get button presses back. This gives you a lot of control, but the downside is that watch-based UIs are a lot less responsive than you’d like.2
  2. The low-resolution monochrome display isn’t as sexy as color touchscreen devices like the iPod Nano or the WIMM Platform. They’ve made the most of it by hiring Susan Kare, designer of the original monochrome Mac graphics, to do the default imagery. You can’t reuse existing designs—for best results any UI is going to have to be custom-designed for this thing.

With that out of the way, here are a few prototypes that I’ve made over the past year or so.3 I hope you’ll excuse the rough graphics in places; I mostly wanted to see how these interfaces would feel if they were always easily accessible on my wrist.

Imagine this: you’re at the airport, hands full of luggage, and you just want to know where and when you need to be at your gate. Wouldn’t it be handy if you could glance at your wrist to find out? Matt Webb called this use “personal signage”, which is a nice way of thinking about it—you can get by with a lot less screen real-estate if your devices know exactly which part of the departure board you’re interested in.

How long do you have to work before your next meeting? It’d be great to be able to see that at a glance.

When you’re driving home, the time that’s really important to you is when you’ll get there. Assuming your phone knew your commute home, it could check current traffic conditions and show you the time you’d get home if you were to leave now.

Since I take the London Underground home instead, what I want to know is how the trains on my line are running.

Of course, the watch has several buttons that can enable you to trigger phone actions. I always text my wife when I’m heading home from work, so I wrote a little app to send a canned message with a single button press as I’m walking out of the building.

If you’re an author who compulsively checks your Amazon rank and social network stats, you could put those things right on your wrist and avoid the distraction of surfing to those sites.

I was curious how my commute time broke down, so put together a custom stopwatch app. I hit the button when I stepped off the bus or got out of the tube, and the app saved that “lap”. (It also used the time of day to know whether to reverse the order of the steps in the list.)

If you want to keep yourself focused on important things, perhaps a little memento mori reminding you how many days you had left (actuarially speaking) would help?

If that’s too morbid, you could try the Pomodoro Technique of staying focused on work in 25-minute increments, and have the watch vibrate when it’s time to take a break.

Finally, a little bit about the development process: to make it easier to quickly build new MetaWatch apps, I put together a little framework called Cicada. It detects watch-compatible Android apps as they’re installed on the phone and automatically adds them to an on-watch menu system.

A Cicada app can run in full screen mode…

…and optionally, the same app can be run in widget mode alongside other apps. Here, the same realtime bus times app is only using the bottom ⅓ of the screen, with other Android applications providing the tube status widget and the digital clock widget.

I’ll talk more about the Cicada framework in a later post.


  1. Disclosure: the MetaWatch guys have provided me with several prototype watches to experiment with over the past couple years. In my day job, I’m employed by Google UK, but my bluetooth watch experimentation is a personal project done on my own time.
  2. You can also modify the firmware that runs on the watch itself, which would be much more responsive, but I haven’t experimented with this yet.
  3. The watches shown here are early prototypes that have slightly different appearances and branding than the shipping devices.