Wearables: 2014 and Beyondby Stephen Barrett on January 15, 2015 11:50 AM EST
- Posted in
- Android Wear
- Microsoft Band
Wearable Use Cases
Inevitably in any wearable discussion with friends or family, one of the first questions asked is “why?” The general public sees the value provided by smartphones clearly, but with wearables that is not always true. In an effort to describe wearable value in general, I will present the top two use cases – fitness and smart watch. Hopefully this provides some context of where wearables are now and where they can go in the future. Future device reviews at AnandTech will have use cases like these in mind when evaluating the quality of a wearable.
Today, fitness wearables have typically provided the most benefit to runners, walkers and cyclists, or just about anyone moving a distance through their own effort (kayaking, canoeing, rollerblading, etc). This is due to a good match of user needs and wearable technology’s specialized ability to meet those needs. A summarized list of care-abouts yields:
- Notification when reaching distance markers – to keep track of progress toward goals
- Notification of speed traveled at each distance marker – to make sure to achieve pace goal
- Overall speed – to make sure to hit pace goal
- Elapsed time – to help schedule a day or meet people at certain times
- Length traveled – to help meet personal fitness goals
- Heart rate – to measure body strain and assist in pacing
- Calorie counting – to aid in personal fitness plan goals
- Map of travel detailing pace – to review pacing and share via social media
- Make calls – to handle an emergency
- Listen to music and podcasts – for motivation and entertainment
- Elevation tracking – to review effort and share via social media
- Connectivity – to interface to other devices like a cyclist’s power meter or a treadmill’s display
Solving all of these with a wrist-worn wearable provides unique value, as the form factor is significantly better than the girth of increasingly large smartphones. Additionally, the display is more conveniently accessible than an arm-band mounted smartphone. However, running or cycling while looking at your wrist is still inconvenient so Bluetooth audio notifications and connectivity to gym bikes and treadmills is desired.
Mapping wearable features to this list yields an imperfect but good result. Note that distance traveled is actually a fairly difficult thing to compute indoors or without GPS assistance, and relies on sensor fusion of compass + gyro + accelerometer passed to a pedometer algorithm.
- Notification when reaching distance markers – Sensor fusion / GPS, Bluetooth audio, Vibration, Display
- Notification of speed traveled at each distance marker – Sensor fusion / GPS, Bluetooth audio, Vibration, Display
- Overall speed – Sensor fusion / GPS, Display
- Elapsed time – Display
- Length traveled – Sensor fusion / GPS, Display
- Heart rate – Pulse oximeter
- Calorie counting – Sensor fusion
- Map of travel detailing pace – Sensor fusion / GPS
- Make calls – Cellular, Microphone, Bluetooth audio / Speaker, Phone contacts sync
- Listen to music and podcasts – Bluetooth audio, Large data storage
- Elevation tracking – Barometer / GPS
- Connectivity – ANT+ / Bluetooth Low Energy
Nearly every need is met by the hardware technology available in wearables on the market today. However, there are a few missing hardware pieces. Cellular functionality has yet to become widely available (outside the Tizen based Samsung Gear S) due to power consumption, miniaturization, and cost constraints; ANT+ support meanwhile is mostly missing. There are a few ANT+ enabled wrist-worn wearables, but none from Apple, Microsoft, or Google’s partners.
As cyclists commonly have ANT+ chest-mounted heart rate monitors, ANT+ power output meters, and ANT+ cycling computers, the lack of ANT+ on a wrist worn wearable seems like a missed opportunity. For example, a cyclist could replace their cycling computer and chest-mounted heart rate monitor with an ANT+ enabled wrist-worn wearable but retain their investment in the ANT+ power meter. The same goes for the many gyms that have ANT+ enabled equipment.
In my experience with the movement use case and today’s wearables, the hardware is very close but the software has not yet come up to my expectations. This is an incredibly competitive target at the moment that has not yet seen a clear winner or consolidation.
While fitness has been a key marketing point of many wearables in 2014, the products involved have yet to pertain to a key demographic of fitness conscious people: weight lifters. This is what I would consider a forward looking wearable target.
I certainly would not consider myself a body builder or gym rat but I do enjoy lifting weights much more than any moving exercise – and I am not alone. There are plenty of people in the world of gyms that spend their time using weights and not treadmills. Therefore, I find myself somewhat annoyed when wrist worn wearables are marketed as fitness devices but have a fraction of the value (or no value) to a weight lifter versus a runner. Personal thoughts aside, compiling a list of a weight lifters care-abouts yields a quite different list that highlights why this demographic has yet to be successfully targeted:
- Heart rate – to measure body strain and assist in pacing
- Exercise tracking – automatic detection of weight usage and exercises performed to provide historical tracking of gains and loses
- Personal record tracking – keep personal records (PRs or ‘bests’) data for each exercise
- Body fat and muscle measurement – keep track of body fat burn and muscle build over time
- Suggested exercises – utilize historical exercise data and muscle atrophy over time and provide intelligent suggestions for today’s exercises. Customizable to constrain suggestions to available gym equipment
- Suggested weight – when starting a new exercise, suggest a starting weight based upon personal information
- Fatigue tracking and warning – track muscle fatigue by muscle at the gym and over time. Utilize data to provide warnings when to stop lifting and when to revisit the gym
- Social features – compare and track with friends
The number one issue here is the lack of technology for automatic weight and exercise tracking. While there are weight lifting smartphone apps with manual data entry, these do not compare to the simplicity of automatic tracking runners and cyclists enjoy. Part of the reason products such as Fitbit became popular is their convenience. There is little more needed from the user than to wear the device and review the acquired data.
This is a solvable problem. Gyms of the future could contain NFC or Bluetooth enabled weights and machines. A wrist-worn wearable could track usage and movement of your body compared to the weights and conclude which exercises you performed and what weight used. Once that data is available, analysis based upon body type becomes possible and suggestions can be made. Combined with today’s heart rate and body fat sensors and weight lifters could find their perfect wearable and their favorite gym. There are efforts in the weight lifting wearable area now (see Push), but without automatic tracking they are currently second fiddle to the moving use case.
The smart watch use case is what I would consider immature. After some failed efforts from 2003 to 2009 from Samsung, Palm, and Microsoft, Pebble awoke the market in 2013 with a Kickstarter campaign. In 2014 the major players of Microsoft, Google, and Apple each targeted this market but none have perfected it. One of the main problems of this use case is parameterizing it. What unique value does a watch offer over a smartphone? Thus, many times smart watch functionality is combined with fitness functionality that can only be offered by a wearable.
Compiling a list of smart watch care-abouts yields:
- Time – need to replace a basic time telling watch
- Customizable watch face – need to replace the tens (hundreds?) of thousands of basic time telling watches
- Physically attractive – if I am going to wear it every day, it cannot look like a toy
- Comfortable – if I am going to wear it every day and sleep with it on, it cannot hurt or bother me
- Water resistant – to survive washing dishes, hands, weather. Ideally IPx7 or greater
- Rich smartphone notifications – keep track of what is happening even if the smartphone is not directly available, such as across the room or in a purse. Optionally dismiss or respond. All notifications should arrive to prevent missing some by relying on the smart watch
- Voice assistant – quick answers like what is the weather or when is the Cowboys' game
- Alarm clock – vibrate function to avoid waking up a partner
- Calendar – easily display my next meeting details such as where it is located
- Messaging – easily send quick messages and replies with SMS or other apps such as Facebook messenger
- Tasks and Reminders – create Exchange / Google tasks by voice and reminders
Nearly all of the actual features of a smart watch come directly from smartphone use cases. The difference is they are slightly tweaked toward the wrist-worn use case. When using a smart watch, the main benefit is getting things done even quicker than with a smartphone. It only takes a moment to rotate your wrist and say “OK Google, Wake me up at 7am” versus finding wherever your phone is, activate it (if no passive listening exists), say the same thing, and put it down somewhere safe. It is amazing to think that shaving these seconds off each interaction can have value, but when you add up each time you touch your smartphone every day it does quickly add up.
However, as many point out, these devices lack the killer app. There isn’t much they can do that your smartphone cannot. The vibrate alarm is one example, but there has to be more. Apple examined some ideas during their Apple Watch keynote such as pairing multiple watches. Taps on a watch sends a corresponding taps to others – useful for spy movies and tense corporate meetings. Until a smart watch specific killer app releases, AnandTech will evaluate the execution quality of the essentials listed above.
Fitness and smart watches were the clearest targets for wearables in 2014, however there are a variety of other wearable technology targets such as personal trainers, hair pieces, eye pieces (Google Glass), and clothing that will be interesting as they mature in the future.
Post Your CommentPlease log in or sign up to comment.
View All Comments
Impulses - Thursday, January 15, 2015 - linkUhh, there are already smart watches that use OLED displays, and smart watches that are perfectly visible in daylight (OLED or otherwise), and smartwatches that respond to motion (pretty much every Wear device? the 360's more sensitive mode burns a little more battery but it's very responsive).
Seems to me you haven't looked very closely at much of the options in the market... The biggest issue is really battery life, but adjusting to 1-2 days of battery life hasn't been a big deal to me. We already did it once when we went from feature phones to smartphones after all...
name99 - Thursday, January 15, 2015 - linkIf that is what you want, buy a Pebble today. I have one and I like it.
But I don't see the harm in others pushing the envelope to try to see what this form factor might be capable of.
To me the Apple solution (offloading all the serious work to the phone) makes more sense TODAY than the Tizen solution. (Google seems at a sort of intermediate point between the two, but I think is pushing the local CPU too hard). And I think a display tech like Mirasol is the way to handle color at low power.
But I think it's foolish to be too dogmatic about these issues. In particular, we don't know the expected ten year trajectory of all the pieces involved, from the energy supply side to the CPU power usage to the expected use cases. I suspect that
- CPU energy usage is actually a much smaller issue than screen and wireless energy usage. Meaning that there's no real win in skimping on the CPU (assuming it is, of course, a power optimized fast sleep/fast wake CPU) BUT the OS and OS/app interaction model are critical in ensuring that almost all the time nothing is running.
- memory may be a substantial power drain. I would not be surprised if the primary reason for Apple's off-load model is limited DRAM rather than a wimpy CPU/desire to avoid using the CPU much
[I also suspect, but maybe this is foolish, that Apple's battery life is going to be substantially longer than what they're suggesting in the press, that they're trying to calibrate expectations so that when they announce the actual battery life is 3 days rather than 1 day, people are awed and impressed. The reason I say this is that, compared to what's in a Pebble and the OS/app model, I can't see any serious sink of energy beyond a Pebble. Unlike Google/Tizen where there is all the "traditional" OS overhead and, I'm guessing, a lot more DRAM constantly draining away.]
- color may be "frivolous" but I suspect it's essential to "cross the chasm". I think accepting the limits of Mirasol (colors, but a limited palette) would be a more fruitful direction for Pebble if they want to remain viable than sticking with eInk.
Thinking "this is a watch plus; we'll architect the system that way" is a sure way to land up on the same path as Palm, Win CE, Nokia and other such "this is a phone plus" companies. These devices will NOT stay as just watches, even if that's the way they are perceived for the first two years or so.
mkozakewich - Friday, January 16, 2015 - linkThe OLED displays already use flicking or shaking to turn on. That's still bad. I'd say we use the two-colour e-ink displays so you can get off (white), black, and red or something. There's no reason you need to see colour on it. The problem here, though, is that it's not really visible at night. Putting some kind of night-light feature on would be helpful. Part of me wonders if one could add a transparent OLED on top just for when it's dark.
Watches and phones already use deep-sleep features. They can last a week if the radios were turned off and no apps were causing the phone to wake every few minutes.
Designing an OS is actually very difficult, which is why you never ever see anything good in consumer electronics. Think of the OSes driving things like consoles or those screens in cars or printers.
Fancy animations are actually important. A watch with no extra graphical features just won't have that visual pull and will feel ugly. People won't want to wear them.
These devices will definitely become cheaper as CPUs are tailor-made for them and the processes shrink even more. In the meantime, though, these things will cost far more than they're really worth. I think the best bet might be to sell them along side a cheapish phone so you can offer a big subsidy for the combination.
mrdude - Thursday, January 15, 2015 - linkWith all of these smart watches being released, Dick Tracy is going to have a fit trying to find the right one. Unfortunately for him, fighting crime and solving cases is going to require a Bluetooth pairing to a compatible smartphone.
mjcutri - Thursday, January 15, 2015 - linkOne device (and manufacturer) that you left out of the discussion is the recently announced Garmin Vivoactive:
Garmin is huge among fitness buffs (cyclists - road and mountain, triathletes, marathoners) for tracking their training and races. They also have a few other devices that pick up the smartwatch theme: the more traditional but more expensive fitness tracker, the 920xt, and the just announced the Fenix 3, which combines the features of the 920xt with a more traditional round watch appearance and the Epix, which should appeal to outdoorsy types with the topographic maps display.
I am hardly a hardcore athlete, but I have been using the old garmin 310xt for a couple years now to track my cycling and running activities and had been using a fitbit one (until I lost it) to track my steps. I am interested in the vivoactive because it looks like it combines the fitness tracker, GPS, and smartwatch into a single device that I could see myself wearing everyday.
Stephen Barrett - Thursday, January 15, 2015 - linkI agree. I couldn't cover everything and that device was actually launched in 2015 so it wasn't a good fit for the article. However, I'll reach out to Garmin and see if I can get a sample.
Thanks for the feedback
mjcutri - Thursday, January 15, 2015 - linkThanks Stephen. I enjoyed reading the article and look forward to AT's future coverage of wearables.
Another advantage of the garmin devices that really appeals to me is that they are stand-alone devices that optionally connect to your smartphone. For me, this means that I can go for a run with just the watch (I hate hauling around my phone when running, which is why I picked up the 310xt in the first place) or I can go fro a bike ride with my phone in my hydration pack and still have access to texts and control my music.
The only downside to them is the Garmin Connect interface. I have been having a hard time lately getting my 310xt to sync, and because of that I had been thinking about moving away from garmin towards some other smartwatch type device (possibly fitbit surge,) but the new devices with the wifi and/or BT sync seem to eliminate the sync issues I have been having with my 310xt.
DBasic - Monday, January 19, 2015 - linkGarmin's vivosmart has many of the desired features except GPS. However, unlike many similar competitors, the vivosmart is 5 ATM rated - which is why I got it so water activities would not be a concern. Pairing/sync on unsupported phones is a pain but does work.
nevertell - Thursday, January 15, 2015 - link"Gyms of the future could contain NFC or Bluetooth enabled weights"
Please, no. Whilst I can imagine NFC working well, bluetooth implies someone would have to charge the weights. And please, just think about that for a second. Also, imagine a gym, usually a gym has multiple kinds of weights to be lifted, and they are stored usually in close proximity of one another. Imagine the bluetooth noise. If you're stacking weights, will you also register each weight individually with your smart device ? Or will the weights have a mesh network and then each smart device will act as a hub that will read the multicast data from each weight and then pick out the weights that report similar accelerometer output data to the smart device's accelerometer data ? Imagine the potential for all the proprietary standards and protocols and the battery drain.
Wearables as sensors work good if little to no user interaction is necessary to obtain meaningful data.
Anyway, IMO the best smartwatch today is the pebble steel, that being said, I've only played with some pebbles and some Samsung Smartwatches. And from my experience with them, I believe that the only functions a smartwatch can do reasonably well are sensor data capture, data transmission to a hub, simple input transfer to a hub and short string display. A smartwatch needn't have standalone apps, because today there are no user interfaces that would work well enough on a 1" display with a maximum of 6 buttons and some gestures and point-and-click touch screen. Of course, if battery and processor tech advances fast enough, maybe there'd be room for a smartwatch that can track hand movements over it (like LEAP motion), then there could be a case for standalone apps, but until then, let's not try and shove a half assed Android on a dual core SoC on my wrist and call it a watch.
JarredWalton - Thursday, January 15, 2015 - linkWireless charging is certainly possible, remember. And ultimately it would require a lot of smarts to detect proximity, verify the weights are in use, etc. It's a potential solution, but I don't know how many people are really tracking this stuff and if it would help. Of course, there's the old saying: "What gets tracked improves." It's why Target, Walmart, etc. monitor the performance of employees, because if they don't most will trend towards doing less rather than more.