Apple Watch's Workout app learns users' strides over time, Turlington Burns diary reveals

Posted:
in Apple Watch edited April 2015
In the third week of her promotional Apple Watch diary, posted Thursday, model Christy Turlington Burns made mention of a previously unknown feature of the wearable's Workout app that adjusts to a user's gait for more accurate results.




Burns wrote that after running with the Watch and an iPhone "a few times," the app will adapt to a person's stride. This should ultimately produce more accurate results, something Burns suggested is especially important given the difference between running outside and on a treadmill.

During the past week Burns was on vacation with her family in the Caribbean, but nevertheless took some time out to prepare for the London Marathon and shoot promotional photos. Images in the diary showed her switching to a pink strap, and using an animated jellyfish as her watch face.

The diary otherwise described her relaxed training regimen, split between indoor and outdoor runs, tennis, yoga, and walking. She also mentioned filling the Workout app's Exercise ring "almost every day," and having to turn off notifications one afternoon in order to get a nap. One of Apple Watch's health features uses the Taptic Engine to gently tap users on the wrist to remind them that they need to stand once an hour.

In the coming week Burns will be in Los Angeles, doing a 20-mile training run in preparation for London.

Apple is using Burns and her marathon training to help build anticipation for the release of the Watch on April 24. She was a prominent figure during the company's March 9 press event, appearing on stage alongside Apple executives.
«1345

Comments

  • Reply 1 of 88
    Hate to say it, but the Apple Watch will probably be less accurate than the iPhone for counting steps, due to its wrist placement.

    This is because we tend to move our wrists a lot more than our hips in day to day use, which means that there will be many erroneous steps counted with the watch. Having your iPhone in your pocket or handbag is much more reliable due to the greater stability in those places.

    In fact, the reason for the Apple Watch's need to train itself is likely due to this intrinsic weakness.

    In Burns's case, the Watch may not be so inaccurate if she uses it only when running a marathon, as she will adopt a regular motion with her arms. But most people will be using the Apple Watch for counting steps throughout the day, not for a specific workout. In that more common scenario, the iPhone will win for accuracy.

    I examined the wrist-worn Fitbit devices and the feedback they have received to arrive at this conclusion. Fitbit do make some non-wrist devices.
  • Reply 2 of 88
    kent909kent909 Posts: 731member
    This is a good thing that the device will be learning over time. The health app on my phone sometimes thinks I have been cycling when I have not.
  • Reply 3 of 88
    magman1979magman1979 Posts: 1,293member
    Hate to say it, but the Apple Watch will probably be less accurate than the iPhone for counting steps, due to its wrist placement.

    This is because we tend to move our wrists a lot more than our hips in day to day use (just look at an Italian gesticulating in everyday speech with his arms), which means that there will be many erroneous steps counted with the watch. Having your iPhone in your pocket or handbag is much more reliable due to the greater stability in those places.

    In fact, the reason for the Apple Watch's need to train itself is likely due to this intrinsic weakness.

    In Burns's case, the Watch may not be so inaccurate if she uses it only when running a marathon, as she will adopt a regular motion with her arms. But most people will be using the Apple Watch for counting steps throughout the day, not for a specific workout. In that more common scenario, the iPhone will win for accuracy.

    I examined the wrist-worn Fitbit devices and the feedback they have received to arrive at this conclusion. Fitbit do make some non-wrist devices.
    You just love proving me and Slurpy right don't you? Between you and Mr. Lawyer, all you have to say about the ?WATCH is baseless, vitriolic shite...
  • Reply 4 of 88
    cpsrocpsro Posts: 3,198member

    Most of the learning probably occurs when the wearer is also carrying a paired iPhone, because only the iPhone has GPS.

  • Reply 5 of 88
    solipsismysolipsismy Posts: 5,099member
    kent909 wrote: »
    This is a good thing that the device will be learning over time. The health app on my phone sometimes thinks I have been cycling when I have not.

    Since the 3-axis accelerometer in ?Watch only needs to assume it's placed on your wrist, whereas the iPhone's 3-axis acceloromater needs a more complex understanding of placement, ?Watch should be a lot more accurate in general. Regardless, if you do a lot of biking I would recommend something like the Wahoo Fitness RFLKR+.
  • Reply 6 of 88
    rogifanrogifan Posts: 10,669member
    cpsro wrote: »
    Most of the learning probably occurs when the wearer is also carrying a paired iPhone, because only the iPhone has GPS.

    But if it means over time you won't always have to carry your iPhone with you that's a good thing.
  • Reply 7 of 88
    mechanicmechanic Posts: 805member
    Quote:
    Originally Posted by Rogifan View Post





    But if it means over time you won't always have to carry your iPhone with you that's a good thing.



    If you read her blog she said exactly that:

    Quote:

     After you run with Apple Watch and your iPhone a few times, the Workout app knows more about your stride. So you can run on a treadmill or outside without your phone and still get a really accurate workout summary.


    So the ?Watch learns in conjuntion with your iPhone and then works well alone.

  • Reply 8 of 88
    Maybe it will just work better than the word suggestions in iOS which Apple also stated would learn which words you use most often and suggest them. I've had this 6 Plus for about seven months now and it continually offers or corrects words that I had no intention of using.
  • Reply 9 of 88
    mac_128mac_128 Posts: 3,454member

    I have to say that picture she took of the watch does it no favors. It looks like a plastic kids watch. This is definitely a case of needing to see that jellyfish screen in person. 

     

     

    Fascinating to see the bands in action. It looks like there's a metal plug in the center where the latch button is. Unlike a lightning connector, I'm assuming they will only attach one way.

  • Reply 10 of 88
    desuserigndesuserign Posts: 1,316member
    Hate to say it, but the Apple Watch will probably be less accurate than the iPhone for counting steps, due to its wrist placement.

    This is because we tend to move our wrists a lot more than our hips in day to day use (just look at an Italian gesticulating in everyday speech with his arms), which means that there will be many erroneous steps counted with the watch. Having your iPhone in your pocket or handbag is much more reliable due to the greater stability in those places.

    In fact, the reason for the Apple Watch's need to train itself is likely due to this intrinsic weakness.

    In Burns's case, the Watch may not be so inaccurate if she uses it only when running a marathon, as she will adopt a regular motion with her arms. But most people will be using the Apple Watch for counting steps throughout the day, not for a specific workout.
    I'm sure Apple has been working on making intelligent algorithms that anticipate all sorts of things you haven't even though of yet, let alone the really obvious things that have occurred to everyone as they contemplate their navels.
  • Reply 11 of 88
    melgrossmelgross Posts: 33,510member
    Hate to say it, but the Apple Watch will probably be less accurate than the iPhone for counting steps, due to its wrist placement.

    This is because we tend to move our wrists a lot more than our hips in day to day use (just look at an Italian gesticulating in everyday speech with his arms), which means that there will be many erroneous steps counted with the watch. Having your iPhone in your pocket or handbag is much more reliable due to the greater stability in those places.

    In fact, the reason for the Apple Watch's need to train itself is likely due to this intrinsic weakness.

    In Burns's case, the Watch may not be so inaccurate if she uses it only when running a marathon, as she will adopt a regular motion with her arms. But most people will be using the Apple Watch for counting steps throughout the day, not for a specific workout. In that more common scenario, the iPhone will win for accuracy.

    I examined the wrist-worn Fitbit devices and the feedback they have received to arrive at this conclusion. Fitbit do make some non-wrist devices.

    You don't actually know that. I would expect that with all the work Apple has put into this, they likely have thought of this as a potential problem, assuming that it actually is.
  • Reply 12 of 88
    mstonemstone Posts: 11,510member
    Quote:
    Originally Posted by Benjamin Frost View Post



    Hate to say it, but the Apple Watch will probably be less accurate than the iPhone for counting steps, due to its wrist placement.



    This is because we tend to move our wrists a lot more than our hips in day to day use (just look at an Italian gesticulating in everyday speech with his arms), which means that there will be many erroneous steps counted with the watch. 

    The watch knows when you are just waving your hand and not walking or running. I would expect it can detect an actual gait within a step or two even if you are on a tread mill and your location has not changed. Just more FUD on your part.

  • Reply 13 of 88
    mstone wrote: »
    Hate to say it, but the Apple Watch will probably be less accurate than the iPhone for counting steps, due to its wrist placement.


    This is because we tend to move our wrists a lot more than our hips in day to day use (just look at an Italian gesticulating in everyday speech with his arms), which means that there will be many erroneous steps counted with the watch. 
    The watch knows when you are just waving your hand and not walking or running. I would expect it can detect an actual gait within a step or two even if you are on a tread mill and your location has not changed. Just more FUD on your part.

    NEWSFLASH: the Apple Watch doesn't have GPS. Therefore, it doesn't know your location. All it goes on is your gestures.
  • Reply 14 of 88
    mstonemstone Posts: 11,510member
    Quote:
    Originally Posted by Benjamin Frost View Post





    NEWSFLASH: the Apple Watch doesn't have GPS. Therefore, it doesn't know your location. All it goes on is your gestures.

    If it is paired with your phone it does and if not it doesn't matter because it can identify a gait versus waving goodbye, talking with your hands, writing on a white board, or whatever. It knows what you are doing which is why it recognizes when you raise your wrist to check the time.

  • Reply 15 of 88
    mstone wrote: »
    NEWSFLASH: the Apple Watch doesn't have GPS. Therefore, it doesn't know your location. All it goes on is your gestures.
    If it is paired with your phone it does and if not it doesn't matter because it can identify a gait versus waving goodbye, talking with your hands, writing on a white board, or whatever. It knows what you are doing which is why it recognizes when you raise your wrist to check the time.

    Perhaps; we shall see.

    However, Fitbit have had years of experience of honing gait patterns, and even then, their wrist-worn devices are still significantly less accurate than the iPhone.
  • Reply 16 of 88
    mstonemstone Posts: 11,510member
    Quote:

    Originally Posted by Benjamin Frost View Post





    Perhaps; we shall see.



    However, Fitbit have had years of experience of honing gait patterns, and even then, their wrist-worn devices are still significantly less accurate than the iPhone.

    http://berkeleysciencereview.com/fit-fitbit/

     

    This is an article I found that tested Fitbit and explains the accuracy variables scientifically. It seems to contradict your argument on a couple points. 

  • Reply 17 of 88
    konqerrorkonqerror Posts: 685member

    The iPhone already does this for distance in Healthkit. Go to Privacy, Location Services, System Services. Look what it says there... "Motion Calibration and Distance".

  • Reply 18 of 88
    shenshen Posts: 434member

    Hey Ben, do they pay you more when you get the first reply to an article?

  • Reply 19 of 88
    shenshen Posts: 434member
    Quote:

    Originally Posted by Benjamin Frost View Post





    Perhaps; we shall see.



    However, Fitbit have had years of experience of honing gait patterns, and even then, their wrist-worn devices are still significantly less accurate than the iPhone.



    “We’ve learned and struggled for a few years here figuring out how to make a decent phone,” he said. “PC guys are not going to just figure this out. They’re not going to just walk in.”

     

    Let us know how your predictions pan out...

  • Reply 20 of 88
    shenshen Posts: 434member
    Quote:

    Originally Posted by mstone View Post

     

    http://berkeleysciencereview.com/fit-fitbit/

     

    This is an article I found that tested Fitbit and explains the accuracy variables scientifically. It seems to contradict your argument on a couple points. 




    Newsflash: Ignorant Apple troll is pulling his argument out of his posterior...

Sign In or Register to comment.