Related
Options

Is UltraTrac still broken?

1271255
over 1 year ago

As far as I understand, UltraTrac is using GPS only once a minute, and the intermediate position is calculated using the accelerometer. But is the accelerometer so bad that the speed is dropping to zero after 20 seconds without GPS?

Personally I dont care about a perfect map and the speed doesnt need to be spot-on, but its actually way off. Would be nice if there would be at least an option to set the interval-rate of GPS.

 

  • 0 over 1 year ago

    No idea how the watch should get that data without gps? I would suggest getting a speed sonsor if you dont want to use GPS which you mount within thr weel. The watch can only recognize that you move (at least in my mind...), but not how fast. Running would at least have arm movement, but this detection is also not really great....

  • 0 over 1 year ago in reply to EnormoDerClown

    its as easy as using the acceleration to determine a change in movement. If you move constantly with 20kph before GPS turns off, than you just look for a change in the acceleration. If there is a deceleration than the user slows down. If there is an acceleration, the user speeds up. 

    of course there is "drift", over time inaccuarcies in the readout will sum up and the deviation between real speed and the integrated speed will differ more an more. But thats just a thing how many decimal places are used to calculate the speed. I dont know how good the sensor is the Fenix 7 is, and how many decimal places are used to calculate the speed. So thats more a question to the developers of Garmin.

  • 0 over 1 year ago in reply to 1271255

    Considering how much extra random acceleration you get on your wrist from bumps on the road etc, and how tiny the acceleration/deceleration from pedalling is compared to that (unless you do an emergency braking), I would be surprised if the accelerometer could be used to detect changes in cycling speed.

    For running it's different. There the watch/HRM strap doesn't mainly even try to measure direct acceleration or deceleration in the direction of movement, but it uses the accelerometer to sense the running style and cadence, and estimates running pace from that.

  • 0 over 1 year ago in reply to Bitti
    Bitti said:
    Considering how much extra random acceleration you get on your wrist from bumps on the road etc, and how tiny the acceleration/deceleration from pedalling is compared to that (unless you do an emergency braking), I would be surprised if the accelerometer could be used to detect changes in cycling speed.

    its the other way around. the mathematics to filter out bumps and so on is already a few centuries old. This can be done pretty efficiently, and then you get a nice curve over time, in all 3 dimensions of the acceleration. And yes, the change in cycling speed is actually a pretty significant value. 

    Getting the same result from running is way harder due to the motion of the arm, but even here its possible to get the speed soley from analyzing the accelerometer without simply counting the steps. 

  • 0 over 1 year ago

    Ultratrac really isn't designed for cycling or any other faster moving activity. It's really only designed for walking/hiking.With slow moving activities the distance between infrequent data points isn't so much so the interpolation algorithms work OK. At 30km/h there will be 500m between each data point.

  • 0 over 1 year ago
    1271255 said:
    the intermediate position is calculated using the accelerometer.

    Do you have a source for your statement ?

    Where did you get this information?

    There is nothing about it in the support article.

    Setting a Garmin Outdoor Watch to UltraTrac Mode for Longer Battery Life

    https://support.garmin.com/en-US/?faq=3yEuP9Wb1H5kIpJY7miJ96 

  • +1 over 1 year ago in reply to bluefish
    bluefish said:
    There is nothing about it in the support article.

    >99% of the things happening in the watch are not documented (at least not for us users, hopefully the developers have at least some sort of code-documentation).

    So its just an assumption based on the way the readings compare to full GPS.

    But my question is directed towards the developers, only they can answer my question, I hope someday a dev will look into the forums (though I have the feeling that never happend until now)

    At this point I will close the thread and either a dev reopens it or the question will never be answered.

    But as a matter of fact (Im a developer myself), what Im claiming IS possible. It is possible to interpolate / integrate the position soley based on the accelerometer for quite some time. Thats a fact that nobody here is able to proof wrong (there is a number of research papers that proof me right)

  • 0 over 1 year ago in reply to 1271255
    1271255 said:
    my question is directed towards the developers, only they can answer my question, I hope someday a dev will look into the forums (though I have the feeling that never happend until now)

    You are in a user-user forum. It is not sure that a developer is reading here.

    The best thing will be you send your suggestion for improvement via the form directly to Garmin:

    https://www.garmin.com/en-US/forms/ideas/ 

    1271255 said:
    what Im claiming IS possible. It is possible to interpolate / integrate the position soley based on the accelerometer for quite some time.

    However, you should keep in mind that the accelerometer and the processing power for calculation also consume power, and thus Ultratrac would then no longer be as battery-friendly as before. But battery economy is precisely what the Ultratrac function is all about.

    Since the fenix 7 has a very good battery life even under normal GPS settings, I have never needed "Ultratrac". Furthermore, you must remember that with Ultratrac no VO2max is recorded.