Results 1 to 8 of 8

Thread: Logging and drift issue

  1. #1
    Low Bitrate
    Join Date
    Dec 2007
    Location
    NY
    Posts
    88

    Logging and drift issue

    Anyone who has made a distance logger, I'm having an issue with drift. Basically while there is no movement distance is being logged because of drift (I'm assuming). I'm hoping to get some ideas of how to compensate. Currently I'm setting it up to make sure the speed is more than 3mph to log the distance, although helpful, it isn't really working all that great.

  2. #2
    North of the land of Hey Huns
    Auto Apps:loading...

    Join Date
    Jun 2004
    Location
    Westminster, MD
    Posts
    1,922
    Are you talking of logging via GPS? If so, I'd make my own distance calculations from the returned location, rather than relying on a "speed" value returned from a GPS program

  3. #3
    Variable Bitrate NeonDev's Avatar
    Join Date
    Feb 2008
    Posts
    431
    Edit: Malcolm beat me.

    are you just using a timer to log distance every so often or does your logging even get triggered by incoming data from some piece of hardware or other process?

    Instead of logging only speeds of > 3mph you ought to log only distances greater than say a meter or whatever fits your needs best.

    say you are using a timer and the difference between your last position and current position is less than _certain distance_ your logging function either doesn't get called or ignores the event; if the difference is greater than or equal to your threshold then do log the distance.

    Does that help or am I misunderstanding you? you didn't really give a lot of detail about your needs and what you wish to accomplish or how you are trying to accomplish it
    check us out at: www.neonboombox.com

  4. #4
    Low Bitrate
    Join Date
    Dec 2007
    Location
    NY
    Posts
    88
    I'm using the speed strictly as a "drift prevention". The distance is being calculated from previous position and current position.

    @NeonDev
    The timer and distance makes "fundamental" sense, but then wont that fall into "how far" is the actual drift? Because at times I've seen drift far larger than 1 meter. .....although I could prob. do a combonation of both the speed and distance traveled.

  5. #5
    Variable Bitrate NeonDev's Avatar
    Join Date
    Feb 2008
    Posts
    431
    if your gps is drifting more than your desired threshold when you are stationary you either need a better GPS (or to relocate it) or to increase your threshold. If you don't care about precision down to a few meters then just increase your distance threshold to say 3 meters sampled say every 10 seconds or so. that way if you move/drift less then 3 meters in 10 seconds it doesn't log it as movement at all.

    I'm not sure you will be able to accurately measure your distance with your GPS (to your desired precision).
    check us out at: www.neonboombox.com

  6. #6
    Mod - Driveline forum
    Auto Apps:loading...
    ClockWorK's Avatar
    Join Date
    Oct 2003
    Location
    Michigan
    Posts
    137
    In short, I would reset the previous position only after ensuring that the speed had gone over the threshhold since the last position update, and not every time the GPS refreshed.

    My threshhold for detecting a stop is 1 knot, and it very rarely thinks I' moving when I'm not. My GPS receiver sits on my dashboard. I might have to try adding a distance-calc to my program now, just to see if it will work..

  7. #7
    Low Bitrate
    Join Date
    Dec 2007
    Location
    NY
    Posts
    88
    @ClockWorK
    That is pretty much what I have. Basically, IF speed passes my threshold, do the calculation, and if there is ONLY a difference from previous, then log it. But at times, rarely but it is still happens, I get "speed drifts", where I will see my gps register 1-3mph and then it will log something. What I am going to do is take NeonDev suggestion and check for a specific distance change; but I'm going to do it AFTER the speed verification test. That way I have two failsafe. It will be an on going test prob for the next couple of months to find the perfect calculation values between the two.
    I'm using i-Blue 737 which I believe is a very good receiver. Attached is how much drift I got in the span of an hour stationary (from home not the car dashboard, but the amount of satellites I have should compensate well...I believe) along with satellites I have in view.

    The green square is where I started from (and I'm amongst buildings in this particular instance)
    Attached Images Attached Images   

  8. #8
    Mod - Driveline forum
    Auto Apps:loading...
    ClockWorK's Avatar
    Join Date
    Oct 2003
    Location
    Michigan
    Posts
    137
    I thought these things were supposed to be accurate to within 15 or 20 yards. You shouldn't have drift of 100+ yards while stationary. I would probably blame the buildings (I'm guessing). I've read that some GPS receivers are better with tall buildings and cityscapes, and provide faster locks.

    I'm not surrounded by tall buildings, and I typcially have 9 satelites in use, 11 in view. Maybe minus 1 or 2 if it's really cloudy.

    I'll think about this some more.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •