GPS Distance errors
 

I have seen several threads about inconsistent distance measurements when using gps on both runs and rides. I am using mapmyride and found an anomaly with a route
that has a turn around. I am curious of anyone that has distance errors are on a route that turns around and retraces steps and if anyone can suggest a way to fix this problem.
I did not understand what the problem was until one day I turned on realtime speed. I have a 1 mile route that has 3 culdesac so I ride around the circle and head back out
on the same road in the opposite direction. What I notice is the speed stays the same till almost the end of the circle, just as I am starting back out on the road part my speed dives
down to 1-2 mph and takes about a tenth of a mile to gradually climb back to the 10 MPH speed I have been going the whole trip around the circle and back out. The circle is
about 50' diameter so I am able to keep my pace the whole way around and back out. It appears the software is getting confused with its gps calculations and for a while thinks I am
going backwards so it slows my speed down. Since speed over time is what distance is I can see how this would make my route appear shorter then it really is, which is what I have
observed. Has anyone else experience this or have any suggestion on how to resolve it?

Ron

Use a GPS device that uses a proper wheel mounted speed sensor

You don't say what the distance the GPS gives nor how you determined the distance was really 1 mile.

You are using a phone, which can do funny things with the GPS readings. Some of them use GPS locations directly. The iPhone may or may not use direct GPS locations. I use GPS on an iPhone (for other purposes) and it's accurate enough.

GPS devices determine location and time. They get total distance by adding the distances between each of the points (measured every second or so). They determine speed from the change in distance divided by the change in time.

At the speeds cyclists go at, GPS not very good for determining instantaneous speeds.

As long as the circumference is accurate, a wheel rotation counter is very accurate for distance and speed. GPS is still accurate for distance. The wheel sensor is better able to measure the extra distance of the bicycle moving back and forth. The wheel sensor takes (roughly) 3 measurements every second and the GPS is generally about 1 every second. GPS can have reception issues if one is in a "concrete canyon" or under heavy tree cover.

There is nothing about cul-de-sacs that would make GPS measurements less accurate.

Try running multiple apps on your phone while you track the ride. Compare then. Run MMR, Strava, Cyclemeter, etc. And see the differences. Does Map My Ride allow you to set your wheel circumference? Do you have it set on default or did you measure yours?

How much is it off by? I'd assume some measure of error. What model phone do you have? Could be a phone issue with GPS recognition. Are there trees in the area that could stifle the signal?

You notice the speed drop while in real time mode...it's not going to be laser precise.

How much of a distance difference are we talking here?

GPS on phone apps are always a crapshoot.

To save bandwidth and power, the ‘tracking’ is fairly granular, so the points aren’t super close together, so if you make a couple turns really close together, like your cul-de-sac u-turn, it may not capture the whole thing

Even though your phone has a GPS chip, it still tries to go off cell tower triangulation, so most apps aren’t ‘true’ GPS. In addition, many of the map-based apps try to force your track onto the street map; I’ve had times (running) where spotty signal had my track bouncing back an forth between two adjacent streets, making my pace 2-3 times faster than it should have been, over that stretch b
Finally, I’ve found that some phones are better than others; my old HTCs were very good, and my IPhone 4 was pretty reliable, but the l6 and 8 I have now were very particular; they need good weather and certain cases/holders to have reliable routes tracking.

I can see with your reply and that you are a senior member that this forum is not a serious forum for people looking help. I will try to find another forum that doesn't have smart asses as senior members. :) Enjoy your trolling.

How do people get through life without a sense of humor?

Without knowing what kind of "GPS" unit we're talking about, all that can really be done to resolve this problem-- insofar as it is one at all-- is to get a high-accuracy satellite receiver which uses both GPS and GLONASS and samples at high speed, like 10Hz. You may be able to find somehing which uses the super accurate GNSS/Galileo satellite network and/or maybe BeiDou network as well as GPS and GLONASS. Maybe we're simply talking about upgrading from an iPhone 6 to iPhone 12, or maybe you need to get a unit with killer antennae like the Garmin GPSMAP 66ST, but with the info we have, there's no way to say much other than that getting around a 50' cul de sac at 10mph does not seem like a data-intensive undertaking.

even two garims don't always match though it is closer to maybe 1/10 of a mile

Increase your distance and small anomalies will become relatively smaller.

Have you ever watched an almost-finished road construction projects where one of those machines comes through and spits a small blob of pain every second to mark where the fog lines, center lines, etc. will go?

Now imagine your GPS does that. And imagine the electronics inside the GPS tries to fit a straight line to the last 3-7 points (to take out random position error that's going to pop up), and puts an imaginary point on that line after every "pfft" of paint. When you go around a cul-de-sac, you may get 4 imaginary paint spots around the curve.. Pretty soon the fitted lines are going nuts, and the distance from the end of one curve fit to the end of the next curve fit gets very small. Your GPS does a simple distance/time calculation of your speed and it looks like you're going very slow. It may also artificially reduce the apparent distance you rode.

A calibrated wheel sensor fixes those problems.

Gps
 

Come on guys - Ron's first post, go easy!!!

Hi Ron,
Welcome to Bike Forums.
GPS in the phones can introduce errors. Dedicated bike computers do a better job.
Also when using a dedicated bike computer you have the option to also use a wheel mounted speed sensor.
The wheels sensor will permit the bike computer to better detect pauses in your ride, and can even be used to determine that the bike is still moving in areas of low GPS signal strength.

All the best

Barry

Hey rsiesta, the GPS receivers we use are accurate enough for what we need them to do. As you've been advised, adding a speed sensor to your wheel will improve this. I know that is not a satisfying answer, but to get one that gives the kind of precision you are hinting at, you'd have to spend more on the GPS receiver than the bike you're riding. Think about it like this: if you aren't guiding bombs and artillery, don't sweat the loss of a few meters here and there.

There's an article in Bicycling that has some information why there are GPS errors, and what you can do to avoid them:

How Your GPS Lies to You

Really? Please leave this thread.

Never mind him. this is not allowed. You have a lot of good replies. Sorry you had to deal with this.

While I havent verified the GPS sat vs wheel sensor, I have looked over some routes that I sometimes dupe back around on & have noticed it does "trace" some missed areas. That is just on the mapped section of the device. The distance, time, mph is realtime, so it is accurate based on wheel revolutions & ride time.

I use Specialized Mission Control and Gaia to track my routes. They usually vary about .1 or .2 miles difference over a 15 to 20-30 mile rides. Elevation data is even more out of whack.

Was writing "don't ride in tight circles", don't know what happened, lost internet connection!

I also use Mission Control, plus rideWithGPS. The results are usually very, very close (last ride reads 46.10 miles in mission control, 46.1 in rwGPS, ride before that, 20.7 in rwGPS, 20.82 in mission control). zooming way in on the GPS track through known urban areas shows that the straight line segments between measured or interpolated points results in a slightly shorter distance than the actual arc the bike is traveling, plus the GPS based measurements are horizontal distances, while the bike's rear wheel sensor (used by mission control in combination with GPS) would show a longer distance over an incline. Just for kicks i once rode a very very wiggly route down a bunch of 20% streets here in SF, and the results were much more divergent between the two apps.

What I don't know is how an app like mission control (or a bike computer which has both wheel sensors and GPS!) integrates the conflicting data. Wheel sensor says 50 feet has passed in the last second (with no additional data), GPS says one second ago you were at XX,YY and now you're at UU,VV, and the distance between those two is 49 feet, not 50.... does anyone know how they reconcile that? Separate data streams which simply don't agree, and the application viewing the data simply chooses?

Depending upon satellite geometry and the number of satellites 'visible' at the time of observation, an un-augmented single GPS receiver can only deliver accuracies with around 10-30 meters. GPS receivers are also considerably weaker in elevation (Z) computation than they are in X and Y. Now, you probably will get much better accuracy, but statistically speaking as long as it's within that 10-30 meter circle, those computed coordinates are about as accurate as you'll get for a single observation. Augmenting these computed GPS coordinates with other stuff like cell towers or known locations of nearby wifi stations allows our cell phones to get within a few feet (~1-2 meters) most of the time, but without going to a survey grade GPS unit with differential correction, you'll have difficulty getting repeatable accuracy better than 1-2 meters. A survey grade GPS with differential correction can get you into millimeter accuracy range, but nobody would want to ride with a device that bulky, cumbersome, inconvenient, and freakishly expensive.

Since these apps/devices that track our rides are computing distances based upon a series of individual observations, each with a potential error of several feet in any direction (including up/down), you can see why you'll never get perfectly matching distances for a nearly identical route measured multiple times.

My question is, do you actually need accuracy less than a foot per observation? Consider that your phone/device is taking faint signals from a bunch of super-accurate clocks spinning around the globe at 14,000km/h and somehow crunching that down to an X,Y,Z location. I'd say that getting within a few feet without any measuring devices at all is pretty darned impressive. Probably best to just enjoy the ride and not worry about a few feet of difference.

Ditto, the article terrymorse linked to. I see this often with my bike computer, Android and iPhone. Depends on weather/atmospheric conditions, local interference, etc.

Some apps are better than others at reconstructing our most likely routes when the device loses and later regains GPS sync. Some apps show us cutting across open fields through barbed wire fences, directly through buildings, etc. That used to happen more often with Strava but I don't see that kind of error so often now. I'm assuming Strava has tweaked its software to better guesstimate our most likely routes based on the surroundings, roads, etc. Back around 2016-2017 Cyclemeter for iPhone was the most reliable app I used.

Recently I've been running laps at the school track during my usual warm-up and cool-down routes. My phone and bike computer GPS usually record the warm-up and cool-down routes pretty accurately. But the track laps will be all over the place. Sometimes it'll accurately trace my steps for a lap or two, then it will seem to show me cutting across the middle of the field, or straying way off the track.

No biggie, these are just workouts. If I was concerned about my lap times I'd carry a stopwatch.

Unfortunately not true.

It is update rate. the GPS only records a position every so many seconds (or fractions of one). It computes distance traveled by taking the distance between the two waypoints and dividing by time to get speed. If the GPS records one waypoint at the entrance to the circle and one at the end, well it doesn't take a math genius to figure out it won't compute an accurate distance.

The Garmin I have varies how often it records the waypoints saving space on the memory card. it also has an option to switch to a fixed, fairly high record rate. Recording the waypoints more frequently improves the accuracy of the calculations.

But fortunately the bias errors like you describe are highly correlated in time, the error changes slowly with time. Thus while the absolute measurement may be off by 30 meters, they are all off by the same amount. Thus distances computed by successive differences are much more accurate than the 30 m bias number would suggest.

I had it explained to me once that for something like a u-turn, the accuracy isn't good enough to pickup the distance. Same for when doing a run on foot on a curvy greenway path. It fails to pickup all the little bends accurately. Despite what was said above about accuracy being fine, just look at the track made if you do some repeated cul de sac loops on the same area.

FWIW, the accuracy on runs is off enough using my bike Wahoo Elemnt that on curvy greenway runs where it is FLAT is where I have my slower runs. Somehow, when I run the perfectly straight streets/sidewalks that are instead on the really hilly terrain.......I'm the same pace/mi or faster.

I've gone out to a flat road and trail near my work for a run and tried the same pace as the curvy greenway and I was noticeably faster. Like 20sec/mi.

If the site or application you use to determine your distance only uses the plots in the log file, then remember that the log files may not have that coordinate that takes you to the furthest distance into the cul-de-sac. As well around sharp corners, it might miss actual corner where you turned.

If you have 1 second recording intervals for your log, then that can still be 20 or 30 feet missed during that one second. Many devices by default are doing more than 1 second interval logging. Some Garmin's by default do what they call "Smart Recording". I've seen enough differences in my logged track with that to know there is nothing smart about it.