Difference is GPS accuracy

[Flatdog7]

I have used my iPhone7 to track my rides but the GPS accuracy has always bothered me. It trivial I know, but annoying to see how it goes all over the place. I recently bought a second hand TomTom Multisport. Tehnd I found out that Tomtom pulled out of making them 2 years ago. Nice to see my lucks holding.

 

Anyway, just by way of trivia, I thought I would post the difference between the two. First image iphone 7, second is TomTom

Edited April 10, 2019 by Flatdog7

[Simon123]

Done at the same time?

 

Reason I ask is that I once left my GPS on, on a table before stopping it.   It was all over, obviously due to movement tracking earth / satellites (didn't google, just a thought), with never more than 10m or so, but everywhere.   

 

I tested various GPS all at once and the results were eye opening.   I did feel like a chop with 3 watches on, a phone in each pocket and two GPS's on my handlebars.  But the debate needed settling! 

[Quagga]

There is a long discussion on the hub about this and someone explained the technical stuff and only sort of understand it.

 

It has to do with many factors but here is just 2 of them

1. The amount of seconds between GPS points being recorded. Mosyt GPS devices are set to record points every 3-6 seconds. Now if you going around a corner you will often see a straight line between 2 points, and that is because the device "wrote" down a GPS co-ordinate and waited 3 or 6 sec and "wrote" down another point.

 

2. Some devices uses a SNAPTO fuction so they dont actually record the the exact GPS location but in order to "try" and stay on a known road or path they will use the "closest" GPS point that is known to the device. So when the device wants to "write" a gps location it will compare it to the closest KNOWN point and then rather use that point.

Now these "road" locations where established by many devices and hence the variants.

[MrJacques]

Current consumer GPS is only accurate to within +- 4 meters.

 

I've had a couple of different Garmin watches (including a high-end FR630) and now a TomTom (which I found to be the most consistent of the watches I tested). There were definitely differences in accuracy. It also depends on various factors like the chipset used, how good the reception is, obstacles like buildings and trees, etc.

 

DC Rainmaker usually includes some accuracy data when he does reviews.

Edited April 10, 2019 by Jacquers

[Stretch]

Where's the screenshot from? It's not Strava and it's not Garmin. Just remember that some apps (like Strava) do their own thing with the GPS data.

 

Further to what's been mentioned above Garmin devices also enable you to use glonass. Tom Tom might too.i have that turned off on mine.

 

There no correction on the data with your phones and watches so as mentioned the accuracy is not great and each device and software processes the data with different algorithms

[dave303e]

That is frequency of recording and then smoothing applied to it hey. If you trim the file down to the raw coordinates and plot them it will be different again.

 

Phone also used AGPS- Assisted GPS which allows it to include cellphone tower signal to locate position, while this aids speed at which a position can be calculated the accuracy is not always amazing

[OnTheRicky]

The big guns like Uber and Google have entire dev teams working on improving GPS device accuracy with fancy post-processing. I don't think companies like Garmin or Polar etc. have the same resources to put behind activity tracking and they probably just rely on the tech to improve and then use it in future product cycles. Look at Polar with wrist-based optical heart rate sensors - they didn't add it to their products until they were happy with it. Also, how do you do this for off-road activities? 

In the big picture, I think GPS accuracy is round about 5% accurate on average, which is more than enough for most use cases. 

 

It comes down to what Quagga explained above... what is missing is clever processing to make sure your device sticks to the "known" route (which companies like Uber and Google have developed.) On a technical level, they essentially build models that behave the way we know the real world should. If you ride your bike/car on a road the processing will ensure that when you take a turn your GPS location gets adjusted to stick to the road. The latest and greatest tech coming to the party is using wifi hotspots to triangulate your position in when GPS connectivity is low, for example in urban areas.

TLDR: It is what it is

[bertusras]

Where's the screenshot from? It's not Strava and it's not Garmin. Just remember that some apps (like Strava) do their own thing with the GPS data.

 

But it is from Strava?

[dave303e]

Suunto has the Fused Track post processing now and it is actually fairly impressive, so they are on it, just nowhere near where the big guns esp guys like Trimble etc

[MrJacques]

https://www.theverge.com/circuitbreaker/2017/9/25/16362296/gps-accuracy-improving-one-foot-broadcom

 

https://www.theverge.com/2018/11/15/18097154/fcc-ajit-pai-gps-location-accuracy-galileo

[Stretch]

Suunto has the Fused Track post processing now and it is actually fairly impressive, so they are on it, just nowhere near where the big guns esp guys like Trimble etc

It's important not to confuse post processing with differential correction. Differential correction uses base stations nearby with known locations and applies that correction to the actual gps data. This can be done during the collection on the fly or afterwards. Trimble et al does differential correction. Post processing is essentially just playing around with the data using map features, cleaning out outliers, smoothing etc etc.

 

The other thing to be aware of your view to the satellites. They cruise around the equator, so if you could to the b northern sky (in SA) if impeded, your reading is affected.

[Jehosefat]

The other thing to be aware of your view to the satellites. They cruise around the equator, so if you could to the b northern sky (in SA) if impeded, your reading is affected.

 

GPS satellites are not in orbit above the equator. There are 6 distinct orbits (iirc) that they use with a few satellites per orbit none of which are directly above the equator. If they placed all the satellites in orbit above the equator, the positioning would be useless anywhere close to the poles and less accurate than it currently is everywhere else. 

[Flatdog7]

Where's the screenshot from? It's not Strava and it's not Garmin. Just remember that some apps (like Strava) do their own thing with the GPS data.

 

Further to what's been mentioned above Garmin devices also enable you to use glonass. Tom Tom might too.i have that turned off on mine.

 

There no correction on the data with your phones and watches so as mentioned the accuracy is not great and each device and software processes the data with different algorithms

 

Its from Strava, the iphone uploads direct and the TomTom app sends to strava.

[Stretch]

Its from Strava, the iphone uploads direct and the TomTom app sends to strava.

Ok... Strava does it's open thing with the data so from Strava you're viewing modified data

[Stretch]

GPS satellites are not in orbit above the equator. There are 6 distinct orbits (iirc) that they use with a few satellites per orbit none of which are directly above the equator. If they placed all the satellites in orbit above the equator, the positioning would be useless anywhere close to the poles and less accurate than it currently is everywhere else.

You are right.... My mind is in the clouds with the satellites... That's referencingg the rtx satellite used for auto correction of gps data... As you were