Innacuracies : Strava vs cycle computer

larkim

Harry - you clearly know your stuff, so I don't want to sound disrespectful, but how does your assessment of 25m accuracy sit with recreational devices which consistently report signficantly less than that - usually, in my experience since SA was switched off, 4-5m. Recreational activities like geocaching simply didn't work in the SA days when a GPSr couldn't send you with a high degree of accuracy to the correct tree in a forest, whereas today that is something which I do with alarming regularity (sad, I know!). The devices / software self reports accuracy - are you saying that that self reported accuracy is inaccurate?

harry-s

No worries larkim, as I say I don't know a lot about the recreational units.
My first thought is that if you are looking for a cache, if the unit gets you within 10 to 20m are you maybe subconsciously drawn towards a likely caching feature (tree or wall etc) because there aren't any other candidates around. I'm not saying that you do this, just wondering if it's possible.
I know in the US they have something called WAAS (wide area augmentation system), which uses a geostationary SV to refine the sensor position, a lot of the garmin hand held units (e-trex etc) have this capability, and this would get you down to a few metres. As far as I know there's nothing like this in Europe, but as I say I'm not really up to speed on that.
I've just had a look at some GPS data collected recently, - this is with a high grade receiver and antenna (which takes a bit of lifting...), mounted in an area with a very good horizon and left running for a few hours. This just about gets it's position down to around 2.5 to 3m (processing the recorded data will allow me to get it down to a couple of millimetres). It seems unlikely that a hand held single frequency unit would get near that at 4-5m.
An interesting test would be for you to go to a well defined spot somewhere and take a note of the co-ordinates your hand held displays, then go back again a few hours later or the next day to the same spot and do the same. Do this three or four times and compare the co-ordinates and see how closely they agree, this should give a more objective indication of the accuracy. Also worth pointing out that GPS is more accurate when it's dark, maybe you were geocaching at night?
Getting a bit off-topic here, sorry chaps....

larkim

WAAS / EGNOS is available to handhelds in Europe so that may account for some of the precision. And point taken about notable features, but not always the case, that's part of the fun of caching!

funkyg

Thanks Harry-S for the info, very informative.

Harry-S wrote:

GPS as you no doubt know, stands for Global Positioning System, - so you are allowed to snort when you hear the media stutteringly refer to it as the GPS system, - I know I do.

That's like when people talk about PIN numbers. If they knew what PIN stood for would they still call it a PIN number or just a PIN? It annoys me, even though it sounds better.

ben@31

Harry-S wrote:

I've just had a look at some GPS data collected recently, - this is with a high grade receiver and antenna (which takes a bit of lifting...), mounted in an area with a very good horizon and left running for a few hours. This just about gets it's position down to around 2.5 to 3m (processing the recorded data will allow me to get it down to a couple of millimetres). It seems unlikely that a hand held single frequency unit would get near that at 4-5m.
An interesting test would be for you to go to a well defined spot somewhere and take a note of the co-ordinates your hand held displays, then go back again a few hours later or the next day to the same spot and do the same. Do this three or four times and compare the co-ordinates and see how closely they agree, this should give a more objective indication of the accuracy.

I've accidentally replicated this by running on a treadmill with a Garmin Forerunner watch ( a sensor on your running shoe still gives you pace and distance when on a treadmill). On the map afterwards you can see a track that spokes back and forth around the treadmill like an asterix symbol, where the accuracy of the GPS has drifted. I presume it would be the same on a turbo trainer.

Harry S... Interesting description you gave us. The way I pictured it...
Our GPS devices has an almanac of where the satellites are going to be.
The time signal tells the GPS device it's distance from the satellites.
As the almanac tells the GPS where the satellites are and the signal tells the distance from said satellites, it can work out where you are in relation to the satellites.

harry-s

That's about it Ben. The almanac comes down with the GPS signal, so every time your unit is switched on and receives a signal from a satt then it's almanac gets updated.
It's interesting (well, I think so...) that the position is calculated from the satts on a global co-ordinate system (lat & long), and that position is independent of planet earth, ie you'd get co-ords even if there wasn't a big lump of rock here. Those co-ords have then to be related to the earths surface, which unhelpfully is slipping and sliding all over the place, so co-ords for a point on the surface will differ over time, as that point moves. For us in the UK we have the OS co-ord system for our mapping, and the OS have a network of firmly anchored receivers across the country, and these are considered 'fixed', and all mapping is related to those. Obviously they aren't really fixed, but it's a good way to keep things consistent. Any accurate GPS co-ords anywhere in the world have to be accompanied with a date, so it's position can be adjusted to suit the tectonic movement in that area. I worry about these things when I'm out on my bike....

Mikey23

Aah, tectonic movements, thats why i havent got any KOMs...

Jules Winnfield

Harry-S wrote:

The GLONASS signals are available to the scientific/professional community, and you too if you want to pay for it, but none of the recreational units utilize it (as far as I know). This does mean though, that if Garmin and all the rest develop or buy software to decode the GLONASS signals, there will be another hike in the accuracy achievable

I will have to correct you there I'm afraid. The Garmin Edge 510 uses both GPS and GLONASS for its positioning

harry-s

I will have to correct you there I'm afraid. The Garmin Edge 510 uses both GPS and GLONASS for its positioning

I wrote the above text five years ago, (and I did say "as far as I know"...).

Hats off to Garmin though for adopting GLONASS, although I wouldn't bet against the Russians turning it off to make a political gesture. In fact it threw a complete wobbler for a day earlier in the year, and most GLONASS receivers failed to get a fix, some did and there were reports of navigational errors of several kilometres (the satts had been uploaded incorrect data about their positions). That's one of the reasons why commercial aircraft don't use GPS for autopilot purposes, the aviation industry doesn't consider it reliable enough.

Jules Winnfield

Sorry Harry, I didn't realise you wrote this that long ago. My bad.

Harry-S wrote:

Hats off to Garmin though for adopting GLONASS, although I wouldn't bet against the Russians turning it off to make a political gesture.

I find it strange though that Garmin didn't use GPS and GLONASS for the 810.

markhewitt1978

I have a similar setup, iPhone in my pocket, cheap computer on the stem. The computer underreads my distance by about 5%. But the answer is that it doesn't matter, personally I go off the Strava distance as it's always further

harry-s

No worries Jules, - I dug it out from the attic somewhere as I thought it might help with a few common GPS q's on the forum, as well as help the OP. To receive the GLONASS signals a different antenna is usually required, or at least modifications to a GPS one, and I guess developing this is what slowed Garmin down. With all the development (and money) that's going into androids these days, Garmin deffo can't rest on their laurels.
I bit the bullet and bought a Garmin (800) last year, and as I feared, the thing drives me nuts. It's a bit of a love/hate relationship, like most gadgets I suppose, - when it works it's great, but I know when the little sod is making things up, and am prone to shouting "liar!" at it, when I think no-one is listening...

BikingBud

Harry-S wrote:

That's one of the reasons why commercial aircraft don't use GPS for autopilot purposes, the aviation industry doesn't consider it reliable enough.

Hi Harry great summary but GNSS, yes GPS is the US DofD owned system, is used by commercial airlines with GBAS including for autloand facilities:
http://www.propilotmag.com/archives/200 ... LS_p1.html

Admittedly this is a blended system using other sensors, such as initial reference systems and radar altimeters, as you rightly highlighted the GNSS derived altitude is likely to be insufficiently accurate whichever model we use of the surface of the earth, most systems use WGS84.

All quite scary to let people know that the pilot doesn't land the aircraft :roll:

I use a garmin 800 and a Suunto Ambit watch they are both quite similar in terms distance and altitude change recorded but they use a barometric altimeter to measure rather than comparing with map data.

elizasusan84

I would Love to win this watch because I’ve just recently started running. This will come in very handy! Thanks so much for the giveaway.

llu02

I use Strava from my phone, but the battery does not last long. Computer much better