Iphone 4 GPS accuracy vs Cycle computer
wesdev
Posts: 104
Hi on a ride in wales last week my mate did 18 miles on his cycle pc and on my iphone 4 it said 16?
Which one would be more accurate? I cranked up the settings on the iphone 4 (everytrail app)
Which one would be more accurate? I cranked up the settings on the iphone 4 (everytrail app)
08 Trek Fuel EX 8
0
Comments
-
The iPhone GPS.
Cycle computers are notoriously used with the wrong true circumference. Mark the floor, put the valve exactly over the mark, roll precisely one rotation forward, mark the floor, measure the distance.
Tyre pressures, tread patterns, a change of tyres since last computer set up, will all give erroneous readings.Boardman Elite SLR 9.2S
Boardman FS Pro0 -
Was it a very hill ride? GPS bike computers aren't smart enough to take account that a hill route from point A to point B effectively includes greater distance if it's hilly (think about it).
Having said that, I'd be surprised to see 2 miles lost due to hills. Have a look at your route overlaid on a map, is it accurate (I assume you can remember where you went)? If it is accurate, then so is your iphone 4.0 -
I was one of said riders with the iPhone4, I uploaded the route to Everytrail and it is perfectly overlaid the route that we carried out. Everytrail seems to track the elevation too, so im sure its smart enough to take that into account when calculating distance. I could be wrong though, look what happened with the iphone attenna!
http://www.everytrail.com/view_trip.php?trip_id=744529
I was shocked to see such a difference, although over 16 or 18 miles 1% out with the cycle computer could mount up and explain the difference?0 -
Everytrail doesn't make the route longer if you go over hills. What everytrail does, and more or less every mapping application does, is to look down on the route from above and work out distance travelled.
Imagine looking down from orbit at your route, it's all in 2d yes? Now imagine coming down to the horizon and looking at your route as it goes over hills and into valleys. Now imagine running a huge piece of string over each hill and valley, so it follows the highs and lows accurately. Now stretch out your 16 or 18 mile piece of string and measure it.
See what I mean?0 -
I totally see your point as I had this very same argument with the OP concerning GPS accuracy when we snowboard and calculating distance / speed etc and I was argued your very point! However, the difference with everytrail is that it appears to work out your elevation as shown in the graphical presentation, so if it knows this, surely it can work out distance?
See what I mean? :-)0 -
It knows your elevation, so it tells you how much you've ascended/descended but it doesn't extrapolate this into extending your overall distance.
Only a perfectly calibrated conventional bike computer can tell you the actual distance covered, but I still prefer GPS
There might be a plugin for SportTracks that will work out the extra distance covered due to riding a hilly route, but for most of us it's purely academic. We can get an idea how much effort was in a ride not just from the distance, but the accumulated climbing.
Phew!0 -
GPS is not fantastically accurate, and 2 miles over 16 or 18 miles seems like tolerance.
The system has an inherent tolerance. but I can't remember how much exactly, I think it's around 5 meters or so.0 -
Well I think 2 miles over 16 or 18 is waaaaay outside of tolerance. Put it this way, I ride a measured course regularly here and it always measures a matter of a few metres outside every single time. I also have a cadence/speed sensor attached to my Garmin Edge 705 which further helps to define the accuracy, but I don't use it all the time (multiple bikes) - and still my 22 mile ride always comes out as 22 miles.0
-
I don't quite see what you're getting at.dodgy wrote:Well I think 2 miles over 16 or 18 is waaaaay outside of tolerance. Put it this way, I ride a measured course regularly here and it always measures a matter of a few metres outside every single time. I also have a cadence/speed sensor attached to my Garmin Edge 705 which further helps to define the accuracy, but I don't use it all the time (multiple bikes) - and still my 22 mile ride always comes out as 22 miles.
The iPhone was not designed as an accurate GPS measuring system. Hell, it struggles to make calls, let alone give accurate measurements.
With a device like that, I reckon 2 miles +/- is to be expected really.
Which is WHY, an expensive GPS like your garmin, with assistance from other means, is more accurate.
Kapeesh?0 -
The app I use gives you speed over ground (wheel speed) and most GPS systems should give this.
GPS is astonishingly accurate and the iPhone GPS chipset works the same way as the Garmin's, or Tom Tom's. The GPS chip reads time signals and takes one away from another. Its accuracy is determined by the satellites, not by the receiver.Boardman Elite SLR 9.2S
Boardman FS Pro0 -
Quick question, how do you know your 22 mile ride is 22 miles?0
-
Bullorks. The iphone has no way of "knowing" wheel speed, without a sensor mounted on the bike somewhere.Bar Shaker wrote:The app I use gives you speed over ground (wheel speed) and most GPS systems should give this.
It calculates it based on changes in GPS info, which, as I pointed out, is not fantastically accurate. Particularly so in the case of an iPhone.0 -
yeehaamcgee wrote:The iphone has no way of "knowing" wheel speed, without a sensor mounted on the bike somewhere.
It calculates it based on changes in GPS info, which, as I pointed out, is not fantastically accurate. Particularly so in the case of an iPhone.
Distance to move from one point to another, divided by the time taken is speed.
Any GPS knows where you start (lat, long, alt) and where you finish (lat, long, alt). The distance between these two is easily calculated. I have thought of a possible dilution of accuracy though and it will depend on the settings in the OP's app. You can set the frequency of GPS position checks. The longer the gap between checks the better the battery life but the lower the accuracy. This is because the GPS will use straight lines between the points. I have mine set at 5sec but you can set it out to 2mins. A very short ride around a circular field could look like a triangle to the GPS.
http://www.motionx.com/home/motionx-technology/Boardman Elite SLR 9.2S
Boardman FS Pro0 -
I know what speed is, and I know that wheel speed can be different. A Car going round a bend at a steady speed, fore example, has the wheels on the inside of the bend and outside of the bend rotating at different rates.
Anyway...
I've not seen any GPS device that would simply take a start position, and an end position, and figure out how far it is between the two, and figure out your distance like that. For starters, that would completely negate any corners in your route, or changes of heading, or even a circular route.
Rather, what they do instead is keep a tally of distance travelled, by measuring at certain time intervals.
BUT, the accuracy (again) of the GPS is the weak link here, as each and every measurement is subject to a few meter's tolerance.0 -
yeehaamcgee wrote:Quick question, how do you know your 22 mile ride is 22 miles?
Because I can see the ride overlaid on Google Maps and other mapping providers.0 -
yeehaamcgee wrote:
Bullorks. The iphone has no way of "knowing" wheel speed, without a sensor mounted on the bike somewhere.Bar Shaker wrote:The app I use gives you speed over ground (wheel speed) and most GPS systems should give this.
It calculates it based on changes in GPS info, which, as I pointed out, is not fantastically accurate. Particularly so in the case of an iPhone.
I'm sure if the GPS on my phone can tell me how fast I'm going, where I've been, and distance travelled, then other newer phones can as well without the use of external sensors?0 -
But the external sensors increase accuracy, that's the point. GPS is not particularly accurate at the scale relevant to cyclists. It effectively becomes a kind of "assisted" or Differential GPS.
Does that take elevation into account when calculating distances?Because I can see the ride overlaid on Google Maps and other mapping providers.0 -
GPS is easily accurate enough and its measurements are repeatable in accuracy.0
-
I was thinking about this, and (especially with lower pressure tyres) would a bike computer consistently overestimate? If your tyre is 'squashed' where it meets the ground, then that makes the circumference of the wheel and tyre slightly smaller. When the wheel is rolling, with your weight on the bike, the contact point is constantly 'squashed', having the effect of using a slightly smaller tyre.
So one revolution move you less than the computer thinks (after setting up by rolling along an 'unloaded' bike and measuring the distance of one revolution) and so speeds and distances are overestimated.
Maybe
Not saying that's caused the discrepancy, just something to think about.
I'd always expect a computer to measure further than a GPS because of the constant little turns of the handlebars, that make the front wheel do a little bit more distance, but don't move the bike left or right enough for the GPS to detect it
Imagine riding really slowly along the white line in the middle of the road. If you turn the bars fairly sharply so that you cross the line, then as soon as you do, give it 'full lock' in the opposite direction, until you just cross the line with the front wheel, then turn back, and continue, you'll measure all the side to side movement from the magnet on the front wheel, but probably not through the GPS.0 -
But in the meantime, the GPS signal could, in a worst case scenario, be measuring first time 5m to the left of where you are, then the next interval is at 5m right of where you are, then 5m left again and so on and so forth. Say it measures twice every second, that's a discrepancy that could lead it to think you're travelling at 10m/s, even when you're standing stock still.
There are algorithms so smooth this data out, but the most basic receivers found in phones will not bother with this. And besides, at the speeds a bicycle is travelling, the signal to noise (signal being actual movement, noise being random deviations based on innacuracy) ratio is very high compared to something like a car travelling at 30 mph.
GPS really DOES have inaccuracies. I'm shocked that you are surprised by this, frankly. There are ways of mitigating these inaccuracies but they are expensive, and will not be found on a consumer mobile phone.0 -
All hypothetical, the fact is if you go out with your iphone, nokia or whatever and record a ride it will be faithful enough to the actual distance.0
-
Give or take a bit. Which is NOT hypothetical.
Nokia phones actually give an accuracy figure, you might want to check that out. It's fascinating.0 -
I see this has caused quite a debate, me and the OP were arguing both sides of this argument and having discussed and researched this a little I can see both will have their innaccurcies.
The iphone apparently has the same GPS capabilities as some of the more expensive Garmin GPS units, but even then I appreciate that the GPS signal is only accurate to a few feet. Just to mention I had the iphone set to its most accurate settings, I think it was taking a waypoint every second or so.
With the cycle computers, as I think has been mentioned before, surely these are only as accurate as you set them up, so instance if you are just 1mm out in your measurements, be it poor measurement, tyre wear, tyre pressure etc then multiply this out over the course of 20 miles and you will get quite a lot of descrepancy.
I guess the solution to my questions is just to use one and although it may be inaccurate, at least its inaccurate every time i ride :-)0 -
Define "Same GPS Capabilities".
Does that mean, for example, It can locate, and track distance travelled, and track altitude etc?
OR, does it mean, it is as ACCURATE as a more expensive Garmin unit? The two are very different things.
As for taking a way-point every second or so, that's still not going to do a great deal to mitigate the inherent inaccuracy of the system.
There are weaknesses to both systems. That's all I'm trying to say.
I did not realise you had such emotional attachment to GPS technology.0 -
iPhone uses assisted GPS, so in the event that a GPS signal was lost, it would use information from cell towers which is a lot less accurate. It's hit or miss as to how accurate it will be. Do the route again, and I bet you'll get a different reading.
2 miles on an 18 mile route does seem excessive. Draw the route out manually on MapMyRide and see what it says.0 -
yeehaamcgee wrote:But in the meantime, the GPS signal could, in a worst case scenario, be measuring first time 5m to the left of where you are, then the next interval is at 5m right of where you are, then 5m left again and so on and so forth. Say it measures twice every second, that's a discrepancy that could lead it to think you're travelling at 10m/s, even when you're standing stock still.
There are algorithms so smooth this data out, but the most basic receivers found in phones will not bother with this.
True, I've left mine (Nokia 6220 classic, so fairly 'old' in mobile phone terms) on by accident when I've been stopped, and when I look back it will have left a jumble of 'instant' journeys around where I was stopped, especially when I've been stopped.
But when I'm on the move it seems pretty good. It even picked up my evasive swerve to avoid a dopey driver, must have only been a couple of feet of movement, but you could see it on the map.GPS really DOES have inaccuracies. I'm shocked that you are surprised by this, frankly. There are ways of mitigating these inaccuracies but they are expensive, and will not be found on a consumer mobile phone.
I don't know if that was in response to me, but I know they're not perfect, I use my computer for distance, and the phone's GPS for route finding/marking. I've got to say, while we're on the topic, the new version of Nokia Sport tracker is really good. It uses OSM maps (not OS), and the battery life is massively improved.0 -
yeehaamcgee wrote:I did not realise you had such emotional attachment to GPS technology.
What's made you say that? :? 0 -
We are comparing two systems here, one which accurately requires a user to input to a device the circumference of a wheel which depends on tyre pressure and the size of the tyres fitted. The other which depends on several billion quids worth of satellites and was used to deliver cruise missiles to specific addresses in a number of locations and countries over the last ten or twenty years. Are we honestly surprised that one is less accurate than the other and is there any doubt which is which?
Come on guys!
Also at 20% gradient the difference between horizontal distance and the length of the hypoteneuse i.e. the distance on the slant, over 100 miles is 0.125 of a mile or 220 yards (200 metres) so I don't think the discrepancy is accounted for by the difference between horizontal and slant distance.0 -
From Garmin's website....owenlars wrote:We are comparing two systems here, one which accurately requires a user to input to a device the circumference of a wheel which depends on tyre pressure and the size of the tyres fitted. The other which depends on several billion quids worth of satellites and was used to deliver cruise missiles to specific addresses in a number of locations and countries over the last ten or twenty years. Are we honestly surprised that one is less accurate than the other and is there any doubt which is which?Today's GPS receivers are extremely accurate, thanks to their parallel multi-channel design. Garmin's 12 parallel channel receivers are quick to lock onto satellites when first turned on and they maintain strong locks, even in dense foliage or urban settings with tall buildings. Certain atmospheric factors and other sources of error can affect the accuracy of GPS receivers. Garmin® GPS receivers are accurate to within 15 meters on average.
Of course....
Neither of which is available for cycling.Newer Garmin GPS receivers with WAAS (Wide Area Augmentation System) capability can improve accuracy to less than three meters on average. No additional equipment or fees are required to take advantage of WAAS. Users can also get better accuracy with Differential GPS (DGPS), which corrects GPS signals to within an average of three to five meters. The U.S. Coast Guard operates the most common DGPS correction service
So, you have a choice of a few mm per wheel revolution, of up to 15meters each sampling interval.
I'd argue that over very long distance, the GPS would be more accurate, but over a shorter distance, the wheel based would be more accurate.0
