Ari Posted July 27, 2006 Report Share Posted July 27, 2006 Really spurring off from the speed camera thread, but a lot of people seem to rely heavily on their GPS (either Nav system or camera warning system) for accurate speed readout. But how accurate are they? Just one scenario I can think of, surely going up or down hills varies the speed compared to road speed? To illustrate. Imagine doing 100mph across a completely horizontal surface. GPS records speed across the ground at 100mph. Now imagine (if it were possible) doing 100mph vertically. You would still be doing 100mph, but speed over the ground would be zero as that 100mph is straight up. So logically therefore, if you were on a 45 degree hill you're travelling 50mph along the ground climbing 50mph vertically. So your car is actually doing 100mph but your GPS would record a 50mph reading. And what about corners? If you have a 180 degree corner and it's 100 miles form start to finish then at 100mph it will take you an hour to complete the journey from the start to the end of the corner. But if you drove straight from point to point at the same speed you'd get there in much less time so your GPS would record a higher speed. Of course it wouldn't in reality because it is measuring point to point speed in intervals much less than an hour, but non the less surely corners must screw up the reading..? Thoughts please gentlemen? Link to comment Share on other sites More sharing options...
Daz Posted July 27, 2006 Report Share Posted July 27, 2006 Regarding the hill scenario, GPS does also measure height so it shouldn't be an issue. Get Lottie to jump out of a plane again with a handheld GPS to test this. I don't know what the sampling rate is of a GPS but it seems to be fairly quick so the corners shouldn't be an issue either. Link to comment Share on other sites More sharing options...
RedRobin Posted July 27, 2006 Report Share Posted July 27, 2006 I relied on using my RA to give me my speed into corners for having fun through miles of twisties, many right-angles, late last night, and didn't notice any particular lag in the speed display. I feel it's accurate enough and the rest is academic. The point being that GPS is more accurate than OEM speedos and it's not so far out to matter. Link to comment Share on other sites More sharing options...
peterb Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] So logically therefore, if you were on a 45 degree hill you're travelling 50mph along the ground climbing 50mph vertically. So your car is actually doing 100mph but your GPS would record a 50mph reading. [/ QUOTE ] Did you ever do trigonometry at school? Cos(45) = .7071, so the GPS would actually read 71mph! Anyway, how many 45degree hills (1 in 1) do you drive up? I reckon that there aren't many hills of a gradient steeper than 1 in 4 on the public roads, and you certainly wouldn't drive them at any significant speed! A 1 in 4 slope would give you a 3% error. Then, as has already been stated, GPS allows a height reading so, theoretically, your GPS device could take this into account (whether any do, I cannot say). Link to comment Share on other sites More sharing options...
Ari Posted July 27, 2006 Author Report Share Posted July 27, 2006 [ QUOTE ] Regarding the hill scenario, GPS does also measure height so it shouldn't be an issue. Get Lottie to jump out of a plane again with a handheld GPS to test this. [/ QUOTE ] Good point well made. [ QUOTE ] I don't know what the sampling rate is of a GPS but it seems to be fairly quick so the corners shouldn't be an issue either. [/ QUOTE ] Yeah I guess it's quick enough not to matter, but non the less, the way a GPS works out speed is to measure the time taken to get from point to point so therefore if the line from point to point is not straight (ie, you're cornering) it must have some effect surely? Link to comment Share on other sites More sharing options...
Ari Posted July 27, 2006 Author Report Share Posted July 27, 2006 [ QUOTE ] [ QUOTE ] So logically therefore, if you were on a 45 degree hill you're travelling 50mph along the ground climbing 50mph vertically. So your car is actually doing 100mph but your GPS would record a 50mph reading. [/ QUOTE ] Did you ever do trigonometry at school? Cos(45) = .7071, so the GPS would actually read 71mph! Anyway, how many 45degree hills (1 in 1) do you drive up? I reckon that there aren't many hills of a gradient steeper than 1 in 4 on the public roads, and you certainly wouldn't drive them at any significant speed! A 1 in 4 slope would give you a 3% error. Then, as has already been stated, GPS allows a height reading so, theoretically, your GPS device could take this into account (whether any do, I cannot say). [/ QUOTE ] I was trying to give an example of how not being horizontal might affect the speed rather than be technically accurate, but good points. Link to comment Share on other sites More sharing options...
peterb Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] I was trying to give an example of how not being horizontal might affect the speed rather than be technically accurate, but good points. [/ QUOTE ] ... and I was pointing out that for all practical purposes, the error is insignificant. For instance, a 1 in 10 hill (still pretty steep) would introduce a 1% error - much more accurate than your speedo! Link to comment Share on other sites More sharing options...
Ari Posted July 27, 2006 Author Report Share Posted July 27, 2006 Yes, that was one of the good points I referred to. Right, I wonder if Lottie is at 20,000ft with my Tom Tom yet... Link to comment Share on other sites More sharing options...
Scotty Posted July 27, 2006 Report Share Posted July 27, 2006 I think the biggest error is due to the sampling rate of the device or at least the processed result of it. If you accelerate hard then the GPS numbers are often playing catch up....at least they were on older systems. To be honest when I'm booting it I'm not normally watching the GPS speed. Link to comment Share on other sites More sharing options...
RedRobin Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] Right, I wonder if Lottie is at 20,000ft with my Tom Tom yet... [/ QUOTE ] .... Don't look up, Ari! Lottie is about to land on you from a great height! Link to comment Share on other sites More sharing options...
peterb Posted July 27, 2006 Report Share Posted July 27, 2006 Well, I hope she keeps calm and doesn't let go of it - the TomTom, I mean! Link to comment Share on other sites More sharing options...
RedRobin Posted July 27, 2006 Report Share Posted July 27, 2006 Scotty, any 'catch up' is insignificant on latest generation Road Angels. It's very quick. Link to comment Share on other sites More sharing options...
patently Posted July 27, 2006 Report Share Posted July 27, 2006 GPS has the height information, but I'm not sure it is used to calculate the speed. So inclines will cause it to under-read, but not by much. Go up a 10 degree incline at an actual 100 mph and it will read 98.5 mph. Looking at my RoadAngel, the sampling rate is far too high for corners to have an effect, I think. Link to comment Share on other sites More sharing options...
Simon690 Posted July 27, 2006 Report Share Posted July 27, 2006 GPS units calculate speed based on the doopler shift of the satelitte signals and not on a time/distance calulation so the accuracy is not affected by gradients GPS speed accuracy depends on a number of factors but it's usually within +-0.5% which is why all the mags and other folks use it when testing cars and bikes etc. I'll pick up my anorak on the way out Link to comment Share on other sites More sharing options...
scooby_simon Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] GPS units calculate speed based on the doopler shift of the satelitte signals and not on a time/distance calulation so the accuracy is not affected by gradients GPS speed accuracy depends on a number of factors but it's usually within +-0.5% which is why all the mags and other folks use it when testing cars and bikes etc. I'll pick up my anorak on the way out [/ QUOTE ] Indeed, also GPS is working in X, Y and Z (Assuming you have atleast 3 Sats connected). So the GPS is aware of the full direction of travel. When out sailing in a tide, I can (sometimes) see the difference in height between going out and returning. Link to comment Share on other sites More sharing options...
patently Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] GPS units calculate speed based on the doopler shift of the satelitte signals [/ QUOTE ] Wow I'm impressed! Link to comment Share on other sites More sharing options...
Ari Posted July 27, 2006 Author Report Share Posted July 27, 2006 [ QUOTE ] .... Don't look up, Ari! Lottie is about to land on you from a great height! [/ QUOTE ] Situation normal. Link to comment Share on other sites More sharing options...
Simon690 Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] Indeed, also GPS is working in X, Y and Z (Assuming you have atleast 3 Sats connected). So the GPS is aware of the full direction of travel. When out sailing in a tide, I can (sometimes) see the difference in height between going out and returning. [/ QUOTE ] [anorak] Actually to get vertical (3D) measurements you need a minimum four sats [/anorak] Link to comment Share on other sites More sharing options...
scooby_simon Posted July 27, 2006 Report Share Posted July 27, 2006 [ QUOTE ] [ QUOTE ] Indeed, also GPS is working in X, Y and Z (Assuming you have atleast 3 Sats connected). So the GPS is aware of the full direction of travel. When out sailing in a tide, I can (sometimes) see the difference in height between going out and returning. [/ QUOTE ] [anorak] Actually to get vertical (3D) measurements you need a minimum four sats [/anorak] [/ QUOTE ] must go and read manual before posting next time.......... Why, I would have expected to be able to work it out via the doppler on 3? Link to comment Share on other sites More sharing options...
sayerbloke Posted July 27, 2006 Report Share Posted July 27, 2006 The Speed-readout on my PDA GPS SatNav provides an interesting* look at the accuracy of car speedos. In my BMW, it reads 3mph less than the car. In my mate's Celica when he had that, it used to read 5mph less. In my brother's Astra when he had that, it used to read almost 20mph less at high speeds! *Disclaimer: Information contained above may not actually be interesting in the accepted sense of the word, so don't sue me Link to comment Share on other sites More sharing options...
rs32 Posted July 27, 2006 Report Share Posted July 27, 2006 have just had navi (Pioneer) fitted and on the "vehicle dynamics" screen there is an option (in addition to the GPS speedo display) to show live info for the following : Acceleration Lateral Acceleration Angular Velocity (how far it has turned in one second) Slope This suggests that if it can measure all these wonderful things, it may be able to use the figures to help calculate true speed fairly accurately. However there's nothing in the manual to suggest that it actually does use the extra data - so could be a complete gimmick Looking forward to measuring Lateral Acceleration at some point though Link to comment Share on other sites More sharing options...
Recommended Posts
Please sign in to comment
You will be able to leave a comment after signing in
Sign In Now