Honda Shadow Forums banner

1 - 20 of 44 Posts

·
Registered
Joined
·
209 Posts
Discussion Starter #1
There was some discussion about checking speedometers against a GPS readout and I got to wondering just how accurate the GPS was. I've done some reasearch and it seems that consumer grade models are good for an accuracy of about 7 meters 95% of the time. Then I wanted to know how often this information was updated, it seems that most consumer grade units update info about every second. Because I'm not too deep into this yet, I can't be certain, but it looks to like a consumer grade GPS is not a very accurate speedometer for a motorcycle.

I may be making incorrect asumptions here, but it looks like it's entirely possible for a GPS to be inaccurate by 14 meters in one mile. Most electronic counters are accurate to within one pulse per mile. Pulse value can and does vary, but the industry standard, as I know it, is 5280 per mile. So within the standard we should be off by no more than 1 foot in a mile with a properly calibrated odometer.

If anyone is interested in making a comparison I would welcome their input.

I realize that big bucks, special use equipment isn't the same, and I wanted to deal with things that average people may purchase.

This is a learning experience for me and I'm almost at step one, so it'll be necessary for me to catch up with those of you who are already familar with the system.
 

·
Registered
Joined
·
336 Posts
As I understand it, most consumer grade GPSs are now accurate to within 3 meters. Also, you have to take into account the fact that the GPS is not just a counter. It is continually recalculating so that, even after a distance of 1000 miles, it should only be off by no more than 3 meters (or 7 meters by your info).

By the same token, they are constantly recalculating speed. If you take one out on the interstate and get it locked in at 60mph and maintain a constant speed, it will take you 1 minute from mile marker to mile marker. For more accurate results, do this over a distance of several miles. It has been my experience that mile markers are sometimes not spaced exactly 1 mile apart.

As an interesting sidenote, I have both a GPS and a bicycle computer on my motorcycle. The GPS and bicycle computer both show me the same speed (withing a couple tenths of a mile per hour) and odometer readings. The Shadow speedometer is always high on both speed and mileage according to the other two devices.
 

·
Registered
Joined
·
130 Posts
Some GPSes are accurate to within 3 feet with correction, such as with WAAS. However, it would seem to me that regardless where you are located on the map, that speed is still a constant. If you're off one foot, then that foot is linear.
Anyway, there should be something like a two second delay to the bird in space, but at a constant speed, it will be accurate. Even pilots trust their GPS for navigation and ground speed.
My thought is that the error is negligeable and that the GPS is much more accurate then your speedo and odo.
 

·
Registered
Joined
·
209 Posts
Discussion Starter #4
spoupard said:
As I understand it, most consumer grade GPSs are now accurate to within 3 meters. Also, you have to take into account the fact that the GPS is not just a counter. It is continually recalculating so that, even after a distance of 1000 miles, it should only be off by no more than 3 meters (or 7 meters by your info).
spoupard,

I appreciate your help. There's a whole bunch of things I'm wondering about and for as long as you want to keep slamming at my brain I'll try to understand it.

I can see your point about accuracy over great distance! That's pretty impressive for sure. If we took that accurate distance and tried to get speed with a time, I believe it would be incredibly accurate. But, would it be able to calculate our gas and rest stops? Or just a block to block speed over the 1000 miles?


One thing, as far as using for a speedometer, I'm thinking of the unit as knowing it's inside an area that's not more than the error. It may be exactly where it thinks it is or up to 3 meters off in any direction.

Am I on the right trail so far?

Thanks,

John
 

·
Registered
Joined
·
336 Posts
Olivereaman said:
spoupard said:
As I understand it, most consumer grade GPSs are now accurate to within 3 meters. Also, you have to take into account the fact that the GPS is not just a counter. It is continually recalculating so that, even after a distance of 1000 miles, it should only be off by no more than 3 meters (or 7 meters by your info).
spoupard,

I appreciate your help. There's a whole bunch of things I'm wondering about and for as long as you want to keep slamming at my brain I'll try to understand it.

I can see your point about accuracy over great distance! That's pretty impressive for sure. If we took that accurate distance and tried to get speed with a time, I believe it would be incredibly accurate. But, would it be able to calculate our gas and rest stops? Or just a block to block speed over the 1000 miles?


One thing, as far as using for a speedometer, I'm thinking of the unit as knowing it's inside an area that's not more than the error. It may be exactly where it thinks it is or up to 3 meters off in any direction.

Am I on the right trail so far?

Thanks,

John
The GPSs accuracy depends on how many satellites it can see at any given time. It requires a minimum of 3 and may be able to see as many as 12 or so at any given time. The more it can see the more accurate it is. Even when it can see only 3, it is pretty darn accurate. Like Saint said, pilots use them almost exclusively for navigation. They must be pretty good! For any given distance, whether it is 1 mile or 100 miles, it will be off no more than just a few meters.

BTW, I'm certainly no expert on GPSs. Maybe someone who knows more than I will jump in.
 

·
Registered
Joined
·
51 Posts
I have 4 different GPS's.........

The cheapest one is the simple Yellow Etrex. It is one of the cheapest units on the market. I have tested them all against each other and I find little if any differences in the accuracy. They all usually range from 3 to 14 ft, not meters. The speed accuracy, I have checked against many auot and M/C speedo's and they are VERY close.

Rick
 

·
Registered
1999 VT1100C2 A. C. E.
Joined
·
9,222 Posts
I may be making incorrect asumptions here, but it looks like it's entirely possible for a GPS to be inaccurate by 14 meters in one mile.
Not exactly -- that would be true of you could traverse a mile in a second. At 60 miles an hour with sampling every second, the GPS could be off by 60 x 7 meters = 420 meters (1,365 feet or better than a quarter of a mile). However, since the GPS is correcting itself with every sampling, it's practically impossible for this to occur if you're using the GPS in lieu of a speedometer. Besides, the GPS would have to be seriously defective to have every single reading off "in the same direction" -- the only way so much error could be cumulative. In fact, any decent consumer GPS will be more accurate than your ability to distinguish between the numbers on a typical analog motorcycle (or automobile) speedometer.

FWIW, 14 meters in a mile would be an error of 0.86% -- less than 1%. Dont' even try to tell me that you can read a Shadow speedometer more accurately than that!

Accuracy, btw, is not the correct term for describing the error in the GPS' reports; it is actually error in the estimated position.
 

·
Registered
1999 VT1100C2 A. C. E.
Joined
·
9,222 Posts
I can see your point about accuracy over great distance! That's pretty impressive for sure. If we took that accurate distance and tried to get speed with a time, I believe it would be incredibly accurate. But, would it be able to calculate our gas and rest stops? Or just a block to block speed over the 1000 miles?
Depends on the particular GPS and software but, generally speaking, any unit you can buy today can do what you are asking EXTREMELY well -- and do it over a distance of a few feet. It should also include traveling through altitude (like going up or down a hill) rather than just "point to point" -- but, again, that depends on the particular GPS unit.
One thing, as far as using for a speedometer, I'm thinking of the unit as knowing it's inside an area that's not more than the error. It may be exactly where it thinks it is or up to 3 meters off in any direction.

Am I on the right trail so far?
Kind of backwards there; the GPS "knows" where it is as a single point rather than an area; the point it reports may or may not truly be where it is due to reception issues and other factors and the error inherent in the estimate of that position (which is typically reported by the GPS) defines an area (a circle when viewed in 2 dimensions but actually a sphere).

And we are actually discussing a GPSr -- the receiver.
 

·
Registered
Joined
·
3,761 Posts
Good GPS receivers on the consumer end have become much better and
many of them are now showing an accurate position within 3 feet.
Many of the Garmin unit are this way and they even tell you what their
off by.

GPS units can be more accurate than a speedometer, but there are cases
even when that is not true.

In an airplane, they are most accurate because of an uninterrupted view of the sky.

On the ground, in a vehicle, if the receiver is taking in signals a showing
your position, here's where the inaccuracy comes in.

Say you are driving along and drive through a tunnel or anywhere else the receiver can lose it's signal.
From the entrance of the tunnel to the exit of the tunnel, as the crow flies,
is only a 1/4 mile.
Inside the tunnel, there are twisties (the road isn't straight).
The GPS unit will lose satellite signal when you enter the tunnel.
It knows the last coordinate were where you entered the tunnel.
As you exit the tunnel, it sees now sees the coordinates where you
exit and runs a time/distance (point of entry to point of exit) calculation and gives you a speed.
That speed AND that actual distance traveled will be incorrect because you
actually traveled more distance due to the twisties in the tunnel.
All the GPS knows is that you when from point A to point B and how long it
took you to get from point A to point B and doesn't know the actual
distance in between. Therefore, your speed will drastically be
reduced for that amount of time in the tunnel and the distance traveled will
be lower than the actual distance traveled.

On a long trip traveling through areas where the sky is inhibited,
this distance (or lack of recording the correct actual distance) will
cause the GPS to read low. As long as you have open skies and
the GPS stays locked on to 3 satellites or can use the WAAS (wide area augmentation system), it will be pretty correctly.
 

·
Registered
Joined
·
190 Posts
I might be wrong about this, (this is more of a distance thing, not a speed thing) but if you use a GPS to give you the distance covered between waypoints A and B, it's straight line distance ("as the crow flies") unless it adds the cumulative distance between each 1 sec update.

The way it sees speed through a curve, is time/distance between updates, so indicated speed while going through a curve would show up as slower than actual since it "connects the dots" between updates. At 60mph (88 fps) it wouldn't be large, but it would be there.
 

·
Registered
Joined
·
3,761 Posts
Kermitdafrog said:
I might be wrong about this, (this is more of a distance thing, not a speed thing) but if you use a GPS to give you the distance covered between waypoints A and B, it's straight line distance ("as the crow flies") unless it adds the cumulative distance between each 1 sec update.

The way it sees speed through a curve, is time/distance between updates, so indicated speed while going through a curve would show up as slower than actual since it "connects the dots" between updates. At 60mph (88 fps) it wouldn't be large, but it would be there.
Maybe I'm too long winded... you just said what I said in about 3 sentences. :lol: :lol:
 

·
Registered
Joined
·
286 Posts
I don't know much about GPS either, so here's another question: Mostly the discussion seems to be about accuracy in the horizontal plane, but what about going up/down in altitude? For example, horizontally you might only travel one mile up a mountain and one mile back down, but the actual total distance might be, say, three or four miles because the crow does not have to fly up and back down. So does the GPS account for the vertical distance?
 

·
Registered
1999 VT1100C2 A. C. E.
Joined
·
9,222 Posts
if you use a GPS to give you the distance covered between waypoints A and B, it's straight line distance ("as the crow flies") unless it adds the cumulative distance between each 1 sec update.
That's only true if the GPS does NOT have satellite lock for the distance you've covered. When the GPS has continuous satellite reception, it adds the cumulative distance between each calculated position -- and most decent units will let you set the "refresh interval" between updates. (The GPSr will, however, "straight line" the distance between updates.)
but what about going up/down in altitude? For example, horizontally you might only travel one mile up a mountain and one mile back down, but the actual total distance might be, say, three or four miles because the crow does not have to fly up and back down. So does the GPS account for the vertical distance?
The GPSr measures distance horizontally AND vertically as long as it has satellite reception. Units with barometric altimeters (atmospheric pressure sensors) do this a bit differently than ones without them -- and, of course, bring with them a whole host of issues with positional errors and interpretation of data.
 

·
Registered
1999 VT1100C2 A. C. E.
Joined
·
9,222 Posts
Some sample info from a recent trip I took in my truck with the GPSr (Garmin GPSMAP 60CS) along follows.

Outbound:
- 254 miles over 5 hours, 1 minute and 43 seconds. The average speed was 51 MPH. The GPSr recorded 2,014 points -- approximately 1 every 9 seconds.
- 65 points were recorded while I was stopped. 38 were recorded with 0 movement while 27 were recorded with false movement (i.e. change in position due to errors in the estimated position rather than movement of the vehicle).
- Altitude was recorded for each of the 2,014 points, ranging from 568 ft above sea level to 1,306 feet. (My GPSr has a barometric sensor so that host of positional error issues comes into play.)

Return (somewhat different route):
- 215 miles over 4 hours, 41 minutes and 4 seconds. The average speed was 46 MPH. The GPSr recorded 1,901 points -- approximately 1 every 9 seconds.
- 38 points were recorded while I was stopped; 24 showing zero movement and 14 showing false movement.
- Altitude varied from 573 ft to 1,330 feet.

If I view my groundtracks in Google Earth, I can find a few points that are not shown on paved surfaces and it looks like I was changing lanes in some curves when I did not do so. Here's an example:

That hard-to-see blue-green line is a section of my groundtrack heading South; after I exited the e-way and turned right, it appears that I was driving in the ditch before turning left. After turning left, it looks like I cut across a few lawns. In fact, I never left the paved surfaces.

But, of course, I cannot determine if the "off-roading" is because the GPSr calculated the wrong position, Google Earth doesn't have it's maps correctly "calibrated", my PC made errors in downloading the GPS data or uploading it to Google Earth or some combination of those.
 

·
Registered
Joined
·
209 Posts
Discussion Starter #15
Wow...


I had some things to do and when I returned I was surprised at the responses! Thanks to all.

I'm still working on the speed function of GPS. Here's my thinking, and I'm sure there's more I need to understand it because these things have been around for awhile and probably have technology that I'm not aware of.

If the GPS has an accuracy of 3 meters, and updates itself every second, the total amount of error possible in that one second is 6 meters. I know that's the worst case, and I only use it to illustrate my point. taking the 6 meters and figuring speed it's possible to be 13 miles/hr off during that 1 second period just due to where the GPS thought it was in comparision to where it really was. If, again the worst case, it measures the next 1 second distance at it's opposite amount of maximum error the speed will go from very high to very low. I'm thinking it can go from 13 MPH high to 13 MPH low in 1 second. That 19 feet doesn't mean anything over a 1000 mile journey, but in the one second where it's calculating speed, there's a big error.


I am no doubt unaware of something that takes all this into account, does anyone know if there is an averaging method?

Again, thanks for all the help!

John
.
 

·
Registered
Joined
·
209 Posts
Discussion Starter #16
adlowe said:
Some sample info from a recent trip I took in my truck with the GPSr (Garmin GPSMAP 60CS) along follows.

Outbound:
- 254 miles over 5 hours, 1 minute and 43 seconds. The average speed was 51 MPH. The GPSr recorded 2,014 points -- approximately 1 every 9 seconds.
- 65 points were recorded while I was stopped. 38 were recorded with 0 movement while 27 were recorded with false movement (i.e. change in position due to errors in the estimated position rather than movement of the vehicle).
- Altitude was recorded for each of the 2,014 points, ranging from 568 ft above sea level to 1,306 feet. (My GPSr has a barometric sensor so that host of positional error issues comes into play.)

Return (somewhat different route):
- 215 miles over 4 hours, 41 minutes and 4 seconds. The average speed was 46 MPH. The GPSr recorded 1,901 points -- approximately 1 every 9 seconds.
- 38 points were recorded while I was stopped; 24 showing zero movement and 14 showing false movement.
- Altitude varied from 573 ft to 1,330 feet.
adlowe,

That's great, I haven't read it all and thought about it yet, but that's hard data. Great! Thanks!

John
 

·
Registered
Joined
·
51 Posts
WOW!!!!!!

This thread has gotten way to technical for me. I just turn the GPS on and go!!!! Tell it to take me back and it does it within a few feet! Punch in an address and it tells me when I get there. Thats good enough for me!

Rick
 

·
Registered
Joined
·
209 Posts
Discussion Starter #18
Re: WOW!!!!!!

West Tn Dawg said:
This thread has gotten way to technical for me. I just turn the GPS on and go!!!! Tell it to take me back and it does it within a few feet! Punch in an address and it tells me when I get there. Thats good enough for me!

Rick
West Tn Dawg,

There's no doubt that GPS can tell you where you are almost or even exactly, I was wondering about using for a speedometer. Can it be accurate enough to keep someone from getting a speeding ticket? Adlowe posted some things I'm chewing over. I didn't know much about GPS but I'm sure finding out now!

John
 

·
Registered
1999 VT1100C2 A. C. E.
Joined
·
9,222 Posts
I am no doubt unaware of something that takes all this into account, does anyone know if there is an averaging method?
Yes, there's averaging. You will not see your speed fluctuating that wildly. The averaging will be determined by the software for the particular GPSr.

BTW, here are the data points recorded that more-or-less correspond to that Google Earth image I posted:
Index.Time.........................Alt...Length...Time..Speed..LegCourse.Position
1935 2/10/2007 12:05:47 PM 887 ft 0.1 mi 0:00:06 65 mph 212° true N40 39.293 W81 26.835
1936 2/10/2007 12:05:53 PM 885 ft 0.1 mi 0:00:06 63 mph 201° true N40 39.212 W81 26.901
1937 2/10/2007 12:05:59 PM 885 ft 0.2 mi 0:00:13 61 mph 197° true N40 39.126 W81 26.943
1938 2/10/2007 12:06:12 PM 884 ft 0.1 mi 0:00:08 58 mph 197° true N40 38.942 W81 27.017
1939 2/10/2007 12:06:20 PM 884 ft 420 ft 0:00:07 41 mph 202° true N40 38.836 W81 27.060
1940 2/10/2007 12:06:27 PM 891 ft 497 ft 0:00:08 42 mph 209° true N40 38.772 W81 27.095
1941 2/10/2007 12:06:35 PM 893 ft 198 ft 0:00:05 27 mph 210° true N40 38.701 W81 27.148
1942 2/10/2007 12:06:40 PM 887 ft 60 ft 0:00:04 10 mph 210° true N40 38.673 W81 27.169
1943 2/10/2007 12:06:44 PM 888 ft 65 ft 0:00:06 7 mph 230° true N40 38.664 W81 27.175
1944 2/10/2007 12:06:50 PM 888 ft 9 ft 0:00:01 6 mph 294° true N40 38.657 W81 27.186
1945 2/10/2007 12:06:51 PM 890 ft 34 ft 0:00:02 12 mph 294° true N40 38.658 W81 27.188
1946 2/10/2007 12:06:53 PM 893 ft 41 ft 0:00:02 14 mph 294° true N40 38.660 W81 27.194
1947 2/10/2007 12:06:55 PM 895 ft 272 ft 0:00:11 17 mph 301° true N40 38.663 W81 27.203
1948 2/10/2007 12:07:06 PM 906 ft 52 ft 0:00:06 6 mph 301° true N40 38.686 W81 27.253
1949 2/10/2007 12:07:12 PM 904 ft 29 ft 0:00:05 4 mph 240° true N40 38.690 W81 27.263
1950 2/10/2007 12:07:17 PM 906 ft 104 ft 0:00:05 14 mph 199° true N40 38.688 W81 27.268
1951 2/10/2007 12:07:22 PM 904 ft 328 ft 0:00:08 28 mph 199° true N40 38.672 W81 27.276
1952 2/10/2007 12:07:30 PM 904 ft 522 ft 0:00:10 36 mph 198° true N40 38.621 W81 27.299
1953 2/10/2007 12:07:40 PM 896 ft 0.2 mi 0:00:15 38 mph 199° true N40 38.539 W81 27.333
1954 2/10/2007 12:07:55 PM 907 ft 0.2 mi 0:00:15 39 mph 200° true N40 38.408 W81 27.394
1955 2/10/2007 12:08:10 PM 910 ft 416 ft 0:00:08 35 mph 203° true N40 38.277 W81 27.458
1956 2/10/2007 12:08:18 PM 918 ft 427 ft 0:00:08 36 mph 207° true N40 38.215 W81 27.493
1957 2/10/2007 12:08:26 PM 910 ft 0.2 mi 0:00:15 41 mph 207° true N40 38.152 W81 27.534
1958 2/10/2007 12:08:41 PM 920 ft 0.2 mi 0:00:17 42 mph 206° true N40 38.019 W81 27.622
1959 2/10/2007 12:08:58 PM 928 ft 0.2 mi 0:00:14 41 mph 209° true N40 37.865 W81 27.720
 

·
Registered
1999 VT1100C2 A. C. E.
Joined
·
9,222 Posts
This thread has gotten way to technical for me. I just turn the GPS on and go!!!! Tell it to take me back and it does it within a few feet! Punch in an address and it tells me when I get there.
But on these cold winter days, chewing on the data can be fun -- especially when reliving a past ride or planning a new one along a similar route.
There's no doubt that GPS can tell you where you are almost or even exactly, I was wondering about using for a speedometer. Can it be accurate enough to keep someone from getting a speeding ticket?
Absolutely -- as long as you've got satellite reception. It can also relieve a lot of pressure that could come from being unsure of where you are, where you're going, how to get there or even how long it's going to take to get there. But only as long as you have satellite reception.
 
1 - 20 of 44 Posts
Top