Mileage discrepancy between odometer and GPS

The number of turns of the wheel is the fundamental data. There is no reason the miles (odometer) would necessarily be derived from MPH. That would introduce errors from two calculations in series putting time in to make MPH then taking back out again to make Miles. It seems unlikely that the MPH calculation would in any way affect the Miles calculation.

 

I don't even use the car speedometer and odometer much. I use a GPS instead. My car reads about 2% faster than the GPS.

 

You guys may find this helpful:
How accurate is a GPS-based speedometer? | GPS Systems

 

Answer:

1) Calibration issues with odometer

2) Even if calibration with odometer is good, very close, it will still vary depending on such as the diameter of the tires. If the odometer depends on a calculated result including the diameter of the tire, per rotation, then as the tires wear, it'll change. So it would likely be set for an interim diameter. So the result will NEVER match the GPS.

3) Even if calibration of the odometer were matched with tire diameter and worked perfectly, it will still ALWAYS vary from your GPS: The odometer rolls, in this hypothetical case, perfectly, all the time...but the GPS does NOT KNOW WHERE YOU ARE ALL THE TIME. Your GPS system probably only locates your vehicle within a few feet---if not that, then a few yards---each time the incoming signal is recalculated. Probably more like a few feet, sometimes, which is very good. Soooooooo........even if you drove in a perfectly straight line, it wouldn't even know exactly your starting and stopping place, but you're also turning, etc., as you drive in your day, and so your GPS thinks you're turning wider or tighter, depending on where it thinks you are...

GPS is VERY good, but it's not exactly accurate.

Odometer is not exactly accurate, either, and tire diameter is not the only variant.

So close is great; exact matches will never happen except, by rare occasion, coincidentally.

Blessings

 

For common automobile tires, diameter doesn't mean much as far as the odometer is concerned. The contact patch is flat, so the familiar circle equations fall apart.

Normal steel belted radial tires have a 'rolling circumference' that is very closely tied to the length of the steel belts beneath the tread. That changes much less than one would expect from computing just the diameter change as the tread thickness wears down.

 

Keep in mind the Prius and most modern cars measure the turning of the wheels, and it's no longer a cable turning - it's an electronic signal. This is used to "drive" the odometer. Then, a correction is applied to increase the displayed speed (to try to keep us from -actually- speeding - and this is by international agreement) and this signal is used to "drive" the speedometer. Pearl S displays about 2% high at highway speeds according to one GPS I use. Further testing is needed to confirm and to extend the comparison to other speeds.

 

Different tire sizes have been the bane of speedometers since the invention of the car, I'm sure, and GPS is only better when it's better.

I haven't heard of car manufacturers using ground-radar for measuring true speed. I'm sure it's been tried, but it would be the most accurate way to measure the actual speed of the vehicle. These have been used on farm machines for decades to calculate and compensate for tire-slippage. They weren't expensive 30 years ago, so it's hard for me to believe they'd be too expensive to use on a mass-produced car.

 

Great information here everyone! I have been searching for information about the Prius mileage on the odo vs actual. I am reading that there are discrepancies and everyone is talking about percentages and "built in" reductions by Toyota. Has anyone tried to figure out exactly the calculation built into the Prius cpu? If Toyota "built in" a variance it would have be based on the actual. What is the actual? If this was able to be figured out maybe we could adjust our own calculations to get more accurate data...

I just put 215/50/17 Yoko Avid Ascends on my 2010 5. I noticed mileage when down as well as mpg (obviously). The car is going less miles on the same amount of gas. I am just trying to figure out what the cpu is basing the mileage on (revs per mile). I could then adjust my mileage and calculate mpg.

A simple formula shows that a 195/65/15 tire that revs 827 per mile will revolve 413500 times in 500 miles. A 215/50/17 that revs 814 per mile will revolve 407000 in 500 miles. The difference is 6500 revs. That equates to 7.985 miles. I could simply adjust my mileage up 8 miles and calculate my corrected mpg. If we knew that the cpu was set on 814 per mile... to me that is the million dollar question. I think we all could benefit with an accurate starting point.

Any thoughts, ideas? Or tell me my math is way off...

 

My favorite calibration method is to compare against the mileposts along a freeway. These posts are placed approximately by the original surveyed road path. Later road realignments can create discontinuities, and some individual posts are misplaced a bit for convenience or obstructions, but these problems can be found and discounted by sampling many posts.

I-90 in Eastern Washington has been good for this. The 160 mile stretch between the Columbia River and the Idaho border has one big bust or discontinuity, a half mile missing near MP218 just west of Ritzville. Some posts are also offset, but otherwise the section is good for odometer calibration.

With OEM tires about half-worn, my first Prius' (2010) odometer read between 0.2 and 0.3% low. My current Prius (2012) with fresher OEM tires was off slightly less. For reasons discussed in other threads, I don't expect significant shifts as tires wear, but haven't yet attempted to check this expectation against actual measurements.

My hiking GPS in battery saver mode produces serious errors, sometimes exceeding 10% on winding roads in narrow valleys. But my automotive unit, and the hiking unit in continuous mode, have landed within 0.1% of the above I-90 mileposts. My methods so far cannot get better resolution than that.

 

I'd think odometer with stock tires should be pretty close, maybe slightly low to err on owner's side, ie: benefit them with warrant issued.

 

Your GPS is not calculating the same distance as your odometer. The surface of the earth is curved (in the "Z" direction) so the road you are on is somewhat longer than the distance your GPS is measuring. Also the GPS isn't 100% contiguous but rather a set of discreet points connected by small straight segments. This error should be small, but it exists.

 

Not so fast! I've got this issue on several of my cars where the speedometer is very close or right on & the odometer is off by 2-4%. I've checked it many times and keep coming up with the same results. I haven't done it on my Prius yet as my wife drives it mostly. Planning a trip up into Canada this summer with it so I'll have a great chance to check it then.

 

The first portion, curvature of the earth, is negligible for our purposes. The second error, the straight segments between discreet points, is larger, but should still be too small to matter. Even that should be swamped by the individual GPS point measurement errors.

They still use the same sensor, but the individual displays have different calibration errors, different offsets from a true reading. For various legal reasons, with OEM tires, the speedometer must not read low, and the odometer must not read high.

 

Speedometer reads speeds higher than both GPS in dash cam and roadside radar signs in my Gen III. It was the same in the Gen II I had before. I know European law insists that cars must never understate the speed for legal protection of the drivers, so an overstate is expected. But ,,,,, the variance seems to be as much as 10% and that's just too damn much!

 

That is why I run slightly larger tires that correct the error.

Bob Wilson

 

I got 15" tires and the speedometer shows 7% lower speed than GPS. I know you said about odometer but if the speed show different the "odo" will show different too.
Europe's regulations I think allow 10% difference.

 

Not necessarily. Each display can be intentionally biased with its own offset or error factor.

In the U.S., the internal computer has a fairly accurate measure of speed, when the OEM tires are installed. However, the speedometer is intentionally offset to read high. From past threads, it appears that Euro market speedometers are also intentionally offset, but typically with a larger offset.

Also in the U.S., odometers typically read fairly accurately, or more recently are intentionally biased to read low. That later practice follows multiple class action lawsuits over warranty fraud, alleging that high reading odometers deprived consumers of some of their warranty protection. Where both of my Prii had/have odometers reading 0.2 to 0.3% low, a newer Forester reads almost 2% low. But that Forester's speedometer still reads slightly high.

With electronics, it is fairly easy to put separate bias or offset or 'error' or 'calibration' factors into each separate measurement.

 

Way back before those class action lawsuits or electronic instrumentation in cars, almost all car speedometers and odometers exaggerated at least a little, and some by a lot (assuming OE tire size). I measured those errors on a few dozen cars then.

I'm glad the lawsuits finally put a stop to the auto industry's long tradition of dishonest odometers. Legally requiring speedometers to overstate speed by more than some reasonable allowance for manufacturing tolerances is ridiculous bureacratic overreach.