Odometer and Speedometer Accuracy

Discussion in 'Prime Main Forum (2017-Current)' started by wyobob, Nov 26, 2021.

  1. wyobob

    wyobob New Member

    Joined:
    Nov 26, 2021
    8
    2
    0
    Location:
    Oregon
    Vehicle:
    2022 Prius Prime
    Model:
    XLE
    I have a 2022 Prius Prime with about 6000 miles on it. I am a scientist and engineer, so I know how to collect and analyze data. I wanted to test the trip odometers for distance (miles) accuracy and also to test the indicated speed (mph) accuracy. I have done so.

    Firstly, both tests were done at highway speeds (50-70mph) and within a 600 mile testing window.

    Odometer Accuracy - For 368.9 indicated odometer miles, I recorded the 368.9 odometer miles, 368.7 GPS miles on my smart phone, and 369 miles between interstate highway milepost signs. The odometers are closer to the other two measurement methods than I would expect. Really close - all within 0.1% of each other. This leads me to hypothesize that the odometers are reporting GPS miles, NOT drive wheel rpm's times assumed rolling circumference times elapsed time. The only way to test this would be to disable the GPS receiver(s) or disconnect the receivers from the car's computer. I do not want to do that, so can someone tell me if my hypothesis is correct?

    Speedometer Accuracy - I tested the indicated speed two dozen times in the 200 miles after the odometer test at 50, 55, and 60mph indicated on flat interstate with cruise control. Since I know that the odometers are accurate to well within 1%, the indicated speeds should also have similar accuracy. They do not. I simply stopwatch-timed 1-5mile runs using the indicated trip odometers which show miles to nearest 0.1mi. The 50mph runs cluster around 74sec/mi, the 55mph runs cluster around 67sec/mi, and the 60mph runs cluster around 62sec/mi. Those are 2.4-3.3% higher than the true speed of the car. In no case did a timed mile time come in lower than expected time if the speedometer were accurate to within 1%. I hypothesize that the odometer distance measurement is from a different source than distance used for indicated speed. They differ by about 3%.

    Can someone tell me if my hypotheses reflect ground truth?
     
  2. Tideland Prius

    Tideland Prius Moderator of the North
    Staff Member

    Joined:
    Oct 2, 2004
    42,856
    14,332
    41
    Location:
    Canada
    Vehicle:
    2018 Prius Prime
    Model:
    Technology
    I can’t speak to the odometer but for the speedometer, bylaw, the readout can display faster than indicated but can never display lower than indicated. Part of it is to account for people swapping wheels and increasing their size (and thus changing the speedometer readout accuracy). Sports cars tend to read higher than actual than regular cars since it’s more likely that owners will upgrade their wheel and tire package.
     
    wyobob and alanclarkeau like this.
  3. schja01

    schja01 One of very few in Chicagoland

    Joined:
    Apr 22, 2005
    1,511
    1,016
    0
    Location:
    Chicagoland
    Vehicle:
    2018 Prius
    Model:
    Prime Advanced
    Every Toyota I’ve owned had a speedometer that read 2 mph high. I never swap wheels or tire sizes.
     
    Db17 likes this.
  4. Dennis_MA

    Dennis_MA Member

    Joined:
    Jul 20, 2012
    170
    85
    0
    Location:
    Amherst, MA
    Vehicle:
    2020 Prius Prime
    Model:
    Limited
    Agreed mine is the same. Indicated 37 MPH is actually 35 MPH on my Prius. It is the same for my BMW.
     
    wyobob likes this.
  5. Mendel Leisk

    Mendel Leisk Sand Pounder

    Joined:
    Oct 17, 2010
    46,033
    32,873
    80
    Location:
    Greater Vancouver, British Columbia, Canada
    Vehicle:
    2010 Prius
    Model:
    Touring
    Speedometers read high, by design and legislation, to ensure you're never going faster than the displayed speed.

    The speedometer more'n likely gets it's (accurate) info from the odometer, then adds a percentage when displaying. If you hook up to the car's OBD port, say with a ScanGauge II, you'll see speed value constantly lower than displayed.
     
    #5 Mendel Leisk, Nov 26, 2021
    Last edited: Nov 26, 2021
    wyobob, AldoON and fuzzy1 like this.
  6. Salamander_King

    Salamander_King Senior Member

    Joined:
    Nov 8, 2015
    9,186
    7,313
    0
    Location:
    New England
    Vehicle:
    2021 Prius Prime
    Model:
    Limited
    I thought 2019 and newer PP LE did not come with onboard GPS. I don't think Toyota would use different methods of ODO reading based on the trim level.
     
    wyobob likes this.
  7. fuzzy1

    fuzzy1 Senior Member

    Joined:
    Feb 26, 2009
    15,474
    8,730
    90
    Location:
    Western Washington
    Vehicle:
    2012 Prius
    Model:
    Three
    That was a bad assumption. Built-in "customer expectation management" can alter the accuracy of the displays.

    Get an OBDII-port engine monitor (I have an old ScanGauge-II, but many people use much newer, better, less expensive products) and compare the ECU internally reported speeds to the dashboard display and to your stopwatch or other references. It is expected that you'll find this OBDII-reported speed to be different than the dashboard display, and much closer to your stopwatch.

    Recheck again after the tires wear down a bit. Just beware that tread depth changes don't lead to as much rolling circumference change as you'd expect from the simple circle equations. This is a separate topic.

    As mentioned by earlier respondents, legislative pressures force speedometer displays to bias high, for speed enforcement purposes, removing the "but my speedometer said ..." excuse for most drivers. As best we have heard on the Prius, the display console simply adds a small increment to the number it gets from the internal CANBUS.

    Separately, civil product liability pressures force the odometer to not read high. Very long ago, I received class action lawsuit notices for two of my cars (different brands) accusing them of warranty fraud because odometers were turning too fast. Neither of my cars suffered the alleged defect, both being fairly accurate, but the makers settled by giving 2% mileage extensions on their warranties. Subsequently, I have purchased a newer model of just one of those brands, and its odometer reads about 1.5% low. I don't recall Toyota being similarly sued (I didn't own one at the time), and have measured my Prius reading closer to just 0.2% low.

    In summary, your car can use just a single speed & distance measuring system, yet still have odometer and speedometer reading different, intentionally, by design.
     
    wyobob likes this.
  8. kevinwhite

    kevinwhite Active Member

    Joined:
    Jun 10, 2006
    318
    197
    0
    Location:
    Los Gatos Ca
    Vehicle:
    2017 Prius
    Model:
    Four
    Both the odometer and the speedometer are derived from a mechanical connection to the transaxle.

    They are however calibrated to different requirements.

    In particular, the European requirements for speedometer accuracy are such that it is required that it never indicates less than the actual road speed. The US requirements I could find had a symmetrical error allowance of +/-5mph at 50MPH.

    "The speed indicated shall not be less than the true speed of the vehicle. At the test speeds specified in paragraph 5.2.5 above, there shall be the following relationship between the speed displayed (V1) and the true speed (V2).

    0 ≤ (V1 – V2) ≤ 0,1 V2 + 4 km/h"

    https://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=OJ:L:2010:120:0040:0048:EN:pDF

    On analog speedometers you can often see the calibration limits marked on the dial as small extra marks;

    There are practical limitations with GPS in that it is not available everywhere, eg underground, tunnels or in urban canyons.

    kevin
     
    wyobob likes this.
  9. fuzzy1

    fuzzy1 Senior Member

    Joined:
    Feb 26, 2009
    15,474
    8,730
    90
    Location:
    Western Washington
    Vehicle:
    2012 Prius
    Model:
    Three
    When using highway milepost signs, check many subintervals and adjacent mileposts along the way. I've noticed numerous anomalies that must be accounted for in the calibration.

    Some mileposts are offset from their correct positions, possibly for mounting convenience. Sometimes it seems they just wanted to make another post serve multiple duties, while other times bridgework makes for difficult mounting, so signs get moved off-bridge to an easier spot. These errors are non-accumulating, so matter only when an offending milepost was used as a start or end point.

    There are also occasional discontinuities. These mostly happen when road segments are moved / realigned / shortened, but sometimes with no obvious changes (or the construction changes pre-date my personal history there). Some of these are marked with a special milepost of the form "NNN.XX AHEAD, NNN.YY BEHIND", but most are unmarked or unnoticed. Twisty winding mountain and river highways seem to have many more these than do open Interstates. I-90 in Eastern Washington has a single such discontinuity between the Columbia River and the state line, where about a half mile is "missing" at MP220 at Ritzville. I haven't been diligent enough to adequately check the western half of the highway. But considering the major grade separations across two valleys, each direction should be checked separately.
     
  10. alanclarkeau

    alanclarkeau Senior Member

    Joined:
    Feb 24, 2016
    6,863
    7,409
    0
    Location:
    near Brisbane, Australia
    Vehicle:
    2016 Prius
    Model:
    N/A
    I severely doubt it - as mine hasn't got a GPS.

    I just changed tyres - and my speedometer changed by ½km/hr against my TOMTOM GPS. Same size. But the original tyres, the GPS vs Speedo read slightly differently depending on pressure (I'd bump them up for higher speed driving). And between new and 60,000km, there was a slight difference too.

    Tyre temperature? Road surface?
     
    wyobob likes this.
  11. rjparker

    rjparker Senior Member

    Joined:
    Jun 6, 2008
    4,221
    2,257
    2
    Location:
    Texas Hill Country
    Vehicle:
    2012 Prius v wagon
    Model:
    Three
    Correct. By design. However the ecu has the correct mph which can be displayed with a variety of obd2 connector tools. Here I am scanning the CAN bus accessed at the obd2 connector to provide accurate mph and engine coolant temp. About $40.

    49CBD9DA-6ABD-47F2-BD2E-AAE515914C6A.jpeg
     
    wyobob likes this.
  12. alanclarkeau

    alanclarkeau Senior Member

    Joined:
    Feb 24, 2016
    6,863
    7,409
    0
    Location:
    near Brisbane, Australia
    Vehicle:
    2016 Prius
    Model:
    N/A
    Where does it get the info for the "correct" MPH?
     
  13. rjparker

    rjparker Senior Member

    Joined:
    Jun 6, 2008
    4,221
    2,257
    2
    Location:
    Texas Hill Country
    Vehicle:
    2012 Prius v wagon
    Model:
    Three
    From the transaxle which has precise rotation measurements to facilitate the hybrid control. MG2's resolver.
     
    wyobob and alanclarkeau like this.
  14. AldoON

    AldoON Member

    Joined:
    May 24, 2020
    105
    75
    0
    Location:
    Milton, ON
    Vehicle:
    2018 Prius Prime
    Model:
    Upgrade
    I noticed the same behavior driving a Hyundai accent across the country. Odometer accurately matched highway markers and GPS. Speedometer consistently read higher. Car didn't have a built in GPS.

    Seems to me like while both odometer and speedometer receive data from the same source (wheel rpm) the odometer displays the most accurate number possible while the speedometer displays a higher value...

    Odometer pulling GPS data would be a bad idea. Bad weather, tall buildings, solar flares and alien radio signals amongst other things frequently disrupt GPS signal...
     
  15. alanclarkeau

    alanclarkeau Senior Member

    Joined:
    Feb 24, 2016
    6,863
    7,409
    0
    Location:
    near Brisbane, Australia
    Vehicle:
    2016 Prius
    Model:
    N/A
    Yep - my (TOMTOM) GPS loses signal regularly. Generally when you most need it.
     
  16. alanclarkeau

    alanclarkeau Senior Member

    Joined:
    Feb 24, 2016
    6,863
    7,409
    0
    Location:
    near Brisbane, Australia
    Vehicle:
    2016 Prius
    Model:
    N/A
    So it is in relation to the output speed of the transaxle - not the speed of the car, which could be affected by tyres, pressures and even tyre grip.
     
  17. wyobob

    wyobob New Member

    Joined:
    Nov 26, 2021
    8
    2
    0
    Location:
    Oregon
    Vehicle:
    2022 Prius Prime
    Model:
    XLE
    Thanks, everybody. I take your points. One point of information, however, with regards to the apparent accuracy of my odometer, that I have a hard time with - As I said, my odometer, my phone gps, and mile marker signs were all within less than 0.1% of each other over 369 miles. That is variance of less than 1 part in one thousand. Since time measurement and drive wheel RPM are very accurate, that leaves an assumption of rolling tire radius to calculate odometer miles, if GPS is not used. My drive wheels have a radius of about 11.7 inches, as measured from the wheel axis to the garage floor. In order for the odometer to have the <0.1% variance from other measurements that I observe, the assumed tire radius could only vary by less than about 0.0117 inch (11.6875/1000) from the actual rolling radius. That is close to 1/100th of an inch. The calculation is easy because tire radius and odometer miles are directly proportional to each other. Yes, I have a new car with factory tires properly inflated, but I cannot imagine that the actual rolling radius of the tires is within 1/100th of an inch of the value that the car's computer assumes. As an example of causes of variance between assumed and actual - I had a 600 pound payload in the car during my tests.

    So, is Toyota's assumption of the rolling radius really that good? And, did I just serendipitously pick a course where everything worked out to fool me into thinking my odometer is accurate to within 1 part in 1000? Or my payload squished the tires just the right amount to fool me?

    As for GPS dropping out temporarily due to obstruction - GPS receivers record points at a user-defined time interval. Every GPS receiver I have ever used just draws a straight lines between whatever points it actually records, regardless of missing a few points for a tunnel or cliff or tall trees. Unless miles of points are missed, missing points is not an accuracy issue, especially in my interstate highway tests. The measured difference between mileposts and GPS was only 0.28 mile out of 369 miles, with the GPS miles being slightly less than the mile marker miles, as I would expect.

    As for mile markers moving around with respect to ground truth - only the accuracy of the first and last one matters, relative to each other. The ones in between don't even need to exist.
     
  18. alanclarkeau

    alanclarkeau Senior Member

    Joined:
    Feb 24, 2016
    6,863
    7,409
    0
    Location:
    near Brisbane, Australia
    Vehicle:
    2016 Prius
    Model:
    N/A
    I guess it's finding the mean - without regard to any standard (or non-standard) deviation). As you go over bumps, the tyres would compress and release continually, never the same for any length of time, unless it's a very good road.

    I recall seeing a video of the tyre of a car under full (savage) acceleration, then braking - and it flexed considerably.

    Can't find it - but here's a Dragster wheel:



    It's one of the side-effects of high-profile tyres - the tyre provides a deal of damping from road surfaces - as a low profile tyre doesn't.

    How that relates - I'm not sure - maybe you were lucky? Interesting question though.

    A wild-goose question - if you were to do the 369 miles on a track turning left continually - like NASCAR - I assume the differential effect of the transaxle would sort that out. Or not. But then, you'd be running different pressures left/right too. (I've read about extremely high on RH, crazy low on LH - but not sure if that's true.
     
  19. Mendel Leisk

    Mendel Leisk Sand Pounder

    Joined:
    Oct 17, 2010
    46,033
    32,873
    80
    Location:
    Greater Vancouver, British Columbia, Canada
    Vehicle:
    2010 Prius
    Model:
    Touring
    It's more complicated. Suffice to say there are published revolutions-per-mile stat's for tires. TireRack for one has them. Any 195/65R15 tire (for example) will have roughly the same rev's-per-mile.
     
    #19 Mendel Leisk, Nov 27, 2021
    Last edited: Nov 27, 2021
    alanclarkeau, wyobob and fuzzy1 like this.
  20. rjparker

    rjparker Senior Member

    Joined:
    Jun 6, 2008
    4,221
    2,257
    2
    Location:
    Texas Hill Country
    Vehicle:
    2012 Prius v wagon
    Model:
    Three
    Conspiracy theories have moved from politics to odometers that work too well.
     
    wyobob likes this.
Loading...