Inaccurate speed reading

[user 5157]

Anyone have an problem with inaccurate speed readings on an i3? Mine is a 2015 i3 REX. Speed reading is consistently 2 mph lower than actual speed (street speed monitors). I have heard this is normal for a BMW.

 

[bmwi3bora]

Yes mine is too. Most of the cars are set in this way I believe to give drivers a safety gap before ticketing.

 

[theothertom]

Lower? Most read higher.

 

[user 5157]

Thanks. After having owned cars that are spot on this will take some getting used to especially at speeds 40 and below. I have not checked if higher speed are also lower as well...no tickets .

 

[JohnKelly]

Both of the i3s we've had have run 2mph under what radar stations show.. so speedometer reads 30mph, and the radar says 28mph, for instance.

 

[alohart]

Speed reading is consistently 2 mph lower than actual speed (street speed monitors). I have heard this is normal for a BMW.

Normal would be for the displayed speed to be higher than the actual speed to make it less likely to be ticketed for speeding.

Apparently, the variance of the displayed speed from the actual speed can be set via coding, so it would be possible to change the relationship between the displayed and actual speeds.

 

[theothertom]

Both of the i3s we've had have run 2mph under what radar stations show.. so speedometer reads 30mph, and the radar says 28mph, for instance.

Wouldn't that be higher, or over, what the radar shows?
Car manufacturers are allowed by law to err on the high side by up to 10%. They aren't allowed to err below the actual speed. So the tolerance is +10%, -0% of the actual speed.
I'm on my 5th BMW and they've all read high, as have 2 Hondas and two Toyotas. No manufacturer wants to read lower or they'll be fined.

 

[user 5157]

Both of the i3s we've had have run 2mph under what radar stations show.. so speedometer reads 30mph, and the radar says 28mph, for instance.

Wouldn't that be higher, or over, what the radar shows?
Car manufacturers are allowed by law to err on the high side by up to 10%. They aren't allowed to err below the actual speed. So the tolerance is +10%, -0% of the actual speed.
I'm on my 5th BMW and they've all read high, as have 2 Hondas and two Toyotas. No manufacturer wants to read lower or they'll be fined.

CORRECTION: My i3 reads 2 mph higher than radar. Sorry for the confusion and thanks for all the comments. I have never owned a car (mostly American made) that were not dead on as far as speed.

 

[jadnashuanh]

I've always sort of wondered if that speed reading is also affecting the odometer. It doesn't have to, as both are controlled via the computer. If so, people are being cheated because the car would reach the warranty limit earlier than reality. I guess on a longer trip on a roadway with mile markers, I could check that out. I'll try to remember the next time I go far enough to try to make an estimate, but even changing lanes over a long distance could add up to an error when there are curves involved.

But, while I would prefer the speedometer to be perfectly accurate, as I would a gas gauge that actually was empty when reading empty, that doesn't seem to happen. Both my BMWs read higher than reality. In today's cars with an on-board gps, it would be possible to calibrate the speedometer and odometer to be quite accurate. I had a Citroen once where it was fairly easy to tell...the gauge needle wasn't damped. You could see it jump when you braked, and fall when you accelerated...once that jump was minimal, you knew you were almost out of gas as there wasn't enough to slosh around and affect the float!

FWIW, some of that programmed error may be to account for optional tires. While the optional OEM tires are typically within 2% in rolling circumference of each other, to ensure compliance in the EU, that may account for padding the speedometer as there is no input to tell it what tire is on it, or its current state of wear. Note, if they used the gps and the ABS sensors, they could figure out the rotation rate and diameter fairly easily, and compensate. I don't think anyone does that, but my Garmin bicycle computer does...the tech is not all that hard to execute. On it, though, it uses a rotation sensor verses ABS sensors. The TPMS don't report on every rotation, either, so you can't use them...it seems they report about every 30-seconds once they detect rotation, which is one reason why it takes a moment to get a reading when you first start out, or to reset things if you changed tires.

 

[user 5157]

I've always sort of wondered if that speed reading is also affecting the odometer. It doesn't have to, as both are controlled via the computer. If so, people are being cheated because the car would reach the warranty limit earlier than reality. I guess on a longer trip on a roadway with mile markers, I could check that out. I'll try to remember the next time I go far enough to try to make an estimate, but even changing lanes over a long distance could add up to an error when there are curves involved....

I did do a quick odometer check using mile markers and the odometer readings seemed OK.

 

[jadnashuanh]

I did do a quick odometer check using mile markers and the odometer readings seemed OK.

How far did you drive? To show up an error, you'd have to drive a fair distance. Say it was off as a result of that 2mph difference, you'd have to drive miles to be off by even a tenth on the odometer, depending on your overall speed. I'd be surprised if it was off, but then, I do not know.

 

[user 5157]

I did do a quick odometer check using mile markers and the odometer readings seemed OK.

How far did you drive? To show up an error, you'd have to drive a fair distance. Say it was off as a result of that 2mph difference, you'd have to drive miles to be off by even a tenth on the odometer, depending on your overall speed. I'd be surprised if it was off, but then, I do not know.

About 10 miles.

 

[Libelle]

I check my car's speedo against a Garmin GPS. My current one is 2kph fast while my wife's previous one was absolutely accurate - very rare (not a BMW).

GPS speeds are from integrating successive positions and are accurate. Borrow a Garmin or Navman and compare them with your speedo. You need to do it at several speeds since accuracy varies - both percentage and absolute.

If you really worry about your warranty and odometer accuracy, just make sure you never stay on the outside of corners. Cross the road and drive on the inside of the bend always. That's the best way to get full value - and life stays exciting!

 

[jadnashuanh]

FWIW, GPS ground speed is only accurate when you're on a straight, level road. Most of them do not take elevation change or the chord of a curve into the equation that I've found.

 

[viking79]

In any driving dynamics a car can drive a GPS will have far more accurate speed than anything else. If you need more precise than GPS (fraction of mph), then you need an inertial measurement unit that combines accelerometer data.

GPS speed calculation is not calculated the same as position. It uses the rate of change in the pseudorange to the satellites, essentially doppler. Some receivers might use velocity from track points, but I think most went away from that 10 years ago or more. I had an old garmin that did that and it would jump speed randomly. They don't use this method anymore.

Although the speed might only update once per second it is tracking the satellites continuously so the velocity is always up to date when it updates its calculations.

 

[i3Houston]

Anyone have an problem with inaccurate speed readings on an i3? Mine is a 2015 i3 REX. Speed reading is consistently 2 mph lower than actual speed (street speed monitors). I have heard this is normal for a BMW.

Same with my 2015 i3 Rex, 2mph below street speed monitors.