Is Bosch off on its speed calculations?

[rudym]

I have a new TREK Super Commuter 8S. It is the speed version motor. Tires and the important stuff is STOCK.
I have noticed over several days and rides using three different instruments, Garmin, Wahoo, Car pacing, that the Bosch Display is Reading 2 - 2.4 MPH faster than actual speed.
Is this a programing issue?

Here is why I am concerned:

#1) corporations tend to cheat - but would they cheat on a bicycle motor?
#2) I am paying for a 28 mph Pedelec and if it is cutting out at 26 MPH, I am not getting what I paid for.

Has anyone else experienced this?
I am going to have the bike shop give it a go and see what they think and or discover but right now something seems fishy.
I edited this initial post to remove unnecessary/controversial tone and verbiage.

 

[Carterk]

I’m pretty sure my 2018 Raleigh Redux (Brose S) is reporting speeds 1-2 mph faster than actual. It is also completely stock.

 

[Over50]

When using my Garmin Virb with my R&M Charger Nuvinci I generally had GPS readings about 1 mph higher vs the Bosch Intuvia display. Precision was hard to obtain since it wasn't a live side-by-side reading. Rather after filming I have to match the GPS files to the video files so there can be a second or two differential in the timing of the files (my video captured the Intuvia display). But on multiple trials I came to the conclusion that the Intuvia was slightly optimistic on its speed readings. I might try a similar comparison this weekend using my GPS watch and one of my Bosch Purion display bikes.

 

[Mr. Coffee]

Why your GPS watch is probably overestimating your mileage

Speed is a function of distance and time. To more accurately characterize the problem I'd suggest seeing if there is an odometer error over known routes. You should easily be within 1-2 percent on a ten mile route.

I'm quite reasonably assuming that the Intuvia's clock is accurate. Any clock error big enough to produce a ten percent error in speed would be easily noticeable through a quick inspection.

I'd also suggest checking the average speed figures versus time over a known route (verify the length of your route through Google maps).

It is quite possible that there is a calibration issue with your Bosch motor. If the wheel size were entered incorrectly at the factory or the bike shop it might explain what you are seeing.

 

[rich c]

A couple thoughts, you aren't dealing with military grade GPS triangulation with your satellite apps. Also not sure what accuracy of tire diameter can be entered into the Bosch system. A couple of my bikes have pretty crude tire diameter options. On my one fat bike, I had to enter 29" tire to get closer to the circumference than entering the 26" tire designation that is on the tire. On my tad pole trike, I can only enter 20" and I know my fat 20" is much larger diameter than a stock 20". I don't like big corporations any more than the next liberal, but tying a slight speed discrepancy to the fuel system software issue with VW and the Bosch fuel system feels like a huge stretch. Also suggesting you aren't getting what you paid for, well that will be between you, your lawyer, and the bike company. You didn't buy a Rolex, why expect or demand such extreme precision with a bicycle? After all it is supposed to be for pleasure.

 

[rudym]

A couple thoughts, you aren't dealing with military grade GPS triangulation with your satellite apps. Also not sure what accuracy of tire diameter can be entered into the Bosch system. A couple of my bikes have pretty crude tire diameter options. On my one fat bike, I had to enter 29" tire to get closer to the circumference than entering the 26" tire designation that is on the tire. On my tad pole trike, I can only enter 20" and I know my fat 20" is much larger diameter than a stock 20". I don't like big corporations any more than the next liberal, but tying a slight speed discrepancy to the fuel system software issue with VW and the Bosch fuel system feels like a huge stretch. Also suggesting you aren't getting what you paid for, well that will be between you, your lawyer, and the bike company. You didn't buy a Rolex, why expect or demand such extreme precision with a bicycle? After all it is supposed to be for pleasure.

Well the 3 comparisons were side by side as in immediate and visual and included a pace vehicle and all confirmed the same thing - that the Bosch Intuvia was reading 2 to 2.4 miles per hour faster, this meant the motor was cutting out at 26 mph vice 28 @Carterk, since we have another stock bike showing similar optimistic Intuvia data I am doubly suspicious that something is amiss. Hey folks, more data please. I am suspicious as my scientific nature demands it. Doubt data, redo and or repeat the tests, see if the same results are achieved. Scientific process of sorts. The best method of testing this would be for lots of peoples who own the Bosch Cat 3 Pedelec motors testing their Bosch Intuvia readouts and sharing them.

 

[rich c]

According to this post, you can adjust the tire circumference +/- 5%.
https://electricbikereview.com/forums/threads/intuvia-wheel-circumference-issue.3511/
Maybe do more research before getting "doubly suspicious"

 

[rudym]

A couple thoughts, you aren't dealing with military grade GPS triangulation with your satellite apps. Also not sure what accuracy of tire diameter can be entered into the Bosch system. A couple of my bikes have pretty crude tire diameter options. On my one fat bike, I had to enter 29" tire to get closer to the circumference than entering the 26" tire designation that is on the tire. On my tad pole trike, I can only enter 20" and I know my fat 20" is much larger diameter than a stock 20". I don't like big corporations any more than the next liberal, but tying a slight speed discrepancy to the fuel system software issue with VW and the Bosch fuel system feels like a huge stretch. Also suggesting you aren't getting what you paid for, well that will be between you, your lawyer, and the bike company. You didn't buy a Rolex, why expect or demand such extreme precision with a bicycle? After all it is supposed to be for pleasure.

@rich c - at issue here is approximately 2 mph which is a lot of difference. yes of course the VW analogy was extreme but it was just an analogy of corporate fiddling. Does the consumer not have a right to expect performance to be on par with advertisement..etc…? At any rate, it will be interesting to accumulate more and more data in this regards... a @Bosch or @Trek rep is always welcome to explain the difference since to me the overly optimistic speed reading favors the manufacturer in several plausible ways. Don't get me wrong, I certainly like doing 26 miles per hour BUT.

 

[rich c]

@rich c - at issue here is approximately 2 mph which is a lot of difference. yes of course the VW analogy was extreme but it was just an analogy of corporate fiddling. Does the consumer not have a right to expect performance to be on par with advertisement..etc…? At any rate, it will be interesting to accumulate more and more data in this regards... a @Bosch or @Trek rep is always welcome to explain the difference since to me the overly optimistic speed reading favors the manufacturer in several plausible ways. Don't get me wrong, I certainly like doing 26 miles per hour BUT.

I take it you have not contacted anyone but the forum about this issue? Just right to the conspiracy theory and their failure to meet your satisfaction? In that last link I posted, Chris Nolte is a respected dealer in New York. If he knows about adjusting the software, your dealer should as well.

 

[rudym]

Why your GPS watch is probably overestimating your mileage

Speed is a function of distance and time. To more accurately characterize the problem I'd suggest seeing if there is an odometer error over known routes. You should easily be within 1-2 percent on a ten mile route.

I'm quite reasonably assuming that the Intuvia's clock is accurate. Any clock error big enough to produce a ten percent error in speed would be easily noticeable through a quick inspection.

I'd also suggest checking the average speed figures versus time over a known route (verify the length of your route through Google maps).

It is quite possible that there is a calibration issue with your Bosch motor. If the wheel size were entered incorrectly at the factory or the bike shop it might explain what you are seeing.

@Mr. Coffee I also have a very bored policeman friend who would love to radar me just because he is a bike nerd also.

 

[Mr. Coffee]

This thread got me curious this morning so for the heck of it I ran some experiments on a fairly straight and level stretch of road near my home.

What I found was:

The Bosch odometer (Intuvia) was very close to my car's odometer which was very close to what Google maps estimated. All three were easily within two percent of each other and since none of them were consistently "high" or "low" I doubt there is any kind of systematic error.

At steady speeds from 9 to 11 mph, the Bosch was very accurate, easily less than three percent and probably more like one percent.

Both a Garmin Oregon GPS and a Suunto Ambit GPS watch gave speed and distance figures that were more than 5 percent high.

This all was done with an R&M Charger GX with a Bosch Performance CX.

I did a few runs with my Pedego Interceptor and found that its odometer was also very close to the other three measurements.

 

[rudym]

I take it you have not contacted anyone but the forum about this issue? Just right to the conspiracy theory and their failure to meet your satisfaction? In that last link I posted, Chris Nolte is a respected dealer in New York. If he knows about adjusting the software, your dealer should as well.

@rich c - you are correct, having just got my new SC8S I am just now exploring it. No intent to be confrontational, just want to gather data from lots of folks with similar Bosch systems. Curiosity. I am aware that adjusting software to defeat the 28 mph cut out in anyway would void an important warranty - I have no desire to do that or encourage that. Incidentally I edited the original post to make it less objectionable. At 5 a.m. I can be testy.

 

[rudym]

This thread got me curious this morning so for the heck of it I ran some experiments on a fairly straight and level stretch of road near my home.

What I found was:

The Bosch odometer (Intuvia) was very close to my car's odometer which was very close to what Google maps estimated. All three were easily within two percent of each other and since none of them were consistently "high" or "low" I doubt there is any kind of systematic error.

At steady speeds from 9 to 11 mph, the Bosch was very accurate, easily less than three percent and probably more like one percent.

Both a Garmin Oregon GPS and a Suunto Ambit GPS watch gave speed and distance figures that were more than 5 percent high.

This all was done with an R&M Charger GX with a Bosch Performance CX.

I did a few runs with my Pedego Interceptor and found that its odometer was also very close to the other three measurements.

@Mr. Coffee - Thanks! Exactly what I am looking for; but percent of what? Percent(s) can be great or little so what did the difference translate to in MPH or KPH? Thanks in Advance! Looking forward to more info.

 

[rich c]

@rich c - you are correct, having just got my new SC8S I am just now exploring it. No intent to be confrontational, just want to gather data from lots of folks with similar Bosch systems. Curiosity. I am aware that adjusting software to defeat the 28 mph cut out in anyway would void an important warranty - I have no desire to do that or encourage that. Incidentally I edited the original post to make it less objectionable. At 5 a.m. I can be testy.

I'm talking a certified Bosch mechanic doing the adjustment. They can plug in a laptop at the shop and do it. Along with any upgrades or diagnostics.

 

[Mr. Coffee]

@Mr. Coffee - Thanks! Exactly what I am looking for; but percent of what? Percent(s) can be great or little so what did the difference translate to in MPH or KPH? Thanks in Advance! Looking forward to more info.

I'll give examples of what I measured and observed.

From my garage door to the nearest four-way stop Google Maps estimates 8.9 miles, the odometer in my Lexus gives 9.0 miles, and the odometer on my Bosch Intuvia measures about 9.05 miles. I'd say all three of those measurements are easily within two percent of each other.

For my "test course" I had a straight and level section of road that measured 2.1 miles on Google and my Lexus and 2.05 miles on the Bosch.

I attempted to ride that course at a steady ten miles per hour and timed it at 12:30. If I really was riding ten mph the whole way and if the Bosch was perfectly accurate I should have ridden that distance in 12:18. A 12-second error on a 738-second measurement is about a 1.5 percent error. After I did this experiment I discovered how to reset the "Average Speed" value on the Intuvia independently and could have produced even better test results if I felt like going out in the heat and doing this again.

During that same section the Ambit consistently clocked my speed at about 10.5 mph and the Oregon was about the same, so they both were roughly five percent "high".

So based on my experiments on a fairly short straight and level course I'd say an indicated "10 mph" on the Intuvia is at least 9.85 mph and at most 10.15 mph. I wouldn't lean too hard on that last digit but that gives you a feel for the possible error range. So at 25mph that would extrapolate to an "actual speed" of between 24.6mph and 25.4mph (after rounding). I would guess that a GPS might give a speed estimate of around 26.3mph when the Bosch estimated 25mph.

The major upshot of my observations is I can't get a 2.5mph error out of the Bosch Performance CX and Intuvia. That tells me it is unlikely there is a systematic or quality problem and your problem is most likely a calibration or setup problem, most probably the wheel size. If you could accurately characterize the error values you are seeing (e.g. by mapping out a known course with Google maps and riding it and comparing your odometer figures with what Google maps estimated for the distance) you might even be able to figure out what the wheel size setting in your Bosch currently is.

 

[AlanDB]

I posted this same concern a while back: https://electricbikereview.com/forums/threads/intuvia-speedometer.22566/

I have since done more extensive testing using different GPS devices and I am more convinced that the Intuvia speedometer registers 1 to 2 MPH faster than actual speed in the 10 to 20 MPH range. I have carefully measured the tire circumference and the setting is correct.

 

[Carterk]

I posted above that my Raleigh Redux seems to be similarly optimistic about its speed- I just want to emphasize that this is a BROSE motor system, not Bosch, so no Intuvia, no collusion ;-)

 

[Mr. Coffee]

Can anyone on this thread tell me what technical basis they have for believing that speed estimates from a GPS, especially a handheld GPS or GPS watch, are likely to be more accurate than other measurements of speed?

 

[rudym]

Can anyone on this thread tell me what technical basis they have for believing that speed estimates from a GPS, especially a handheld GPS or GPS watch, are likely to be more accurate than other measurements of speed?

@Mr. Coffee - I don't believe gps in general are more accurate - in my case they were my first observation that something was different.

The Honda CRV pacing me at 20 mph and then 25 mph and then 28 mph has a very accurate speedometer and I was surprised that the cars speed was the same as the Garmin and the Wahoo GPS units.

Radar tomorrow - I hope? Lets see what my local law enforcement buddy thinks about me doing 28 mph in a 25 mph zone. If I am a good guesser based on my 3 observable speed comparisons, he will clock me at 25.8 mph or there about and want to issue me a citation for doing 8/10ths of a mile over the posted speed limit. But at the time he clocks me I will have the bike at 28 mph according to the Bosch Intuvia.

All this may or may not amount to a hill of beans BUT, it will sure be fun trying to figure out where the error is coming from.

Equipment or algorithm ? Mechanical or math? perhaps they are the same.

 

[AlanDB]

I have tested my Garmin Oregon GPS speed in the following way in my car on open highway, flat terrain, over a 10 mile stretch using a stopwatch. and mile markers. I set the GPS to record track points at the lowest setting, every 1 second. Set the cruise control on my car as close to 60 MPH as possible (using the speed display on the GPS ... my car also exaggerates the speed on the speedometer). Then I use the stopwatch to mark the start and stop point as I pass the mile markers. Then it is easy to calculate the average speed for 10 miles using the time recorded on the stopwatch. I then take the track log file from the GPS and truncate the points outside the range as I can best determine them based on the UTC values recorded on track log file (my stopwatch is a Casio Wave Ceptor that syncs time once a day, and of course the GPS gets its time from the GPS satellites). In my test, the GPS track log showed the distance at 10.02 miles and the speed at 60.3 MPH. My calculation using the 10 mile distance and stopwatch showed my speed as 60.46 MPH. Incidentally, my Hyundai speedometer showed the speed between 62 MPH or 63 MPH on this test.

I realize this is not a scientific test and there are many variables that can introduce errors. But it is about as good of a test as I can do with the equipment I have at hand. Also, a car going 60 MPH with cruise control is not the same as bike riding where the speed is much slower (avg. 11 MPH in my case) and varies much more dramatically, and bike speed can change by a significant percentage in one second. But still, what I have observed is pretty consistent. The Intuvia display always shows a higher speed than the GPS ... anywhere from a few tenths MPH up to 2 or more MPH.

Another point is that speed inaccuracies caused by satellite drift position error on a moving GPS would tend to increase the average speed calculation on a GPS, in that it would increase the distance traveled in a given time (on a recorded jagged line rather than a strait line.)