Spring 2020 Ebike Reliability Survey Results

More data is always beneficial, something that most scientists look for is solid data.

A more useful approach would be to populate data of 50 or 100 users for each brand and then segment it into different component failure and then start the analysis.
If you would like to gain name and credibility as a source of reliable information, then providing the source of data with statistical analysis in the form of histograms or charts would be helpful.
This provides the readers deeper perspective, not just a number.

It would be wise to populate baseline data points for each brand, no need to rush.


Thanks for the tips Ravi! I’m used to working in much larger datasets in my day job as a mechanical engineer. Due to the limited dataset, the amount of stats possible was limited. I still feel that this data is valuable information for folks interested in ebike reliability.
From my day job, I regularly present complex data sets to folks that aren’t engineers. They need a simple result to understand, at least at first. Some folks, like you and me, would prefer the complete dataset so you could make your own evaluations. Most folks aren’t like that though and I wanted to present a ranking and mileage per failure that would be easily understood.

Perhaps with my next survey, I’ll have enough responses to populate 100 or so for each brand and start the analysis there. Thanks again, with these suggestions, the survey will improve continuously with each iteration.
 

Ravi Kempaiah

Well-Known Member
Thanks for the tips Ravi! I’m used to working in much larger datasets in my day job as a mechanical engineer. Due to the limited dataset, the amount of stats possible was limited. I still feel that this data is valuable information for folks interested in ebike reliability.
From my day job, I regularly present complex data sets to folks that aren’t engineers. They need a simple result to understand, at least at first. Some folks, like you and me, would prefer the complete dataset so you could make your own evaluations. Most folks aren’t like that though and I wanted to present a ranking and mileage per failure that would be easily understood.

Perhaps with my next survey, I’ll have enough responses to populate 100 or so for each brand and start the analysis there. Thanks again, with these suggestions, the survey will improve continuously with each iteration.


It is a very good start for sure.

I will look forward to more analysis from you later this year. Have you considered partnering up with Pedelec Monitor?
If you need introduction, I am happy to connect you to him.


This person has access lots of real world testing data from users all over Europe and some users have over 25,000 miles. That is when true reliability starts to show up.
Kind of like most cars are ok until 50,000 miles. A Lexus and Chevy may fare similar under 50,000 miles. At 150,000 miles, the equation changes.

Even a simple histogram conveys so much information. A picture is worth a thousand words 😉
 
It is a very good start for sure.

I will look forward to more analysis from you later this year. Have you considered partnering up with Pedelec Monitor?
If you need introduction, I am happy to connect you to him.


This person has access lots of real world testing data from users all over Europe and some users have over 25,000 miles. That is when true reliability starts to show up.
Kind of like most cars are ok until 50,000 miles. A Lexus and Chevy may fare similar under 50,000 miles. At 150,000 miles, the equation changes.

Even a simple histogram conveys so much information. A picture is worth a thousand words 😉

I wasn’t aware of that site, I’ll be sure to check it out. Thanks!
 

Sierratim

Well-Known Member
More data is always beneficial, something that most scientists look for is solid data.

A more useful approach would be to populate data of 50 or 100 users for each brand and then segment it into different component failure and then start the analysis.
If you would like to gain name and credibility as a source of reliable information, then providing the source of data with statistical analysis in the form of histograms or charts would be helpful.
This provides the readers deeper perspective, not just a number.

It would be wise to populate baseline data points for each brand, no need to rush.

I humbly want to be the first to volunteer to be an ebike tester under this new and innovative plan! I'll be happy to supply a shipping address for the test bike(s) upon request 🤣
 

Cyklefanatic

Well-Known Member
It often happens that the people who spend more have more complaints, just because they have higher expectations. That happens to me more often than I wish...
I am not so sure about that. Try to get a Mercedes owner to admit that their cars are not as reliable as a corolla. Parts and service are horrendously expensive yet Mercedes owners remain highly satisfied and repeatedly buy again. The very high cost makes it difficult for them to admit faults. They bought an image and they don’t want to tarnish it.
 

Phyz

Active Member
On the other hand, doth hell have any fury like the buyer of a high end model that fails repeatedly and requires significant time in the shop?

On the other other hand, Range Rover owners.
 

Cyklefanatic

Well-Known Member
More data is always beneficial, something that most scientists look for is solid data.

A more useful approach would be to populate data of 50 or 100 users for each brand and then segment it into different component failure and then start the analysis.
If you would like to gain name and credibility as a source of reliable information, then providing the source of data with statistical analysis in the form of histograms or charts would be helpful.
This provides the readers deeper perspective, not just a number.

It would be wise to populate baseline data points for each brand, no need to rush.

Put me on the test riders list. I’ll ride and evaluate for free.🚲😁
 

Ravi Kempaiah

Well-Known Member
E-bike industry has had a horrible history of exaggerating range numbers.

To this day, you will find companies advertising 130 mile range on a bike with 500whr battery. If the customer takes it at face value, they will be disappointed.


1588950457872.png


@EbikeTestLab ,

Here is some advice for your next survey.
There are 4 main component categories on any E-bike. If you address those, rest are all fairly generic.
The first 3 categories make or break the E-bike fun experience.

  1. Motor + controller hardware.

  2. Pedal assist system + display electronics

  3. Battery + BMS

  4. Bicycle stuff - hubs, wheels, brakes, fender etc.
If the first 3 are reliable (let's say...at least for 5000 miles /1 or 2 years depending on their warranty period), then a bike can be considered fairly good irrespective of the brand.

5000 road miles, not mountain bike miles. e-MTB's need much more maintenance just like dirt bikes.
Bicycle component stuff like hubs, brakes, wheels, attachment hardware fall into various quality/price brackets, and they can fail well before 5000 miles but they can be serviced quickly.
There is another significant component that is related customer experience and reliability.
  1. Customer service + warranty response of the company.
Even the best products fail sometimes but how responsive is the company for any issues matters quite a bit.
While this is more of a service category than hardware category, this plays a key role in keeping the bike rideable.
This should also be factored into the equation.

In my mind, these 4+1 category should give a fairly good idea.

All of this is very intensive in terms of time and resources to do it right. Perhaps you could become the testing arm of EBR and get some revenue to conduct these detailed studies but in the end, you will be seen as a reliable source of info that both the manufacturers and customers will seek.
 
E-bike industry has had a horrible history of exaggerating range numbers.

To this day, you will find companies advertising 130 mile range on a bike with 500whr battery. If the customer takes it at face value, they will be disappointed.


View attachment 51864

@EbikeTestLab ,

Here is some advice for your next survey.
There are 4 main component categories on any E-bike. If you address those, rest are all fairly generic.
The first 3 categories make or break the E-bike fun experience.

  1. Motor + controller hardware.

  2. Pedal assist system + display electronics

  3. Battery + BMS

  4. Bicycle stuff - hubs, wheels, brakes, fender etc.
If the first 3 are reliable (let's say...at least for 5000 miles /1 or 2 years depending on their warranty period), then a bike can be considered fairly good irrespective of the brand.

5000 road miles, not mountain bike miles. e-MTB's need much more maintenance just like dirt bikes.
Bicycle component stuff like hubs, brakes, wheels, attachment hardware fall into various quality/price brackets, and they can fail well before 5000 miles but they can be serviced quickly.
There is another significant component that is related customer experience and reliability.
  1. Customer service + warranty response of the company.
Even the best products fail sometimes but how responsive is the company for any issues matters quite a bit.
While this is more of a service category than hardware category, this plays a key role in keeping the bike rideable.
This should also be factored into the equation.

In my mind, these 4+1 category should give a fairly good idea.

All of this is very intensive in terms of time and resources to do it right. Perhaps you could become the testing arm of EBR and get some revenue to conduct these detailed studies but in the end, you will be seen as a reliable source of info that both the manufacturers and customers will seek.

I totally agree on the range numbers, which is one of the main reasons I started EBTL. I just need to figure out a way to test more bikes! Court sent me the Ancheer last year, which was a great partnership. I hope to do more.

Thank you for your suggestions. In regards to the failure categories, I did something very similar with my algorithm where failures were ranked based on severity. My severity rankings line up very closely to yours. It's not a big deal if a brake cutout switch fails, but it's a huge deal if a motor stops working. I also collected primary usage data but didn't factor that in, but will add that to the algorithm next time. If enough responses are generated, the results could be broken out by usage or type of ebike (commuter vs. mountain biker, etc).

Thanks again Ravi, your insight is appreciated!
 

Alex M

Well-Known Member
There is another significant component that is related customer experience and reliability.
  1. Customer service + warranty response of the company.
Even the best products fail sometimes but how responsive is the company for any issues matters quite a bit.
While this is more of a service category than hardware category, this plays a key role in keeping the bike rideable.
This should also be factored into the equation.

In my mind, these 4+1 category should give a fairly good idea.
I agree that #1-#3 should be assigned most weight in ranking. #4 should be assigned much less weight, or perhaps not included in the ranking at all but commented on in footnotes.

"+1" can't be avoided but it's hard to tell how include this in the ranking. Many people don't care about company warranty response as long as there is a response, i.e. reimbursement or parts. Mostly those who buy consumer-direct models. All they care about is getting the part(s), on or out of warranty. Luckily such parts are usually generic. There is sometimes a lament from somebody who bought X or Y consumer-direct bike and is getting a run-around from (very responsive) phone reps who diligently but not very thoughtfully go through the list. The bike might be as reliable as 3 times more expensive model with dealerships, but when consumer doesn't realize that he either must locate a suitable shop or do troubleshooting himself - this is an issue separate from reliability.
 

Cowlitz

Well-Known Member
How many miles do you prefer on a bike in the survey?

I didn't feel like I'd had my Gazelle long enough. I still don't. I bought it last September, rode it through October and maybe part of November, then winter occurred. It only has 800 miles on it. Is that enough?
 

Deacon Blues

Well-Known Member
I think #4 should matter. Little thing that fail can be very frustrating, especially if there is a number of them.

For example, almost every time I power on my Pededgo battery, then press the power button on the display, it won't come on. I have to power off the battery and then power it back on and then power on the display.
Also, when I try zeroing out my tripmeter it doesn't work 9 out of 10 times.

These are little things, and don't disable the bike, but they are a pain in the butt and makes me wonder about Pedego electronics. I suspect that somewhere down the road the display will fail completely.
 
How many miles do you prefer on a bike in the survey?

I didn't feel like I'd had my Gazelle long enough. I still don't. I bought it last September, rode it through October and maybe part of November, then winter occurred. It only has 800 miles on it. Is that enough?

Yes! Due to the algorithm developed, low miles with no failures don’t hurt the overall EBTL miles per failure and it’s another data point to use. Low miles with failures certainly affected the miles per failure calculation.
 

6zfshdb

Well-Known Member
almost every time I power on my Pededgo battery, then press the power button on the display, it won't come on. I have to power off the battery and then power it back on and then power on the display.

All three of my Pedego bikes did the same thing when they were new. Pedego tech support says it is because the battery was in "sleep" mode. It has to be powered on and off and on again or put on the charger to "wake" it. I'll admit, I don't really know what sleep mode is but the problem went away after around 500 miles or so.
 

Alex M

Well-Known Member
I think #4 should matter. Little thing that fail can be very frustrating, especially if there is a number of them.

For example, almost every time I power on my Pededgo battery, then press the power button on the display, it won't come on.
What you're talking about is #3 - as per Ravi. It absolutely should matter, because it's puzzling, annoying and causing delays in usage.

#4 is Bicycle stuff: hubs, wheels, brakes, fender etc. While all this also matters, it sounds more like maintenance, well within an average user skills. Now, what might cause a fender failure, - other than a grenade or an adverse tree trunk , I honestly don't know :).

People new to this might think that ebike is "almost" like an e-car, push the button and go, but it actually needs more care and service than a regular bike and much more frequently than a car. Luckily, bikes are largely user-serviceable.
 

Alex M

Well-Known Member
Rather than bike brand I would be more interested in knowing which motor system is most reliable. If the spokes or fork are faulty I can fix that but I am stuck with that motor system for the life of the bike.
I was about to say the same, but then Ebiketestlab would have to drastically change what they are doing now.

Yes, motor comparisons are useful, but motor isn't everything.
There are also controllers - not just reliability but also adjust-ability - how much you can change whatever can or should be changed.

You are not really stuck with a particular motor for life - not with a hub motor. Installing a hub of a different brand and wattage is not (much) more difficult than replacing a fork. This could necessitate replacement of controller as well, as users are often interested in more speed or power.

And there are also batteries and BMS.

Unfortunately this becomes rather technical, most users are not knowledgeable in anything electrical. Besides, the sheer variety of Chinese batteries, BMS and controllers - both quality and design - make the task difficult.
 

Cyklefanatic

Well-Known Member
I think #4 should matter. Little thing that fail can be very frustrating, especially if there is a number of them.

For example, almost every time I power on my Pededgo battery, then press the power button on the display, it won't come on. I have to power off the battery and then power it back on and then power on the display.
Also, when I try zeroing out my tripmeter it doesn't work 9 out of 10 times.

These are little things, and don't disable the bike, but they are a pain in the butt and makes me wonder about Pedego electronics. I suspect that somewhere down the road the display will fail completely.
My Pedago latch has no powering up issues but the trip meter is useless. I’ve had two of them for 4 years and they are still working fine so the trip meter is sure isn’t affecting anything else.
 

Ebiker33

Well-Known Member
Interesting, spokes failing aren't a big deal any LBS can help you out with those, motors and controllers failing is another story, and battery failure before 1000 cycles isn't good either.
However as somebody that know about batteries, all batteries should be lasting at least 2.5 years. After 1000 cycles consider that bonus time