Technician

Don't really understand the obsession with the Bickerton brand. It seems they do a 20 in wheel folder in green but that bike has completely different frame geometry so no...it's definitely not a Bickerton bike. I'd like to know what brand it is though and add it to the list.

Winfried

On Bromptons, it's the rear triangle that can break (not mine… yet):

downtube

@berlinonaut the video is grainy and unclear. I am not sure anyone will be able to figure out the manufacturer.

However, I think we can use statistics on the known failures. The probability of a Tern failure can be computed from earlier failures & Tern's sales numbers. Tern claimed they sold 100,000 bikes months before the 2014 recall. Several people have made lists of known failures ( I understand the majority of failures were not recalled until AFTER the failure ). I have been in this industry for 20+ years, and I have never seen this type of failure before 2014 ( at the time of Tern's second recall ). Therefore I believe the accepted industry catastrophic failure rate would be less than .00001% ( feel free to use any value you want for this variable....the outcome of the test will not change )

Tern's explanation is based on improper welds from a factory that is no longer used. However two recalls were from two different factories ( according the the CPSC website ), additionally there are outstanding failures of non-recalled bikes. This doesn't add up, hence nothing they share will affect the known information ( statistically speaking ).

Overall with a probability and no explanation we can perform a hypothesis tests. Please see Hypothesis Tests . The results are not good, we can conclude that these failures are statistically significant. Unless all these old Tern's are removed via recall I can statistically prove there will be more failures. I would also expect problems to increase as the weather warms ( since people will ride more ).

Thanks,
Yan

berlinonaut

So if you CAN compute the probability: How big is it then? You are the maths guy of the two of us but I'd assume to get any number you'd first have to know production numbers. You say in spring (?) 2014 it was 100.000 Terns (but deliver no link to proof it). But I'd assume these 100.000 include all Tern bikes, not only the models affected by the recall. Assumed you solved that issue, you then could use the numbers of the known failures to calculate the risk, but then you'll run into two other issues:

- Not all incidents became public and with those that became public if is often unclear when the bike's frame was manufactured (i.e. in opposite to wehen the bike was bought). Therefore your calculations for the risk will possibly too low or inadequately high.

- As the risk is possibly not static it may change over time: Affected bikes may fail early, therefor percentage of bikes in risk of failure from total production may sink over time. Or not - this depends on the root cause of the problem and the factors that make the incident happen (structural damage, mileage, rider-weight, age,...) and we do not know those.
Therefore in my opinion whartever you could compute is a static risk for a given moment in time only (based on incidents that already happened) and still this number is probably totally wrong as you do not have a sufficient basis for the calculation. You are right that there were breaking frames outside the recalls - but were they affected by the same risk-factors? Or have maybe the ones recalled had a higher risk involved?

Therefor I'd be massively interested if you could lay out in more detail how you can compute the risk (as you say you can do it).

Regarding the industry failure rates: Brompton currently does a recall of the F-A-G-delivered -bottom bracket on ~145.000 bikes (while not everyone of these 145k bikes is affected and needs a new bottom bracket). They say the failure rate would be 0,002% (1 in 5000 bikes) - but (what they do not say): that is the failure rate until today, based on historical data, not the risk for failure. And, as they say this is a fatigue-failure, it can be assumed that numbers would rise over time and therefor the risk is higher

Still they say the numbers are within the range of what the industry considers ok in terms of safety standards: So it is probably fair to assume that your assumption of seems to be way too optimistic.

Your last two claims seem a bit trivial to me - it is obvious that failures will happen as long as bikes are on the roads and once you remove them the failures will stop. And also that that more failures will happen when there are more bikes on the road (like when people ride more). However: Risk management is not about total avoidance of risk but to hedge it to an acceptable level. A recall of ALL Tern bikes (no matter of necessary or not) would with a high probablity lead to Tern going bust and therefore a total loss for all customers. Obviousloy not a good solution.

What an acceptable level of failure may be is an interesting question...

Joe Remi

Can, eventually. The chances of a steel Brompton failing suddenly and catastrophically are too low to even discuss. You'll know a crack is forming long before it makes it all the way around to total separation.

downtube

@berlinonaut I can run a hypothesis test. In Tern's forum Josh Hon claimed the bikes are safe due to high sales of 100,000. Statistical testing contradicts his claim. There is no reference because the forum is closed and possibly gone.

We can define x as the true value of a Tern bicycle failing 10 years after purchase with 1000 miles ridden per year. This variable does not change, and we can run hypothesis tests too find a confidence interval for the variable.

We are not talking about frame failures. We are discussing catastrophic frame failures. I had never seen one before 2014. Give me an estimate for this variable and I can run some tests.

Are you encouraging us to allow Tern to not fix this problem....so they remain viable?

Thanks
Yan

berlinonaut

I need no hypothesis test to know that a claim "something is safe because we built 100.000 of it" is bull****. But again you do not deliver any proof for your claim and I also can imagine that you left out the framing context of what Josh Hon said (which, if delivered, would possibly let the citation sound less stupid).

You said you can compute the risk and i ask you to do it and to explain how to do it. Regarding the 100.000 bikes made until early 2014 (assumed the number is correct): How many bikes of those 100.000 until 2014 failed, that have not been part of a recall until today? Because those are the ones that we are talking about. 1? 10? 100? 1000? How many will within the next ten years? As you say you can calculate the risk and also that there would be lists of the incidents you should be able to do it. It is not my job to proof your argumentation - it is your job!

I am clearly not saying that Tern made or makes everything brilliantly right - surely I am not in this boat. Still you seem to be on an agitating crusade against Tern that is only very loosely based on facts. Understandable to a degree having in mind the lawyer-story that you laid out here in the forum - but on the other hand by your behaviour and argumentation in this thread at least with me your credibility has suffered already. Being a natural scientist you should know what a proper rational argument and proof looks like.

downtube

Does anyone have a backup of the forum saved? If so please share the applicable quote here.

Please provide a value for industry wide catastrophic frame failures and I will provide a hypothesis test and confidence interval for Tern's failure rate tomorrow.

I am not worried about anyone's view of my credibility.

Thanks
Yan

downtube

@ berlinonaut I found part of the thread at Tern Link D8 frame shearing in two .

Are you familiar with the thread?

Thanks,
Yan

berlinonaut

I know this collection, yes. But the thread you linked to is 15 pages long. If there's something specific you want to point attention to it would be nice to link directly to the according posting and/or to paraphrase what you want to point out.

downtube

I recall a statement of 100,000 sales from Josh Hon. I see people have quoted that number from him, however I am unable to find his quote.

Thanks
Yan

Technician

Tern set to ship 100,000 units this year | Bicycle Business | BikeBiz


Could be this.

tds101 

First posted failure: Fri Apr 04, 2014 1:15 am


Last post before thread lockdown: Mon Jun 02, 2014 6:24 pm


Thread is 15 pages total,...

downtube

I don't think so. I recall a post referencing UK sales of around 100,000. FYI the problem seemed isolated to the UK at first. Additionally people in the thread quoted the 100,000 number. Very strange.

Thanks
Yan

Technician

That 100,000 figure in relation to Tern folders seems to have been quoted regularly over the last 3 or 4 years by various people involved in cycling. It's just a round number like a hundred grand and rolls nicely off the tongue, in reality I doubt very much if it applies to actual bikes built or distributed anywhere. Brompton claim a figure of 100,000 bikes built and I expect some vehicle manufacturers use the same number for quoted car sales.
Whether true or not who knows, it just sounds impressive.

downtube

Can anyone give me an estimate for a catastrophic frame failure. Maybe .001%? I never saw one before 2014, and I have been active in this industry for over 20 years.

Thanks,
Yan

jur

I am fairly sure a whole bike range was recalled due to frame failures and discontinued. Exact same mechanism. Can't remember the brand.

downtube

Origin 8. This happened around the same time as Tern's issues. I am not sure they are independent events.

I have never heard of anything before this stuff started.

Thanks
Yan

berlinonaut

Huh? Bromptons have frame numbers that count upwards since the beginning (with one jump of 100.000 in the frame-numbers at around 1997). They've reached 500.000 bikes built in the middle of 2017. While Brompton's production numbers have not been published via the news consistently they have been mentioned in various talks by Will Butler Adams and can partly be found in their yearly balance of accounts that is published by law in the uk as well.

In other words: Look at any Brompton with a recent production date (it is coded inside the serial number), subtract 100.000 from the frame number and you know how many have been built until then. Pretty transparent and no rocket science.

The only thing I remember about Brompton and 100.000 bikes is that the yearly capacity of the new factory they recently moved into is 100.000 bikes - double of what the old factory had (and roughly double of what their current yearly production numbers are). They say they want to reach those 100.000 bikes a year in the early 2020ies as a goal.

berlinonaut

I fast-forwarded myself through bigger parts of this thread - quite an interesting (while long) read. There's indeed a post of Josh Hon in there with the 100.000-bikes number. He replied to one of the reports of a broken frame on March 25th 2014:

So he neither said that 100.000 Tern bikes were sold in the uk back in 2014 nor did he say that the frames would would be safe due to the fact that they had 100.000 bikes built (as you claimed earlier in this thread). So you have doing two false claims against Josh Hon in the last couple of postings here alone.

Other than that the thread was pretty interesting. Remember that it was copied over from the official, public Tern-company forums and the posts, lasting from Dec. 2013 to June 2014 were before the recall of the Link-Series in 2014.

There were in between 5 and 8 broken Tern Link frames reported in total in this thread over the course of these seven months, all but one happened with bikes of uk-riders (while the posters in this thread were from all over the world). The one exception happened in Israel.
Plus there were reports of a Tern Verge and and a Tern Eclipse with a broken frame, again from the uk, with one of those two bikes possibly seeming to have missed the recall that had already taken place earlier.

So this seems to be kind of a strong correlation to the uk. The question is why. Assuming that the sales of Tern bikes at this time were not exclusively in the uk but also in the rest of the world this strong correlation of failure with bikes sold and used in the uk must have a reason and I would not assume that this would be that they ride on the wrong side of the road over there. As Tern only entered the market in about mid 2011 the bikes have been quite young when they snapped, one having only been 10 months in service since purchased. At least one of the riders whose frame snapped was an elderly women of low weight, claiming to be anything but a harsh rider. But her bike had something like 2600kms done since purchased and it seemed that the other bikes that snapped had been used relatively intensively as well - maybe beyond average and therefor they were the ones to break first. The broken parts of the frame looked pretty identical with the different bikes, hinting to a bad weld from what one could see.

Tern in form of Josh Hon, Amanda Uy and partly Mark Bickerton (as Tern representative for the uk) said Tern would investigate the issue and by examining the frames trying to find a root cause. This took quite a while and made people upset, the more as communication by via the forum as well as outside the forums with the riders that were harmed was not happening to a level that people understandably would have liked. So they did a pretty bad job in terms of PR and that Evans, the uk Distributor, sometimes behaved more than just embarrassing towards the riders with broken frames made things worse. People from all over the world showed up in the thread, expressing their anxiety and consternation and speculating wildly about the issue as well as about Tern as a brand. As this grew the forum-behaviour was changed - you then had to register and login to be able to see and read the threads. Something that upset people in the threat.

In June 2014 1700 Tern Links were recalled with the explanation given that the frames in question had been produced in the early days of Tern in 2011 and 2012 at a factory they were no longer using.

So from the first post of a broken frame until the recall it took roughly six months. Painful for those affected but in my opinion still within a reasonable timeframe, especially given the fact(s) that the frames hat probably to be shipped to Taiwan for investigation (time consuming), only the first incident was reported in Dec. 2013 (the rest followed later over the course of the following months, second at the end of January 2014, third second half of Februar 2014, forth end of March 2014 end so on), therefor the significance only developed over time. The investigation of the root cause had to take place as had the decision what to do to solve the issue on the one hand and limit losses on the other plus the preparation of the worldwide recall. Basically they had to find out what the bikes in question have in common (correlation), and which of the possible correlations may be the root cause and then which (and which amount) additional bikes may be seriously endangered, following the correlation. Recalling the whole production usually is neither an option nor necessary, on the other hand you want neither to recall too few bikes nor to have to extend an existing recall later.

While during that time Tern consistently assured they were investigating the issue they could no doubt have been far better in terms of reaction times, wording, communications, PR, caring for those that were affected and make their customers generally feel cozy.

So, judging from the tread alone to me Tern's behavior regarding the issues with the Link in 2014 seems reasonable, at least from an engineering perspective regarding the timeframe. The reason given for the issue seemed possible and a plausible cause and - to be honest - I felt much better in regards of my own Bickerton Mk X (that is based on a Tern Link frame) after reading this thread yesterday than before with this thread here at bikeforums.

Unfortunatly there were those other recalls for certain variants of the Eclipse and the Verge in 2013 for some bikes produced between September 2011 and March 2012 (170 bikes if I remember correctly), later expanded for bikes between April and July 2012, 2016 for 220 bikes produced between January 2012 and April 2016. Those were only very small numbers of bikes recalled and while a recall is at the same time a sign for a serious issue with a product or a company as it is for responsible behaviour of a company it still is a disappointment that sums up emotionally to the image of the brand.

Just because the failures may show some similarities in terms of symptoms they still may eventually be based on different root-causes.

Objectively still I would not consider this a total fail of the brand: Many brands have had issues in their early years, including Brompton and Riese-Mueller/Birdy and others, though maybe less dramatical issues and definitively with less public. The fact that Tern scaled production quite big from the start in combination with sourcing the frames from various external factories (as it is common today) plus the communication possibilities via the internet are all things that are different for Tern in comparison to those other companies in their early years.

What is left today are the frames that broke recently that have not been part of a recall. I have no clear view on the numbers and the models - new and old incidents get permanently mixed up as do different models and root causes and so - at least to me - it is currently not possible to get a helpful picture regarding the current situation. The only thing that seems clear is that there have been serious issues aside of the recalls, but I can't say if they are also significant. That said: From a consumer's perspective a single broken frame is too much. From an industry's perspective not necessarily.

Maybe it would be a good idea to create a list of the incidents of bikes that have not been part of any recall and still broke. This should include:

- what did happen
- under which circumstances
- when did it happen
- where did it happen (in which country has the bike been bought)
- when has the bike been bought and what is the frame number and model-year
- what is the exact model
- anything else of interest

Could help us to get a clearer picture. Maybe such a list exists already - would be helpful to know.


Now something completely different: What I recognized to my astonishment when reading to this old thread copied from the Tern forums was that Yan/Downtube was posting in there. Which is in my opinion pretty unusual: A CEO of a bike company posting in a competitor's company-forum. Not only that but he did it under a false flag and fake identity ("bikerrrguy"), at least in the beginning. I would consider both bad style already. As did users of that forum. But it does not stop here. He did a lot of speculation and accusations w/o any proof to a degree that other users in this forum called him a troll - and stil he did not stop.

In April 2014 he claimed to have proof that Tern used a certain factory that would notoriously produce bad quality under a different brand called SGM. Other users said that he had not even the slightest proof and that the SGM brand's quality was far from what he claimed.

He accused Tern and Josh Hon to be a liar and for criminal activities and more, basically bullying Tern in their own forums (as a competitor!): Tern Link D8 frame shearing in two - Page 8

Tern Link D8 frame shearing in two - Page 7

Tern Link D8 frame shearing in two - Page 7

Tern Link D8 frame shearing in two - Page 7

Tern Link D8 frame shearing in two - Page 7

Tern Link D8 frame shearing in two - Page 9

He even falsely accused another user of the forum to be Josh Hon, posting under a fake identity for PR reasons as this user was not willing to share his level of speculation and rage. And he had then to row back quickly as a third user presented a photo of himself and the user in question.

Keep in mind: This is not a disappointed customer full of frustration - this is a competitor using Terns public official company forums that are intended for Tern user support for his personal crusade against Tern as well as to foster his own business, making one false claim after another. Already more than three years ago, long before the actual situation regarding breaking frames had evolved. This leaves me a bit speechless and after reading this I'll take Downtube's posting here at bikeforums with a much bigger grain of salt in future. While this has nothing to do with the Tern issues on an objective level it has well with the rage that went to this thread in this forum here as well. I do not know if Yan has been thrown out of Tern's forums but if he was I would not be surprised....

As a sidenote: Ususally Yan is pretty proud and convinced about the quality of his Downtube bikes. In the thread from the Tern-forums he wrote:

Tern Link D8 frame shearing in two - Page 8

While another user in the same thread wrote later:

Tern Link D8 frame shearing in two - Page 10

I cannot judge on the quality as I do neither know details nor have ever seen or ridden a Downtube and neither do I know if this is a generic issue but these lifespans do not sound too impressive to me and at the time when the issues showed up they should have already been fixed according to what Yan/Downtube claimed. That he as the single owner of a relatively small bike company "doesn't know" if he had any frame issues does sound a bit weird, especially regarding what he expects other bosses of bicycle-companies to know and do (as does the fact that he did not know about the existence of (mandatory) EN/ISO-tests for bicycles and if his own bikes had been tested to them - as it turned out in another thread). Also it would be interesting to compare Yan's (probably wrong) critcisim with what the situation at Downtube may be...

Technician

Interesting post berlinonaut worth reading several times, I'm sure you'll get a good reaction from it
Regarding the 100k figure...It's just a number many have referred to over the years when talking of units sold/exported by many brands of machine and not just bikes. I'm sure Brompton has built and sold many more folders than that. In fact I reckon more than that get stolen in the UK in the course of a year ! A bike thief's dream machine, if you can't ride it away you can wheel it. There's a very good reason why I won't be buying one.

HGR3inOK

+1
Those of us who have been insulted/derided by Yan Lansky in this forum are well-aware of his methods.

-HANK RYAN-
Norman, Oklahoma USA

downtube

I understand I am free to speak as I wish in this country. Was I not allowed to enter the Tern forum? Now lets go through all of your statements:

#1 Quote:
I am certain Tern has not sold anywhere near half of the claimed 100k bikes. (...) Tern has no factory or engineering staff, it is a marketing company. They outsourced frame manufacturing and assembly overseas. The factory they chose has known issues in multiple markets. I am posting a picture of the SG3 Storm sold in the Philippines. There are many of the exact same failures on facebook. Josh knows about these problems ( let's get real I know about them ). He has done nothing.(...)
I do not see a recall ever taking place....I know Josh, I've dealt with him when he was at Dahon. I expect this will turn into a criminal matter. I hope the UK folks are contacting the authorities. I am doing my part in the states.

Several people above just agreed that 100,000 is a nice round number that was not necessarily connected to sales. The SGM Storm failures are almost identical. I think it makes sense that Tern would be aware of other companies frame failures. Also Tern does not have any factories of their own. I was also told that they laid off most of their engineering staff by someone with direct knowledge. After many months of waiting I did believe that Tern would not address the problems.....failures are still occurring. I still believe this will be a criminal matter.

#2 Quote:
I expect this entire forum to be removed from the server shortly. Josh is probably trying to sell the company to his father.

The forum was removed for 6 months, came back and now it has morphed again. The second part of the quote was not necessary or accurate.

#3 Quote:
In short he stole Tern from his father.

Did you read the complaint of Dahon vs Tern? After you read it, tell me what you think. I have a copy I will gladly email it.

#4 Quote:
I can guarantee there are at least 5 more frame failures ( probably over 10 ) that are being hidden from the public.

It would be foolish to believe that all frame failures were publicly posted on Tern's forum. I have learned of many that were posted elsewhere on the internet. I agree that my language was not the best, but the statement is accurate.

#5 Quote:
I assume Josh signs your checks given your attempt to deflect the issue from the faults and on to me. (...) This has criminal intent written all over it moving forward.

I was wrong on this.

#6 Quote:
After this I understood Tern was dead. My wife reiterated do not get involved. Then Thor started posting some nonsense on this forum, so I called him and told him to stop. I shared all information with him. At the end of the conversation I believe he knew Tern was finished. Tern is finished anyway,...

I can not understand how they can be viable long term given the risks. There was just another failure a couple weeks ago, and I recall another about three months back. I would not invest in their business....would you?

#7 Quote:
in 10 years I have had 2 broken frames. Both had spider cracks around the seatpost. I had about a small number of front suspension frames develop similar problems until I changed the seat tube in 2007/2008 and I am not sure I had any breaks since.

This was not well written. I meant that there were two broken frames and some spider cracks around the seatpost. I have had less than 10 faulty frames in 14 years.

#8 Quote:
Tern has no factory or engineering staff, it is a marketing company. They outsourced frame manufacturing and assembly overseas.

This was true when it was written. I believe they have a staff now, but I don't believe they have a factory.

I have reread my posts and I am not proud of lots of the things I said. However, I am happy that I stood up as a voice for unknown victims. I would not feel comfortable keeping quiet.

FYI the thread was much bigger that the content at forumotion.info. My summary of the forum is:
1. Failure rates increasing ( from 1 per month at the beginning to 4 per month at the end)
2. People asking for an update from Tern
3. Hardly any responses from Tern, with no rational for the limited recall. People are still terrified to ride them.
4. At the end there were two crazy users that were supporting Tern as a great company with no logic or rational. Most everyone ignored them, and soon after it was closed to new posting.

I have a copy of the entire forum at home in NC. However I am away teaching until Dec. I will gladly share my backup when I can.

Thanks,
Yan

downtube

Thanks! I left this forum for several years ( around 2010 ). I am getting close to leaving again.

Yan

masch

Well as a Tern user for a year (have not ridden a bike for 20+ years before this other than short city trips less than 1 hour total over 20 years), I was really won over by Tern's design, and reviews. I wanted to replace my Link with a Verge in 2 years.

But my experience with the bike breaking apart while going downhill at an intersection and the support (lack of) from the Tern company, certainly put me off of them. I don't speak for anyone else but just mine and my wife's.