Motor doping thread

[ScienceIsCool]

Re:

Best not use thermal imaging then if tyres are gonnna show up like motors lol? End of the day, when the sun comes out, the road is hot and that is conducting far more than 15w into your tyres. On a cold day without sunshine, the thermal loss will be greater than the gain, so the tyre will never get warm anyway. Just ride a bike in a straight line and use a temp gun on the tyre before and after, they really don't heat up at all.

Uh, no. Solar irradiance is ~1000 Watts/m^2. A bicycle tire's incident area couldn't be more than 0.001 m^2. It's ~ 2 cm wide, is a roughly 1 meter circle (i.e., half is pointed down and away from the sun), and only a small portion is pointing towards the sun. So I'd say on a bright, sunny day the sun is directly warming the tires by less than a Watt.

Also, tires are made of non-conducting compounds, which means that they don't cool down very quickly. Lucky that or your tires would go squishy and half flat if you rode in really cold weather.

Anyways, the whole point is whether you can discriminate the heat signature of a motor (in use) from the background. I'd agree that you'd need to test in all kinds of conditions before you could claim that it's possible. That said, the photos we've seen certainly are suggestive of a motor. My biggest beef is what do you do with these photos? Stop a rider mid-race and disassemble their bike? What if it's just a wrecked bearing? Wait until after the race and a couple of bike changes? Ugh.

John Swanson

 

[sniper]

Great posts John, cheers.
Small correction/addition: a wrecked bearing to my knowledge cannot explain why the heat signature extended into Roglic' seat post. A small battery can.

 

[Tienus]

Interview: UCI presidential candidate Lappartient aims to ramp up fight against hidden motors

https://cyclingtips.com/2017/07/interview-uci-presidential-candidate-lappartient-aims-ramp-fight-hidden-motors/

 

[Catwhoorg]

Re:

Motors exist in bikes. Scaling it is simply a dollar cost issue.

That is probably the single most profound statement in the clinic this year,

 

[sniper]

http://bike.shimano.com/content/sac-bike/en/home/road/wheels---hubs/rear-hubs/fh-9000.html
This presumably was Roglic rear hub, fwiw.

 

[samhocking]

That clearly is distorted by Strade 2 then John. The tyres and rims are the same temperature according to Strade 2's camera for both Roglic and the test subject. i.e. both pictures show tyres and rims the same temperature, yet Roglic shows his wheels and rim the same temperature as the motor, yet the tyres and rims in the picture with a known motor in seat tube shows the tyres and rims the same temperature as the frame. i.e. tyres and rims and frames do not heat up when ridden and as I have tested with a temp gun myself. Clearly the Roglic image is using not only a narrower calibration on the camera, but also using data visualisation techniques with colour range that they don't equally apply to the test subject. Clearly it's distorted by this calibration, resolution, distance from the motor and not knowing the history of Roglics wheel, but fundamentally both tyres are the similar temperature to the rim, but only Roglics tyres and rims are the same temperature as the motor, depsite Strade 2's own test subject having a known motor for comparison. This is the distortion technique used.

 

[sniper]

Not sure why you keep comparing the two. One is rider in race. Other is test subject.
All other riders in race had same heat signature on wheels at same point.
But didn't have the glowing hub plus seat post.
This is only the fourth time im telling you.
Time to take note.

 

[samhocking]

So where is the picture of Roglic's glowing hub and seat stay, but not glowing tyre and rim then?

 

[hrotha]

I imagine people keep comparing the two because that's what Strade 2 wanted us to do, otherwise they wouldn't have put them side to side like that. Anyway, do you have any screenshots of Roglič in the same frame as other riders?

 

[samhocking]

Strade 2 didn't put the two side-by-side, but same Flir camera used, this is the whole point about them distorting the image data to fit their claims.

Core of that documentary was this:

Test subject riding glowing Vivax motor:

Race footagehighlighting a glowing motor in hub

Both are filmed using same Flir camera. Both test subject and race subject with claimed motor show tyres and rims more or less the same temperature as each other, yet the race subject, the tyre and rim are same temperature as claimed motor, but Strade 2 already kindly show us what a real motor and real rim and tyres look like, because they kindly already filmed a test subject for us. What they're looking at as a hub motor, is their own distorted calibration pushing the data visualisation towards what they want to try and show. The reality, is if you set both cameras up the same, from the same distance and use the same calibration surely race subjects motor should look more or less like the test subject where static objects like a rim is not the same temperature as a thermally active objects like an electric motor.

 

[samhocking]

This was the side-by-side shot I think was shown in documentary too. Clearly the hub on right is slightly warmer than on the left, but it's no different than the rim and tyres. Rider on right is a bit nearer camera too, so overall he will be a bit brighter, that's just how the cameras work at distance.

 

[Random Direction]

Sam - honest question - do you have a background in image interpretation, medical imaging, remote sensing, or thermal imaging? I worked in two of these fields for 15 years and can't figure out why you are saying that the images were distorted to fit Strade 2's purposes.

On the comment: "Clearly it's distorted by this calibration, resolution, distance from the motor and not knowing the history of Roglics wheel,"

When doing image interpretation, a scale is always used to display the imagery so differences are visible to the eye, in this case with the hottest areas shown in white, and the darkest in black. Normally a linear stretch is used, with the colours evenly distributed to the range. If some form of distortion were used, then either (a) if an exponential stretch were used the coldest background would blend in with the other cool areas (i.e. most everything save the hottest areas would be black) or (b) there would be jagged steps between colour gradients.

Not sure how thermal (or spatial) resolution call into account the Strade 2 report - these would result in blurriness instead of bias.

Distance to the motor would not change the thermal response - sensitivity to the thermal response would be lower, but not biased in any way unless the background to the image were far hotter (obscuring the response).

As for history of the wheel - it would have been in operation for a while, and therefore heated up. Kinda like F1 drivers warm up their tires before the green.

So, as far as I can tell, each of the four arguments made is false: calibration, resolution, distance and history of the wheel.

Again, let me know your experience in image interpretation.

 

[sniper]

Great post+analysis, Random.
Stade 2 produced a solid piece of investigative journalism. But reality hurts, and for many cycling fans understandably difficult to come to terms with the reality of motors.
Arguments? Who cares about arguments.
It's easier to simply dismiss everything wholesale as "utter balls", "phantoms", and "conspiracies".

 

[Marxten]

I actually made an account because I love the discussion here, but I feel some arguments aren't made yet.

A lot of the previous discussion was focussed on Roglic in the Stade 2 documentary. I remember him not having sponsored wheels (Dura Ace) but a Vittoria, supplied by the neutral service at the moment he was filmed. (Of course, Vittoria Tires still sponsor his team as someone else commented at the time) This makes the question of moto-fraud more difficult, especially since he wasn't really competing for the win at the moment (testing with non-sponsor wheels makes no sense imho, you'd put a lot more people in the know that way).

Also, we saw a quote for how much a motor would cost: 100k to 200k for a working model. Roglic' team, LottoNL, doesn't really have a big budget. They're often estimated around 14 million, but that includes the ice-skating team and on Dutch fora the cycling budget is estimated closer to 10, and even with the new sponsor next year they'll be closer to 12 than 14. Yearly financial data is also available at our chambre of commerce: https://www.kvk.nl/orderstraat/product-kiezen/?kvknummer=301171870000 .
If you want to use a motor, you'd better have a team that either has no publicly available financial data, a team that invests a lot in 'technological advancement' or a team that lets riders use their own (non-sponsor) equipment and staff. (I remember Dennis and Contador using Autobahn discs and Contador having personal mechanics. I also see how Contador's time trials have gone up a level again after getting on the Trek Speed Concept, which is not known as a fast TT bike)

 

[red_flanders]

Re:

...
I worked in two of these fields for 15 years and can't figure out why you are saying that the images were distorted to fit Strade 2's purposes.

On the comment: "Clearly it's distorted by this calibration, resolution, distance from the motor and not knowing the history of Roglics wheel,"

When doing image interpretation, a scale is always used to display the imagery so differences are visible to the eye, in this case with the hottest areas shown in white, and the darkest in black. Normally a linear stretch is used, with the colours evenly distributed to the range. If some form of distortion were used, then either (a) if an exponential stretch were used the coldest background would blend in with the other cool areas (i.e. most everything save the hottest areas would be black) or (b) there would be jagged steps between colour gradients.

Not sure how thermal (or spatial) resolution call into account the Strade 2 report - these would result in blurriness instead of bias.

Distance to the motor would not change the thermal response - sensitivity to the thermal response would be lower, but not biased in any way unless the background to the image were far hotter (obscuring the response).

As for history of the wheel - it would have been in operation for a while, and therefore heated up. Kinda like F1 drivers warm up their tires before the green.

So, as far as I can tell, each of the four arguments made is false: calibration, resolution, distance and history of the wheel.

...

There has been very interesting critique of the Stade 2 documentary in the past. Perhaps this is what Sam is referring to. Worth a read:

viewtopic.php?p=2053665#p2053665

Relevant points to your questions:

3) Dynamic color scale of thermal images
It gets worse. The color scale that the camera provides is not static, it is self-adjusting based on the currently measured temperature range. You can see in the video on the right side how the upper and lower limits fluctuate constantly and considerably. And whenever an very warm object like a motorbike or the front of a car comes into sight, the upper temperature limit shoots up, the colors in the image go crazy and they have to cut quickly. These cases are no problem because the effect is so obviously disruptive that it renders the image completely useless. But the very same effect comes into play all the time in a subtle way that is (especially in combination with the characteristics of the color scale discussed above) extremely damaging!
As a result you can't reliably compare two different thermal images that are shot with a (even slightly) different color scale range. You could see a glowing hub in one shot and hardly any difference in color in an other shot although the temperature differences were exactly the same for both bikes. Just a purely artificial visual effect because in the first shot the temperatures of the two areas happened to correspond to a region in the color scale with a large gradient while in the other shot they didn't.
These guys just took a thermal camera off the shelf and went with its default settings that are not really suitable to answer the kind of questions they were interested in under the given circumstances. With just a little bit of thought they could have customized the settings to produce actually useful material. But they didn't.

 

[jmdirt]

Re:

Great post+analysis, Random.
Stade 2 produced a solid piece of investigative journalism. But reality hurts, and for many cycling fans understandably difficult to come to terms with the reality of motors.
Arguments? Who cares about arguments.
It's easier to simply dismiss everything wholesale as "utter balls", "phantoms", and "conspiracies".

Have you considered that Stade's stuff is not solid? I know that seat post/BB motors exist, and I can't imagine that they haven't been 'tested' in pro races (IMO + YouTube). I don't think that hub motors exist in a form that could be used in a pro race YET (IMO).

 

[Catwhoorg]

My only technological concern (weight of batteries) has been allayed.

One of our control system engineers pointed me to silver-oxide as the current battery of choice for high energy to weight ratio installations.

At 130 Wh/Kg, this comfortably fits in the required energy storage for a game changing difference, into a stack that would comfortably sit in a seat tube just above a motor.

 

[Random Direction]

Re: Re:

...
I worked in two of these fields for 15 years and can't figure out why you are saying that the images were distorted to fit Strade 2's purposes.

On the comment: "Clearly it's distorted by this calibration, resolution, distance from the motor and not knowing the history of Roglics wheel,"

When doing image interpretation, a scale is always used to display the imagery so differences are visible to the eye, in this case with the hottest areas shown in white, and the darkest in black. Normally a linear stretch is used, with the colours evenly distributed to the range. If some form of distortion were used, then either (a) if an exponential stretch were used the coldest background would blend in with the other cool areas (i.e. most everything save the hottest areas would be black) or (b) there would be jagged steps between colour gradients.

Not sure how thermal (or spatial) resolution call into account the Strade 2 report - these would result in blurriness instead of bias.

Distance to the motor would not change the thermal response - sensitivity to the thermal response would be lower, but not biased in any way unless the background to the image were far hotter (obscuring the response).

As for history of the wheel - it would have been in operation for a while, and therefore heated up. Kinda like F1 drivers warm up their tires before the green.

So, as far as I can tell, each of the four arguments made is false: calibration, resolution, distance and history of the wheel.

...

There has been very interesting critique of the Stade 2 documentary in the past. Perhaps this is what Sam is referring to. Worth a read:

viewtopic.php?p=2053665#p2053665

Relevant points to your questions:

3) Dynamic color scale of thermal images
It gets worse. The color scale that the camera provides is not static, it is self-adjusting based on the currently measured temperature range. You can see in the video on the right side how the upper and lower limits fluctuate constantly and considerably. And whenever an very warm object like a motorbike or the front of a car comes into sight, the upper temperature limit shoots up, the colors in the image go crazy and they have to cut quickly. These cases are no problem because the effect is so obviously disruptive that it renders the image completely useless. But the very same effect comes into play all the time in a subtle way that is (especially in combination with the characteristics of the color scale discussed above) extremely damaging!
As a result you can't reliably compare two different thermal images that are shot with a (even slightly) different color scale range. You could see a glowing hub in one shot and hardly any difference in color in an other shot although the temperature differences were exactly the same for both bikes. Just a purely artificial visual effect because in the first shot the temperatures of the two areas happened to correspond to a region in the color scale with a large gradient while in the other shot they didn't.
These guys just took a thermal camera off the shelf and went with its default settings that are not really suitable to answer the kind of questions they were interested in under the given circumstances. With just a little bit of thought they could have customized the settings to produce actually useful material. But they didn't.

Catwoorg. Yes, the dynamic scale of images used in real world applications is an issue for interpreting images, but doesn't make the images themselves useless, in particular for a trained interpreter. One can control for the effect visually (by a good interpreter), understanding how different background effects change the overall image results on the subject of interest, or digitally by splicing the very 'hot' or 'cold' portions of the image off and then stretching /interpolating within the remaining image range. Nothing odd or unique about both of these methods.

I'd have to look at the Stade 2 images in detail to see if this could have been a limiting, however my initial view was that the rim/wheel temperate issue was simply the test subject's bike not being in race / long ride conditions, and the background temperature issue being a non-issue - the ranges were there within the bike, and not obscured by what was outside of the bike.

Sam might of been referring to the dynamic scale issue, however the arguments used didn't hold merit. Scale could still be a problem, just not how it was referenced.

 

[ScienceIsCool]

Re:

My only technological concern (weight of batteries) has been allayed.

One of our control system engineers pointed me to silver-oxide as the current battery of choice for high energy to weight ratio installations.

At 130 Wh/Kg, this comfortably fits in the required energy storage for a game changing difference, into a stack that would comfortably sit in a seat tube just above a motor.

??? - For many many years you could buy high power LiIon cells with excellent energy density. This 18650 (~1.5 x the size of an AA cell) can handle 20 A discharges and has an energy density of 205 Wh/kg. There's puh-lenty to choose from.

http://www.molicel.com/hq/download/DM/DM_INR18650A-V3-80078.pdf

John Swanson

 

[red_flanders]

Re: Re:

Yes, the dynamic scale of images used in real world applications is an issue for interpreting images, but doesn't make the images themselves useless, in particular for a trained interpreter. One can control for the effect visually (by a good interpreter), understanding how different background effects change the overall image results on the subject of interest, or digitally by splicing the very 'hot' or 'cold' portions of the image off and then stretching /interpolating within the remaining image range. Nothing odd or unique about both of these methods.

I'd have to look at the Stade 2 images in detail to see if this could have been a limiting, however my initial view was that the rim/wheel temperate issue was simply the test subject's bike not being in race / long ride conditions, and the background temperature issue being a non-issue - the ranges were there within the bike, and not obscured by what was outside of the bike.

Sam might of been referring to the dynamic scale issue, however the arguments used didn't hold merit. Scale could still be a problem, just not how it was referenced.

I'm not suggesting they are useless, nor is the author of that very good post. The suggestion is that they are quite problematic, and much more difficult to interpret than some here suggest. You will often hear the Stade 2 documentary cited as "proof" of motors in the images in question. It is not, for the reasons outlined.

You should watch in detail. The issue isn't what is or isn't obscured but the exaggeration and fluctuation of very minor differences due to the dynamic scale, colors used, and the small differences being highlighted.

I think you are correct this doesn't address Sam's post. But the Stade 2 documentary, while interesting and suggestive of possible motor doping, is in no way conclusive. The experts they very briefly speak to suggest the same. They clearly avoid going into detail with the imaging experts, for obvious reasons.

 

[sniper]

Stop talking sense, Random Direction!

Also, I still haven't heard a single plausible reason - except a rear hub motor - for why the heat signal extends into the left seat post.
If it were due to some kind of distortion or scaling issue, you'd not expect it to affect only that little piece of left seat post connecting to the rear hub. What would be the odds?
A rear hub motor+battery is the only thing that can plausibly explain it, in my view, and in the view of the independent French thermal imaging expert that was consulted by Stade 2.
(Wait, he could be part of the conspiracy! Maybe he was paid by Stade 2 to provide misleading information! Yes, that's plausible!)

And are we really still discussing the tire temperature?
As mentioned tons of times, all riders passing that same point have roughly the same heat signal on the tires. Only one guy has an additional heat signal coming from his rearhub+seatpost.
The testing subject, as Random points out, has no bearing on the issue whatsoever.

The images, plus Roglic' 2016 Giro D'Italia performance - are you kdding me? - plus his documented bike issues during that same Giro, all point unequivocally towards a rear hub motor. Denying that without providing a plausible alternative scenario, at this point, is just silly.

 

[King Boonen]

Re:

I actually made an account because I love the discussion here, but I feel some arguments aren't made yet.

A lot of the previous discussion was focussed on Roglic in the Stade 2 documentary. I remember him not having sponsored wheels (Dura Ace) but a Vittoria, supplied by the neutral service at the moment he was filmed...

I would be interested to know if that was really the case. Any footage/pictures?

 

[Tienus]

A Belgium article from 2011 :
http://www.dhnet.be/sports/cyclisme/le-vrai-secret-du-velo-de-cancellara-51b78659e4b0de6db97fd696
In short:
The real secret of Cancellara are his bearings which can give him an advantage of 2,5s / km. The brand is called Gold-Race and its an invention of Giovanni Cecchini.
Denis Migani is being interviewed, he is the son of a bike shop owner who used to be the mechanic of Pantani. He was probably also the mechanic of Indurain as there are a pink and yellow jersey from him hanging in the shop.
Denis Migani has been friends with Cancellara who visited the area on vacations with his family when he was a child. Cancellara got introduced to Cecchini by Migani.

The bearing story sounds like nonsense to me and Velonews agrees.
http://www.velonews.com/2011/03/bikes-and-tech/why-the-latest-cancellara-mechanical-doping-story-is-even-more-stupid-than-the-first_164067


Giovanni Cecchini started in the early 90s waxing and repairing skis and snowboards. He was also working with the italian ludge and bobsleigh teams. He was present at the 1998 Nagano winter games.
I am not sure when he first started working in cycling. He is hinting at working with double giro winner Ivan Gotti in a FB post. It is likely he was allready preparing Cancellara's bike in 1998 when he became the junior world tt champion. In 1999 Cecchini prepared the bicycle of Cunego for the world championship in Verona. Cunego won the roadrace and Cancellara the tt again. Not a bad result, Damiano wrote a thank you card for Cecchini.
http://ja.ink361.com/app/users/ig-1168908085/goldrace_official/photos/1144820249954105900_1168908085


http://www.gold-race.com/news/galleria-immagini/?lang=en
In more than 25 years of activity we have won 4 Olympic Games, 24 World Championships and 5 World Record. Giovanni Cecchini has earned the title of Olympic Technician of Italian, Swiss and German team. The rest has yet to be written…

Plenty of fotos on FB.
https://www.facebook.com/pg/GoldRace/photos/?ref=page_internal
You see pictures of him with track, bmx and wheelchair chaimpions and also gran fondo winners.
You can also see him with Cancellara and CSC. It looks like the tt bike sleeps with him in the hotel room.
Pictures of Basso winning the giro 2006. Cecchini takes the tt bike to his hotelroom again.
Checchini was also working on the tt bike of Paolo Savoldelli when he won the giro in 2005. He took it in the back of his car to the Frugeri bike shop. He also brings it to what looks like his garage. There is a box in the garage which has Scarponi and alta montagra (High mountain) written on it. He is also working on the bike in a room with what looks like the giro neutral service bikes and wheels.
Before the 2005 tour he is working on Ullrichs tt bike in his garage and in the Frugeri shop. He is even checking the rear wheel with a stethoscope.
He is also doing maintenance at the track for Kristina Vogel, Maximilian Levy and Rene Enders. This time he uses a stethoscope and 3d glasses to check the wheel. It works wel apparently, plenty of world championships amongst them and Vogel even winning gold at the olympics while working with him.

Cecchini was also present at PR 2007.
http://ja.ink361.com/app/users/ig-1168908085/goldrace_official/photos/693317110921705823_1168908085
I posted about suspiscious bike changes during that race.
viewtopic.php?p=2059900#p2059900



Cecchini does some of the work in the Frugeri cicli bike shop. He also wears shirts with team Cecchini and Frugeri written on it and he has a Frugeri toolbox. Flaviano Frugeri seems to be the other half of team Cecchini. He can be seen in many fotos with Cecchini and he was also present at several GT with him. He was for example working on the bike of Sastre at the vuelta of 2007. Cancellaras tt bike at the 2006 world championship in Salzburg. He was also working at the team CSC bikes during the 2007 tour.

The Frugeri.it website is ofline but there is webarchive.
http://web.archive.org/web/20130929223348/http://frugeri.it/index.php?option=com_content&view=article&id=47&Itemid=54

There is an automatic starting slideshow of interesting pictures.
This part of the website is called Assistenza, sound like a suitable name for it.

Tra i professionisti a cui sono stati eseguiti interventi sulle biciclette, spiccano i nomi di: Fabian Cancellara, Carlos Sastre, Andy Schleck, Ivan Basso, Yaroslav Popovich (nella foto), Jan Ullrich, Paolo Savoldelli, Axel Merckx e tanti altri campioni.

Among the professionals involved in bicycles are the names of: Fabian Cancellara, Carlos Sastre, Andy Schleck, Ivan Basso, Yaroslav Popovich (pictured), Jan Ullrich, Paolo Savoldelli, Axel Merckx and many others.


Since 1998 Flaviano Frugeri has worked for the Italian national team. He also worked four years for the Swiss national team.
https://www.facebook.com/photo.php?fbid=10204527857052771&set=pb.1123452478.-2207520000.1502441298.&type=3&theater

Frugeri works with road racers, track rider and triathletes
.
Flaviano Frugeri with president of the Italian triathlon federation Luigi Bianchi in his shop:
https://www.facebook.com/252485174778482/photos/a.261319333895066.83152.252485174778482/1682541835106135/?type=3&theater

And his dad with the same guy another day in the shop:
https://www.facebook.com/252485174778482/photos/a.261319333895066.83152.252485174778482/1010545625639096/?type=3&theater

His dad also used to work for the Italian national team.

 

[Marxten]

I think it was somewhere in this thread, but who know how much I'd have to read back... Also not sure if there where pictures present.

I did however find the article where LottoNL demands an apology and claims that Roglic had flatted early, got a spare rear wheel and just as he made it back to the peloton he was dropped again (and that's where he was filmed).

Ah, found it: page 2 & 3 of the Roglic topic in the Clinic. http://forum.cyclingnews.imdserve.com/viewtopic.php?f=20&t=30858&sid=8724e5a3c8c308528a93fdecdd0efea3&start=20
Pictures can be found in other articles, but not very good ones.

 

[sniper]

The only source for Roglic riding with a Vittoria wheel is Lotto Jumbo.
No pictures to my knowledge, even though that should be really easy to prove and could have really helped Roglic' case.
No law suit against Stade 2, either.
It would have been a slam dunk if there were proof of Roglic riding with a Vittoria wheel.

I'm confused here, Marxten, as you said you made an account only because you "feel some arguments aren't made yet". But the vittoria wheel claim has been addressed extensively, as you yourself now suddenly seem to have remembered.