iBike Pro - New PM

RapDaddyo said:

About as accurate as the HAC. Uses gradient/speed/airspeed.

Search this list:
http://lists.topica.com/lists/wattage/read?sort=d&start=37910

Woofer said:

About as accurate as the HAC. Uses gradient/speed/airspeed.

Search this list:
lists.topica.comreadOpen ↗

FWIW, I sent an email to the CEO of the company and he responded. They are aware of the limitations of the HAC4 and believe that their device takes a completely different technical approach and that their device is impressively accurate. He said that the Cooper Institute in Dallas is conducting a set of tests of their device against PT (and perhaps other PMs). He said the test results should be posted on their web site at the end of this month. They consider their measurement sensors and algorithms to be proprietary, so it'll remain a black box for the moment. I look forward to seeing the Cooper Institute test results. I will wait to pass judgment on the device after I see some hard data and some ride reports. You may be right that it is no more accurate than the HAC4, but I'm willing to wait and see some data.

Darn it, I should have taken out a patent:

http://www.cyclingforums.com/showthread.php?p=1270367#post1270367

rob of the og said:

Darn it, I should have taken out a patent:

cyclingforums.comshowthread.phpOpen ↗

Yep, you should have. Acccording to the company's CEO, they have attempted to solve the equation in Kraig Willett's piece biketechreview.compm review.htmOpen ↗. In his words, "... we are using innovative sensor technologies and proprietary algorithms to determine ALL of the forces of his [Willett's] power equation." Whether they have done that and how well their solution works remains to be seen, but that's what they say they were trying to do.

RapDaddyo said:

A new PM product was launched at the Interbike show -- the iBike Pro. These people are taking a completely different technical approach to measuring power, and their price point alone (MSRP$349) makes this product very interesting if it measures power with any reasonable accuracy vis-a-vis PT and SRM. I don't know anything more than what is on their web site, but would be very interested in any commentary on the accuracy of this product. I plan to contact the company and ask for any comparative data they have vs. the PT. If that looks promising, I have a buddy who's about to buy a new bike. He can't afford the PT and is considering the new Garmin bike computer. I may redirect him toward this product if it looks to be at all accurate in measuring power. If he gets one, I'll strap it on my bike and compare the data with my PT SL on a few rides. I'll report the results. If you get one, how about posting a user review here? Here's their website ibikesports.comdefault.aspxOpen ↗. BTW, I have no relationship with this company and no financial interest in this product's success or failure.

So, the big improvement over the HAC4 here is the wind sensor. At first you think, "wait, that can't work; what happens when the wind changes!", but then you realize that it all pretty much works out; if you turn into a 10mph headwind, well, the aerodynamic component of the resistance you're overcoming just got 10mph harder, so it will get it basically right.

It's a clever idea. According to the discussion taking place on the topica list, you need to calibrate the unit with a "coast-down" from approximately 20mph to 5mph. It also sounds as though there's no way for this unit to take acccount changes in position or rolling resistance. Position can vary would change the appparent power quite a lot between the tops and the drops. It would also be weird to roll over some bad pavement and see your "power" drop.

Still, this sounds like a quantum leap over the HAC4, because it will at least give vaguely meaningful results on the flats. It will be interesting to see people's training results with this system; if it's good enough to be an effective power training tool, I'd be more than a little tempted to save the ~$700 and all the finicky little peeves one hears about with respect to the PowerTap. I've always said that a power measurement device "for the masses" could be incredibly popular.

To rip off an interesting comment from the topica discussion, when coupled with a strain-gauge based PM, this could be a really neat tool for measuring aerodynamics. Download both files, overlay the graphs from the two meters. The greater the ratio of iBike power to SRM power, the better the aerodynamics.

Awful. Obviously power varies a large amount depending if your sitting bolt upright on the tops or on aerobars with your nose down to the bars.

The polar power seems better to me.

-Bikeguy

kmavm said:

According to the discussion taking place on the topica list, you need to calibrate the unit with a "coast-down" from approximately 20mph to 5mph. It also sounds as though there's no way for this unit to take acccount changes in position or rolling resistance. Position can vary would change the appparent power quite a lot between the tops and the drops. It would also be weird to roll over some bad pavement and see your "power" drop.

iBike seems to have added a much larger FAQ section to its website. Here is what they say about the riding position issue:

"Most riders stay in the same position about 90%+ of the time, so the percentage of the total ride where there is a wattage difference won't be great. The iBike Pro assumes that the cyclist remains in the same riding position, so changes in riding position will not be reflected by changes in the iBike's wattage readings.

However, remember that the overall wattage "error" caused by ride position is moderated by other wattage components. For example, let's suppose that, at a particular point of the ride, the iBike is reporting 200W. Remember that the iBike "knows" where the watts are going. Let's suppose the breakdown is: 75W due to hill climb, 75W due to aerodynamic resistance, 30W due to acceleration, and 20W due to frictional forces. Now, finally, let's suppose that the rider is out of his normal riding position and that this is causing a 10% underreporting of aero watts.

Then the iBike should be reporting 208W, not 200W. However, the total watts error at this particular instant of the ride would be 208/200 = 4%, not 10%! Think of it this way: 10% of the time the iBike might be off by 4%. Over the course of the entire ride this is 0.4%, which is insignificant. In fact, our accuracy is well within the range of competitive products."

What I find interesting is that while competitive products freely admit they have accuracy tolerances (for example SRM's amateur product ($2,049) claims accuracy of +/-5%, and their pro product ($2699) +/-2%) the cyclist has NO IDEA when his $2000 SRM is over or underreporting watts. At least the iBike guys are telling us when there might be an issue! To me, that's a level of honesty you can't get elsewhere.

Also, I saw a link somewhere where they talked about the road surface issue. They said they have a way to handle it.

Does the iBike support playlists?

kyzyl2 said:

To me, that's a level of honesty you can't get elsewhere.

Before you beat the drum too loudly about honesty, I hope you can see that the aero example you chose is the least relevant to the rider and the most favorable to iBike's marketing staff. Fortunately, that's not the example they use in their FAQ (or else they changed it after you looked because it made no sense).

Personally, I'd be interested to see how well it deals with the turbulence created by being positioned directly behind another cyclist. If it's mounted on the handlebars then it's not seeing the wind that's hitting the riders head and shoulders. If it handles real-world situations as well as or better than existing products then that's great. I'll wait until the comparison data comes out before starting the cheerleading or trying to shoot it down.

frenchyge said:

Before you beat the drum too loudly about honesty, I hope you can see that the aero example you chose is the least relevant to the rider and the most favorable to iBike's marketing staff. Fortunately, that's not the example they use in their FAQ (or else they changed it after you looked because it made no sense).

Personally, I'd be interested to see how well it deals with the turbulence created by being positioned directly behind another cyclist. If it's mounted on the handlebars then it's not seeing the wind that's hitting the riders head and shoulders. If it handles real-world situations as well as or better than existing products then that's great. I'll wait until the comparison data comes out before starting the cheerleading or trying to shoot it down.

Hey...I pulled that quote off their website FAQ this morning! I don't know what FAQ you're talking about, from what website or forum and at what point in time.

Their website says that the iBike's coast down procedure properly accounts for the effects of riding position including head and shoulders. They also say that their pressure sensor properly accounts for the relatively lesser amount of opposing aerodynamic pressure that exists for riders in the pack.

Their FAQ about riding position freely admits that, when that riding position changes significantly (they estimate the rider is out of his primary riding position about 10% of a long ride: do you agree?), there will be an "error" in the wind component ONLY of power measurement. The example also shows that the overall effect wind-related error is amelioriated if the cyclist is expending power for other purposes (which is almost always the case). And they conclude that the overall impact of riding position changes on measurement accuracy is well within the limits of existing power meters.

Perhaps the more fundamental point here is that ALL power measurement devices make approximations and ALL make power measurement errors. It looks like to me that the iBike guys are trying to be straightforward about where they make their approximations, and what the accuracy impact will be.

Doesn't bother you just a little bit that power meters costing 5 times as much as the iBike have measurement errors as much as +/- 5% (by the manufacturer's own admission), and that as a cyclist you have NO IDEA when, where, or why these really expensive meters are giving you erroneous readings?

All this said, I agree with you that the "proof is in the pudding"; we shall see what the iBike comparison tests show.

Even before they post their test and comparison data for this new PM product, it's interesting to consider what the relevant comparison should be. I think it does this product a disservice to compare it with the PT and SRM PMs, because they are at such different price points. This product should be viewed as a completely different type of on-bike instrument and the question is, "What is its value?" Asking whether it is more or less accurate than a PT or SRM is really the wrong tack. It provides new data, acquired in a completely different way than anything on the market. What is the value of that data? Probably not much for someone who owns a PT or SRM. But, what about the majority of cyclists who don't now have a PM and aren't likely to buy one at today's prices? What degree of accuracy is required to be of any value at all? Is it acceptable that the device produces power numbers that are off by as much as 20% at any point in time (e.g., actual power is 200w but the device says 240w) but averages out to 5% over time (e.g., 10 minutes, 20 minutes or an hour)? Maybe it shouldn't be compared with today's PMs at all. Maybe it should be compared with other computers in this price range, which typically have HRMs for management of intensity of effort. What degree of accuracy makes this device more useful than a HRM for training purposes? Those of us who own PMs will be asked by our PM-less cycling friends (who don't want to pay $1K+ for a PM) whether we think this device is worth the money. What should we say? To say, "Well, it's not as accurate as a PT or SRM." seems to be a cop-out. Our friend should say, in response, "Well, I'm not prepared to buy a PT or SRM. But, is this thing worth $350. Does it provide me with more useful data for training than my HRM?" What will we say in response?

kyzyl2 said:

Hey...I pulled that quote off their website FAQ this morning! I don't know what FAQ you're talking about, from what website or forum and at what point in time.

...

Doesn't bother you just a little bit that power meters costing 5 times as much as the iBike have measurement errors as much as +/- 5% (by the manufacturer's own admission), and that as a cyclist you have NO IDEA when, where, or why these really expensive meters are giving you erroneous readings?

All this said, I agree with you that the "proof is in the pudding"; we shall see what the iBike comparison tests show.

They seem to have changed the FAQ on their website at some stage today, it now gives an example for a rider at a constant speed on the flat where aerodynamics are a much bigger component. This is a good sign, as they are now picking an example where their kit is at more of a disadvantage and still showing a reasonable accuracy.

Of course, the statement about honesty assumes that the widget is otherwise 100% accurate, which is isn't going to be. They quantify one area of error, which you can allow for, but there may/will still be errors which you can't allow for.

Their site also says that they won't be in a position to send any products out until December, so there's nothing to lose in waiting for more detailed studies to come out. The thing that I'd like to see is the comparison between zero cadence and zero power. It seems to me that if these can be independently shown to correlate then that would be a good sign that it's getting the resistances right.

I don't currently own a PM, but I have been educating myself on the subject. I definately want to get away from the HR zone(s) method as I believe training with power makes more sense.

Here is my question for all you PM owners and gurus. Shouldn't I be more concerned with a device that is consistent vs accurate? Should I really care if I'm pushing 250w vs 235w? As long as the device always says I'm pushing 250 for that effort and doesn't drift or vary. Is this in fact what you guys are saying when you say accuracy?

For me, I think the most important aspect in choosing a PM is one that gives me the same data for the same levels of work and does it consistently. I don't want to think I'm training in one zone when in fact I'm actually training in a different one because the PM can't stay calibrated (for lack of a better term). Should I really care if I'm actually putting out the number the device reports as long as that number is consistent with the measuring device?

If I am interpeting the FAQ correctly, this is where the iBike Pro will suffer. When you conduct your tests to determine your zones the results will be based on position. In other words, if you are in the drops your data will only be accurate for that position. Depending on the size of your shoulders, head etc you could have a ? percentage error during your workout if you use another position.

Thoughts? Remember, I'm a newbie at this power thing so feel free to correct, chastise, or whip me with a wet noodle.

Rocket^ said:

If I am interpeting the FAQ correctly, this is where the iBike Pro will suffer. When you conduct your tests to determine your zones the results will be based on position. In other words, if you are in the drops your data will only be accurate for that position. Depending on the size of your shoulders, head etc you could have a ? percentage error during your workout if you use another position.

Thoughts? Remember, I'm a newbie at this power thing so feel free to correct, chastise, or whip me with a wet noodle.

Most riders spend most of their time riding in the same position (the iBike FAQ estimates 90%). If the cyclist does the coast down calibration in that standard position, then the aerodynamic power component of the iBike will be completely accurate. Only when the rider position shifts is an error introduced.

By the way, if you stand up because you're on a steep hill, the aerodynamic power "error" caused by changing riding position will be swamped out by the watts you're putting out to get up the hill!

So it really depends. You might want to check out the iBike website www.ibikesports.com.

frenchyge said:

Before you beat the drum too loudly about honesty, I hope you can see that the aero example you chose is the least relevant to the rider and the most favorable to iBike's marketing staff. Fortunately, that's not the example they use in their FAQ (or else they changed it after you looked because it made no sense).

Personally, I'd be interested to see how well it deals with the turbulence created by being positioned directly behind another cyclist. If it's mounted on the handlebars then it's not seeing the wind that's hitting the riders head and shoulders. If it handles real-world situations as well as or better than existing products then that's great. I'll wait until the comparison data comes out before starting the cheerleading or trying to shoot it down.

I'm all for a fair shake and would be happy with +/-5% for a device like this. They *were* claiming 0.5% on the website but that's clearly unrealistic.

My frontal area on the TT bike is 40% !!! higher on the cowhorns than while tucked on the aerobars. If we assume Cd is the same (and it could be likely be higher on the horns), my aero drag and power would increase ~40% in that position. At TT speeds of 45 and climbing at ~30kph, this would introduce very large errors. I would ballpark 20-30% (depending on speed of course). While standing, the error would increase even more ...

I believe the postion change issue will be more troublesome than the variation in rolling resistance. And this ALL assumes they have a reliable means of determining hill gradient. And you have a perfectly flat (and I assume windless?) place to perform your coast-down test.

Power just isn't that easy unless you're directly measuring force/torque and applicable linear/angular velocity.

I think the ibike will prove to a power predictor or approximator rather than a true Power METER.

rmur

kyzyl2 said:

Hey...I pulled that quote off their website FAQ this morning! I don't know what FAQ you're talking about, from what website or forum and at what point in time.

This one: ibikesports.comfaq.htmlOpen ↗ question #3, 10:27am CST.

kyzyl2 said:

Perhaps the more fundamental point here is that ALL power measurement devices make approximations and ALL make power measurement errors. It looks like to me that the iBike guys are trying to be straightforward about where they make their approximations, and what the accuracy impact will be.

It's true that they all have error, but precision, repeatability, and the impact of changes *during the ride* are more relevant to the rider than the fact that there's a relatively fixed error from drivetrain friction or calibration offset. The danger is that things change during the ride that the rider doesn't realize, and if his PM doesn't pick it up (or worse, shows an offset in the opposite direction) then he can't understand or compensate for it.

kyzyl2 said:

Doesn't bother you just a little bit that power meters costing 5 times as much as the iBike have measurement errors as much as +/- 5% (by the manufacturer's own admission), and that as a cyclist you have NO IDEA when, where, or why these really expensive meters are giving you erroneous readings?

Well it might've.... but I didn't buy that one. As far as having no idea when or where the errors are occuring, if your PM is *precise* then the errors aren't randomly occuring during the ride. If they're known and constant then understand and compensate for them and move on. Comprehensive testing, in real-world conditions (like this one: biketechreview.compm review.htmOpen ↗ ), goes a long way to determine the precision of the various devices and make the consumers comfortable that their PMs are not randomly inserting +/-5% errors into their readings. I happen to know from my profession that getting precise readings of wind speed is very difficult when there are sources of turbulence nearby, so I'm skeptical that the iBike's precision would hold up under real-world cycling conditions. If it does, then great.

RapDaddyo said:

Even before they post their test and comparison data for this new PM product, it's interesting to consider what the relevant comparison should be. I think it does this product a disservice to compare it with the PT and SRM PMs, because they are at such different price points. This product should be viewed as a completely different type of on-bike instrument and the question is, "What is its value?" Asking whether it is more or less accurate than a PT or SRM is really the wrong tack. It provides new data, acquired in a completely different way than anything on the market. What is the value of that data? Probably not much for someone who owns a PT or SRM. But, what about the majority of cyclists who don't now have a PM and aren't likely to buy one at today's prices? What degree of accuracy is required to be of any value at all? Is it acceptable that the device produces power numbers that are off by as much as 20% at any point in time (e.g., actual power is 200w but the device says 240w) but averages out to 5% over time (e.g., 10 minutes, 20 minutes or an hour)? Maybe it shouldn't be compared with today's PMs at all. Maybe it should be compared with other computers in this price range, which typically have HRMs for management of intensity of effort. What degree of accuracy makes this device more useful than a HRM for training purposes? Those of us who own PMs will be asked by our PM-less cycling friends (who don't want to pay $1K+ for a PM) whether we think this device is worth the money. What should we say? To say, "Well, it's not as accurate as a PT or SRM." seems to be a cop-out. Our friend should say, in response, "Well, I'm not prepared to buy a PT or SRM. But, is this thing worth $350. Does it provide me with more useful data for training than my HRM?" What will we say in response?

Again, no-one's seen any proper data yet... but I would imagine the iBike will be most accurate when you're climbing, and the slower/steeper you're climbing, the better it'll be, because the other resistances will be small compared to the gravity component. So, if you tend to do most of your training up and down hills, as I do, this will probably be pretty good. If you live in a flat area and do your quality work at high speed on the flat, then the error will be bigger, and maybe unacceptable.

10% would be at the limit of acceptability for me, that's +/- 20W at 200W, which is ok since we all know that you don't hold an exact figure anyway, so we're probably all working with zones of between X and X+deltaX. For me, during a long endurance ride I might aim for 200W, and so anything between 180 and 220 would be alright. Since that's just for endurance and the quality stuff is done in the hills where the error will be less, that for me would be alright. It seems to be that there would be a rapidly diminishing return in paying 5x the amount for more accuracy that I might not really use. If I were a big flat land sprinter that wants to do loads of sprint kicks, or a timetriallist needing to bang out 350W for long periods on a pancake flat road, then you might need to pay the extra. If I was aiming for 500W, and this might mean anything between 450 and 550 then I wouldn't be very happy - for the above suggestion of 20%, this would be between 400 and 600, which I would say is of no use to anyone.

The consistency issue raised by Rocket is very important of course. I usually go to get tested at a local lab on a Kingcycle, so I would need to have a Rosetta Stone type test to check what to allow for on the road, but that's the same for all PMs if you test on a different set-up to the one you train on. Presumably since it asks for a roll-down test at the start of your ride, then the iBike will be consistant within it's own terms for a given situation. Of course, whether it's consistant between different situtations - is 300W reading uphill into a headwind really the same as 300W reading downhill with a howling tailwind, and would you ever know? Since it's relying more heavily on different instruments for its readings at different points there will be issues with the different degrees of accuracy that each can provide...

Uh...what approximations do PowerTaps and SRMs make?

The Ergomo makes one assumption because it can't measure the right leg so I consider it the best of the power guessers along with the iPod and HAC4.

There have been studies that have validated the accuracy of PTs and SRMs. Also, it's quite easy for any user to perform a calibration.

Here are the numbers from the same 2x20 workout on a nearby mountain for 3 consecutive weeks:

9/04 295 271
9/11 295 270
9/18 295 270

Interesting that the same workout with roughly the same amount of motivation yields such similar numbers.

On Sept 3, I performed a calibration and got a slope of 19.5403 Nm/Hz. On Sept. 30, I decided to service the BB which required removing the cranks. After reinstalling the SRM, I performed another calibration and got 19.5403.

People complain about the price of SRMs but have no problem buying frames or wheelsets that are in the same price range yet do little to increase their performance. I bought my first SRM when I was a college student. It was a tough pill to swallow but I never regretted it. The last time I bought a new frame was in '99 but I ride past all kinds of people here with fancy carbon rides like they're standing still.

One time I passed someone I knew but didn't recognize on a climb. When we were talking at the top, he said that I rode by him so fast that he thought, "That's [censored]!"

None of this would be possible without the SRM.

Despite all the newcomers to the market, SRMs are selling faster than ever. Obviously, the market supports their pricing.

Lastly, I'd like to see consumer comparisons of the iBike with powermeters. I'm wary of studies commissioned by manufacturers.

kyzyl2 said:

Perhaps the more fundamental point here is that ALL power measurement devices make approximations and ALL make power measurement errors. It looks like to me that the iBike guys are trying to be straightforward about where they make their approximations, and what the accuracy impact will be.

Doesn't bother you just a little bit that power meters costing 5 times as much as the iBike have measurement errors as much as +/- 5% (by the manufacturer's own admission), and that as a cyclist you have NO IDEA when, where, or why these really expensive meters are giving you erroneous readings?

All this said, I agree with you that the "proof is in the pudding"; we shall see what the iBike comparison tests show.

rob of the og said:

10% would be at the limit of acceptability for me, that's +/- 20W at 200W, which is ok since we all know that you don't hold an exact figure anyway, so we're probably all working with zones of between X and X+deltaX.

It certainly depends on what you want to use it for, and precision is more important than accuracy. Coggan's power zones are 15% of FT in width, so an accuracy of +/- 10% bascially means that you're "probably in the right *zone*." I'm not sure that'd be good enough for me, personally.