validation of the powerStroke method

Post Reply
rons
Posts: 28
Joined: Sat Dec 08, 2012 7:54 am

validation of the powerStroke method

Post by rons »

Not to throw water on your fire, but when ibike launched you were quick to show it correlated to the standard power meters. In this case I am not yet 100% confident this really tells you anything? Has it been compared to say a set of Look power pedals where you really are measuring force as it relates to both legs? Further 16 data points per second is not that impressive when you think about 90 RPM, that is 12 readings were revolution or every 30 deg in the circle. So from top dead center to 90 deg you get 3 readings. I have read research papers that look at pedal force using 120 Hz (readings per second) that show the force profile for an elite cyclist. I think it important to think critically about new ideas. Not that you are wrong, but that you need to prove your concept not blindly believe. I am interested in the concept but do not see enough data to prove that correlation of front to back and side to side oscillations really relate to efficiency or power production nor pedalling dynamics. Even the best riders do not produce an ideal round power output to the pedal.

So though I am not saying this function does not work I have not seen enough data to give me cause to believe yet. Of course you are unlikely to share the math, so it is simply a black box. The Newton however is a very nice power meter and fits the bill and niche it is designed for extremely well. Over the years it has become easier to use and is now almost as easy to start using as any DFPM. To add $200 (40% of the initial investment) for a software option, that is not visible on the bike, that has questionable utility when sitting at your computer is not a given. Let's say I am in need of convincing.

Do not get me wrong, I think my Newton is a great device. Let's be honest there are still lots of doubters but I have seen plenty of research and can see the patents to see how and why the thing works. Let's see something to drive belief in the PowerStroke. It is a cool idea.

So let the discussions begin both pro and con!
Velocomp
Velocomp CEO
Posts: 7804
Joined: Fri Jan 11, 2008 8:43 am

Re: validation of the powerStroke method

Post by Velocomp »

I think this is a very fair and specific post. There's a lot to comment on, both PowerStroke and otherwise, so let me take a bit of time.

"In this case I am not yet 100% confident this really tells you anything?"

We try to be as specific as possible about what PowerStroke measures. We devote 10 pages of the instruction manual to what is being measured, and how to interpret it. And there's a video, too...

Remember: until PowerStroke, no one has ever measured dynamic motion of the bike and rider. What can be done with this new and exciting data is wide open to exploration and innovation.

"Has it been compared to say a set of Look power pedals where you really are measuring force as it relates to both legs?"

The question includes an assumption: that the Look pedals work as advertised. How have their claims been validated? We haven't seen the data. In fact, because of some of the things we learned while developing PowerStroke, we are extremely skeptical of any pedal-based power measurement device. Why? Um, that's proprietary, but suffice it to say: even after years of development, Garmin has been unable to bring its pedal device to market. Brimm has been unable to bring its pedal device to market. We think we know why.

"Further 16 data points per second is not that impressive when you think about 90 RPM, that is 12 readings were revolution or every 30 deg in the circle. So from top dead center to 90 deg you get 3 readings."

If you were using ANY crank-based PM, you'd get exactly one reading per 360 degrees, vs twelve readings per revolution for the PowerStroke. And we improve resolution effectively to 36 readings per second (nine readings between 0 and 90 degrees) using some very clever analysis. We think that is very impressive!

That said, when we can we will make a posting to describe more details of what we're doing, and how it works.

"I have read research papers that look at pedal force using 120 Hz (readings per second) that show the force profile for an elite cyclist."

We have, too. We are using that research as part of the "Ideal Motion" plot shown in the PowerStroke window. To be clear: our plots show the MOTION caused by forces, not the forces themselves.

"Not that you are wrong, but that you need to prove your concept not blindly believe."

Point taken. We have some patents pending, so we cannot be too specific about some things, but we certainly will try to be more descriptive.

"I am interested in the concept but do not see enough data to prove that correlation of front to back and side to side oscillations really relate to efficiency or power production nor pedalling dynamics."

In every sport we think of, outside cycling, athletes and coaches spend years trying to improve the dynamic motion of the sport (golf swing, swim stroke, throwing motion...you name it). In cycling, almost zero time is spent honing the only dynamic motion involved: the pedaling stroke and the cyclist's movement on the bike! Why? In part, it's because that motion has not been measured! If dynamic motion is critical in every other sport, should it not be in cycling, too? PowerStroke provides the first-ever means to measure dynamic motion.

The reason you haven't seen more data is because, until now, no data existed!

"Even the best riders do not produce an ideal round power output to the pedal."

What we have seen, over and over in our months of PowerStroke testing, with many different riders of vastly different abilities, is that the best riders (defined as: podium-taking competitive cyclists) have distinctly smoother pedal motions than amateurs. The winners may not be perfect, but their PowerStroke plots are d*** good!

"Of course you are unlikely to share the math, so it is simply a black box."

We will try to more specific; it's cool and very robust technology. But you're right, too: we aren't going to publish the source code :D

"The Newton however is a very nice power meter and fits the bill and niche it is designed for extremely well. Over the years it has become easier to use and is now almost as easy to start using as any DFPM."

Thank you! We are working now on a major FW and SW release for Newton that will make your compliment 100x more true! Coming soon...

"To add $200 (40% of the initial investment) for a software option, that is not visible on the bike, that has questionable utility when sitting at your computer is not a given."

Actually, it involves firmware, software, and hardware changes. In fact, this is the most complex technology we've created, since the original iBike itself.

This is PowerStroke 1.0, which we wanted to introduce at the beginning of the season. Anyone who has followed us over the years knows that we constantly improve core technologies, and provide nearly all of those improvements without additional charge. PowerStroke is a new and very core technology. You'll see more coming from us soon.

But even with the $200 charge for PowerStroke (our best product), PowerStroke price is lower than the cheapest version model of our competitors. Furthermore: if you buy any competitor product, for your money you'll get exactly one thing: a power sensor. You'll need to spend extra on a computer head, and install 3rd party software, and pay your mechanic to swap-put your pedals,wheels, cranks, etc. And with your DFPM you won't get CdA; you won't get Effort Speed; you won't get out-of-saddle, drafting, insights,or lots of other interesting other data; and you won't get access to new technologies such as PowerStroke. AND YOU NEVER WILL, because competitor products are intrinsically limited in what they measure.

"Do not get me wrong, I think my Newton is a great device. Let's be honest there are still lots of doubters but I have seen plenty of research and can see the patents to see how and why the thing works."

iBike theory was invented in the 1500's (by Newton); but reducing his theory to practice took a long time!

Give us a bit of time to get our next big thing out there, and then we will spend more time on PowerStroke explanations.
John Hamann
rons
Posts: 28
Joined: Sat Dec 08, 2012 7:54 am

Re: validation of the powerStroke method

Post by rons »

Thanks John. I appreciate the time you took. I was not saying one or the other of the look pedal system was correct, but it is more a thought experiment that if you look at two systems and get similar results then there are reasons for further belief. Similar to good ole math, you can solve a problem using different mathematical approaches but they should both achieve the same answer.

When I talk about the $200, to me the hardware in the Newton I purchased is mine and assuming I go for no further enhancements that cost is sunk for both of us. So I reiterate that the PowerStroke is software and that once developed is not a cost to build the unit you sell to me. In other words then the cost of development must have been enormous when the hardware costs $500 and a software add on costs $200. From a marketing point of view that is a hill for me to climb. Another thought to gain acceptance and user base is to offer it with a limited run trial license then if I want it I pay. Once I am hooked I will probably find the versatility.

For instance this afternoon I was on the trainer (it was snowing outside today :( ), not a set of rollers, and I noticed that my bike was oscillating left and right in sync with the pedals. If I could isolate just that aspect of PowerStroke vs pedal velocity, power and the calculated pedal force it would be possible to see a plot of left and right pedalling dynamics similar to your left right aspect of the plot. So based upon this "very limited" impression I do think that powerstsoke might be able to work on a trainer, but based upon working with people doing bike fitting it is much easier to show something in real time vs expecting them to relate data to something that occurred even minutes ago. So that would be my holy grail for PowerStroke to be able to display in real time on the unit or on the computer screen.

So again, thinks for taking the time to respond. I still think there needs to be a broader field of knowledge to make it something that people will believe in similar to the Newton. As I mentioned there are still doubters in the effectiveness of the Newton, though the evidence is out here now. Also, as you note, it collects far more data than any other cycling device of any sort and far more than any DFPM. So I am a Newton-o-phile and can't wait to see the next things coming. I only hope they will not cost as much as the PS add on.
Velocomp
Velocomp CEO
Posts: 7804
Joined: Fri Jan 11, 2008 8:43 am

Re: validation of the powerStroke method

Post by Velocomp »

rons wrote:
From a marketing point of view that is a hill for me to climb. Another thought to gain acceptance and user base is to offer it with a limited run trial license then if I want it I pay. Once I am hooked I will probably find the versatility.
I have a friend who is an extremely accomplished artist. He paints magnificent landscapes, always while standing outdoors. He told me that, once, someone walked up to him, while he was creating one of his paintings, and asked, "how long did it take you to paint that"? His answer: "About 25 years".

We've spent 9 years improving and extending our technology. We think about things others can't even imagine, because we have totally proprietary and unique technology. We create products that reflect our passion for precision measurement and expert guidance. And we try to provide solid value for everything we do.
John Hamann
cjonbike
Posts: 113
Joined: Sat May 03, 2008 1:37 am

Re: validation of the powerStroke method

Post by cjonbike »

And we Thank You for all you do!!!
rons
Posts: 28
Joined: Sat Dec 08, 2012 7:54 am

Re: validation of the powerStroke method

Post by rons »

+1 on the thanks for your efforts!

However, for you artist friend and all of the rest of us ,beauty/value is in the eye of the beholder. We have all spent a lifetime to be what we are, the value is in what we do today, yesterday is history and open to interpretation.

So keep up the innovations, the Newton unit is hands down a great value proposition in my humble opinion, as are your options for other upgrades and the main software Isaac is free so that is the most significant reason one might see the PowerStroke is an expensive add on. But hey that is just the opinion of one small voice and your sales will be the real test of public opinion. So I am sure you had long discussions to come up with the price and had to decide on your strategy. I appreciate the responses and can see this function is close to your heart.

Looking forward to the future of iBike power products and glad to be a Newton owner!
jayt
Posts: 45
Joined: Wed Jan 23, 2008 11:19 am

Re: validation of the powerStroke method

Post by jayt »

First, I think the Newton is a fantastic product. I started with the iBike Gen I (and Gen II, skipped Gen III). You have finally hit a product that I can recommend without reservation. You solved the battery problems, you are wireless out of the box, calibration is minimal, easy and does not have to be constantly repeated. I can finally put the ting on my bike, make sure (if I want) the altitude is correctly set and away I go. It pretty well self-corrects for tilt problems, I can ride and believe the data (the only caveat so far is I was in a strong cross wind and the readings seemed a bit off, but that is a very particular situation). All of that to say, good product and at a good price.

When you came out with the PowerStroke, I was at first excited, and I was running in my mind how much would I pay (I figured it would not be free, and to be honest I don't mind paying something for a major feature), however, for me, $200 is pretty much out of the question versus the (perceived) utility. I would see the PowerStroke measurement as an interesting thing to know and maybe re-test every now and then, but it is not a primary thing I would watch and be trying to monitor and correct all the time. To me a possible better marketing strategy would be to have offered at a much lower price point, build support and enthusiasm for you measurement (after all you are pretty much the only folks that can do this one) and drive more sales of your main device, the Newton. I'm sure you will have some sales but I think it is probably pretty limited. I'm in product R&D and I understand needing to recoup your R&D investment, but there is more than one way to do that. I greatly appreciate that you did not charge for you OS 2 upgrade, but chose to give us all a better instrument. As always, I wish you the best.
daidnik
Posts: 33
Joined: Tue Jun 30, 2009 8:35 pm

Cost/Benefit of Powerstroke feature

Post by daidnik »

I'll throw in a contrasting viewpoint in the Cost/Benefit analysis which seems in principle the focus of this thread.

I first got the Gen3 without the software for $200. Great product; novel product; gave me what I wanted namely an objective tuning metric on which to understand and optimize budgeting of power output. I once told a friend after he tried a burst of power on the hill that pooped out as I was pacing him:

"I know what you can do better than you know what you can do because I can measure your output." This was entirely true as well.

Later I added the software key for about another $250. It seemed like a lot of money, but I thought why not add some more money be able to save the data for reference over time. In this upgrade, again I got more than I had expected. I saw how the 'indirect' method of calculating the riders output by measuring the forces 'he' does work against actually provides the most meaningful data set for analysis, i.e. elevation profile & wind. Again, I was happy with the Cost/Benefit.

Did upgrade to Newton for another $400. Seemed like a lot of money, but my Gen3 mount was failing and by now, the iBike until was in the 'must have' category. New mount much improved, battery life great; added cadence and found it more useful than I thought. I could count it pretty well, but now I don't have to think so much.

When the Powerstroke idea came out, I read about it and understood the basic principles and realized that it was an 'uncharted data realm', but since the iBike had the hardware infrastructure to 'tap into' this uncharted realm, I decided to give the manufacturer the benefit of the doubt and throw in the $200 to support the effort.

In my case, I was/am recovering from severe Tib/Fib leg fracture that was inflicted on my by a car bumper side impact just over a year ago. I've been riding for months now, but nerve recovery is still going on and will be for another few months. What appealed to me in this context was to exploit the increased data acquisition rate on the iBike and to look "inside" the pedal stroke cycle to see and try to understand symmetries/asymmetries in the dynamics of motion.

Now, regarding the cost/benefit analysis on this feature addition, the picture may be less clear depending on what you expect(ed).

Does it enable you to see "inside" the pedal stroke cycle? Yes!

Will strain gauge pedals enable you to see "inside" the pedal stroke cycle? Sort of.

Unfortunately, merely measuring the applied force vs time on the pedals will not tell us very much about the dynamics of the system in response to the applied force even though that is really the most important data to be acquired.

Once again the 'indirect' approach to measuring work done per unit of time wins again. It enables us to see the response of the system. This is a richer data set than force vs time even with Left/Right discrimination as with the pedals.

The only hard part of the Cost/Benefit analysis on this feature is that 'immediate gratification' will likely not be as strong as some features. The data is not the easiest 'dimension' of the picture to understand. I am hoping that as Powerstroke starts to catch hold, people may post their data sets and a better understanding of this new 'uncharted realm' may become more clear.

Back to my specific situation, my riding is over varied topography and longer rides; 80-100 miles with 6000+ ft integrated climbing. I can tell you that in just two of those rides with the Powerstroke that I could see the effect of the one leg being stronger than the other especially with regard to the lateral motion. From that and a little study of nerve sensation on the injured (Left) side and a little psychic intuition, I realized that I needed to focus more on pushing my big toe side in my footing on the pedal which would also strengthen the Soleus muscle.

To have a means of capturing data to show 'balance' in the dynamics of the pedal cycle has some value. Just how much value it brings you will take some time and study and hopefully this subject will 'heat up' a bit on the forum.

It IS a very interesting data set!
rons
Posts: 28
Joined: Sat Dec 08, 2012 7:54 am

Re: validation of the powerStroke method

Post by rons »

I certainly do not disagree that the dataset is interesting, what would help to move someone like me is have the relative effect of pedalling as a display on the Newton in real time. To affect a change in dynamic motion the best feedback is immediate. I grant that I can not comment on live data in my own situation since I am not so keen to toss $200 to find out. To me and I think like some others out there if the software was made available on a limited time free trial (15 days) which would revert after that time without a full paid key you might hook me. To see if I like it at $200 is too much for the entry into that dataset. I may be completely wrong on the marketing strategy and time will tell. What I love about iBike is they are on the move and that is exciting. The depth of iBike data is the crack cocaine of devices. Is it the easiest to work with? probably not but on a cost benefits the device is hands down the best, no argument.

Bottom line: PowerStroke neat innovation, cost /benefit equation still need to be convinced.
daidnik
Posts: 33
Joined: Tue Jun 30, 2009 8:35 pm

Re: validation of the powerStroke method

Post by daidnik »

I agree with you on that one but I don't think that will be forthcoming soon.

Increased cost/complexity of the display would certainly be a factor. It also might be too much data to 'digest' in real time. The increased computational analysis would be huge judging from the time it takes Isaac to chug thru the data using a PC that has to be several powers of 10 faster than the Newton.

I think that once the data set is analyzed for the Powerstroke, then it can be pulled into the any Isaac install above the req'd ver level and viewed. Here I am thinking that Isaac analyzes the data upon upload from the Newton using the profile to cite the position of the cadence sensor; does it's proprietary extrapolation to increase the resolution to 32 points/sec as Velocomp cited earlier in this thread. Then, I think saved file is portable.

Given this, I will try to post some of the past rides with the Powerstroke option and you can check it out. It really does require some study.

What I think the potential improvement to be gained from this data is more of an 'overall efficiency' factor. For me that equates to either an increased avg power on these ~103 mile rides (same route each week); an easier time climbing those last hills on the way back, or both. It's early in this analysis, but I think I may be seeing some improvement in last Sat ride over the previous week.
Post Reply