That it is a good price. But I would want to see it independently tested and verified before shelling out for it.

What is the story with Velocomp, anyway? Their iBike unit provided the same functionality...did they go bust and are trying to kickstart their way back into business, or something?

Every edition of the iBike, including the first, claimed that it was as good as a strain-gage power meter.

Every subsequent edition of the iBike claimed that it was better than the previous edition.

To be fair, they're explicitly note a direct force power meter is required for the aero portions here. No different than anyone else in the industry.

---

-
My tiny little slice of the internets: dcrainmaker.com

Sure. Past performance is not a guarantee about future results but I was just pointing out that they have a historical record that includes claims about their products.

I got in at the $299 level before it filled up. At that price I figure it's worth a try. It may not be as accurate as something like the Notio, but that one's expected to be closer to $1K and I haven't heard anything about a release date. At least I can start testing with the Aeropod, I can always upgrade to something more accurate later if I feel the need, once this product market starts to shake out and mature.

jsk wrote:

I got in at the $299 level before it filled up. At that price I figure it's worth a try. It may not be as accurate as something like the Notio, but that one's expected to be closer to $1K and I haven't heard anything about a release date. At least I can start testing with the Aeropod, I can always upgrade to something more accurate later if I feel the need, once this product market starts to shake out and mature.

I just got one for 349$ all others were sold out already, Ray video shows working prototype:
https://www.youtube.com/watch?v=L7i-jklh7wE


There is no big risk, additionally they have Issac software and experience from the past products which at 350$ makes them super interesting.
I'm sick and tired of waiting, completing questionnaires, signing up for #metoo lists from other companies, begging to be able to give them 1000+$ for totally new products that will most likely have a lot of kinks to iron out.

Alternatively, you could save $200 and start collecting data as soon as this arrived at your doorstep:

https://www.ebay.ie/itm/201383461691

This isn’t my forte, but isn’t that pitot tube awfully close the frame/underside of the aero bars? Is that considered clean air or do you need to modify to mount further down and out?

---

Team Zoot - Great Lakes

Probably needs half a dozen license keys and in-app purchases to make it do anything. I really dislike their business model of trying to upsell on every single tiny feature.

ibike was only displaying 2 decimal places 0.00 new toy is showing 0.000 most importantly it has Garmin integration via IQ apps.

This appears to be simply an update of the Powerpod they've been selling. The iBike variants have been able to do CdA measurements for many years.

There are two key accuracy/precision aspects of these devices. The first is making really good wind direction and and speed (or dynamic pressure) measurement. The other is positioning the device far enough from the effected zone of the rider+bike and at an elevation that represents the center of drag of the rider+bike. So well forward of the rider+bike (probably a couple feet anyway) and around the armpad height (just guessing).

If you see one of these that is so small and well integrated that you could imagine riding with it on your bike all the time, you know it's a toy. You'd be better off with proper Chung VE testing, even if there is wind. Maybe much better off. If that is too much trouble then use BBS or something similar to get an estimate and call it good enough.

I think there's a market niche for products like this. They're a gateway drug. Some people will be happy with it, others will move on to more precise ways to measure drag. There's room for everyone. A decade ago, only a handful of people really understood how much speed they wasted because it's hard to "see" aero or rolling drag. These devices make things that were previously invisible, visible. That's a good thing.

I have had an iBike for a number of years and though it could do some of what this can do the display was only 2 digits ie 0.34 so that may have been due to the limitation of their sensors at that time. I have to say the iBike worked OK but has it quirks, I also have a 4iiii Precision which I like a lot. The two together are fun to look at the data after a ride. As they say in their kickstarter page, and I have no reason to think otherwise, that sensors etc have come to a higher level and it is this higher accuracy or precision (I am not certain which it really is) has allowed them to come up with better resolution of CDA. I have no doubt that they will deliver to their price point and like everything else you hope you get good value for money. There are aerodynamics experts who understand the whole story of measuring CDA in a dynamic environment and who can talk to the potential of this really working but since it is just a tweaking of their current offering I think that it should be a reasonable option. At some point there will be other commercially available competitors and then one can see how well they work in the right test chamber which for now is the big thing that I think is lacking. So though it is an interesting development there should be some independent method to confirm that the thing is giving correct data. At least I would want to see that before I paid anything for it. Finally one has to remember the device alone can not derive CDA it has to have a good power meter to work with so that is one true fly in the ointment, and I leave that item to DC Rainmaker's professional assessment.

This is John Hamann from Velocomp...full disclosure up front.

A friend told me to check out this thread. I haven't visited slowtwitch for about 10 years now, so reading some of the comments is like a trip down memory lane!

I respect the perspectives and knowledge of the contributors to this forum, and so I thought I would directly respond to some of the points made.

Believe it or not, it's been about 14 years since Velocomp began. That is a lot of time to make a lot of mistakes, and to make a lot of improvements. We've done both.

Our first CdA product, iAero, was introduced 10 years ago. Ten years ago we used analog sensors, which had lots of stability, resolution, and calibration challenges. Ten years ago, a good DFPM was...what, $1500 or more? Ten years ago, we were still in the early days of figuring out how to do real-time signal processing from a lot of different sensors.

So, yeah, 10 years ago, some of iAero's performance attributes fell short of expectations; we discontinued the product about a year after introducing it. But its underlying foundations were sound.

We've worked very, very hard to improve, and it is an irrefutable fact that every device we have made has been better than the previous one. And AeroPod is better than PowerPod:

1) we have spent a lot of time working on the wind port design. A pitot tube approach is better, particularly in cross winds, than our traditional approach

2) We are using completely new sensors in AeroPod, particularly the wind sensor and accelerometer. I would say more about this but I really don't want to give away details that have taken us years to figure out.

3) Folks, measuring wind is not easy, but we have 13 years experience in knowing what works, what doesn't, what to look for in testing, and how to test. One of the things I'm looking forward to, some day, is seeing data from competitive devices. We think the school of hard knocks will be equally tough on anyone who tries to measure wind.

4) AeroPod signal processing algorithms have been changed substantially. In fact, while doing CdA testing we discovered a few subtle, yet material, errors in our CdA coding that ultimately led to compromised performance. We found these bugs only because we were pushing ourselves to get better results. This ultimately led us to some significantly new signal processing algorithms we are using in AeroPod.

5) For testing validation we have been using ride files from high-level triathletes who have been tested in wind tunnels. Their results so far vary only in the third digit.

6) One of the posters on the forum made a comment about wind sensor design and calibration. That's a solid point; accurate CdA data requires good calibration, a good wind sensor, and careful design of wind port and case. Everything we have learned about all of this, over 13 years, is going into AeroPod.

7) For our "release 1.0" we are going to hold Crr constant. We are fully aware that this is a limiting assumption, but it is a start. Once we have things settled down we will focus on relaxing this assumption.

8) We are also aware that yaw angle is a factor. While it is relatively easy to add wind angle measurement, it is expensive. Until we see the response of the market to AeroPod, we don't think it makes sense to add more cost to the first generation product.

9) We are trying to be careful about representations. We DO measure CdA out to 3 places, and we seem to be getting good accuracy and consistency in our testing. But CdA is NOT an easy measurement, and it necessarily relies on good DFPM measurements, too. We have been using a variety of DFPMs in our testing, and I can tell you that some of them fall far short of what's required. But there are some really solid performers, too.

10) To argue that AeroPod is a "tweaking" of PowerPod is, to me, like saying iPhone X is a "tweaking" of the original iPhone.

We know you guys have high standards and will be hard on AeroPod...that's fine.

We look forward to your evaluation of AeroPod, and if you have ideas that will help make it better, we will be all ears.

---

John Hamann

I think that $$$ will be extracted from cyclists. But no one will be happy with it as a means of of measuring drag, unless the novelty of getting a variable number that is within 5-10% of reality (maybe) is all someone wants. iBike has had this functionality for many years, but it isn't good enough to get actionable data on positioning and equipment.

The more precise devices that actually can be used for CdA determination, will be considerably more expensive, not so small and well integrated, and will require a good deal of time and care to use properly. I don't think Aeropod owners are going to jump to something like that. The market for those will be people who already do VE testing and want to improve the accuracy in more challenging environments.

I guess one thing in Velocomp's favor is that the Aeropod will function as a powermeter (not a real good one, but better than a aerometer), so when the novelty of seeing CdA numbers wears off you can stick it on a spare bike for that purpose. I had the iBike (from Gen 1 to 3 I think), and they did get slightly better with time. My biggest beefs when I gave up were terrible chipseal performance (which they have "fixed" recently) and the sole reliance on the barometric altimeter, leading to substantial slope variability (~5%). I don't know if they've addressed that one, but it could be done using GPS to scale the barometric readings.

Will you be providing a list of recommended power meters that perform well? It would be a shame if your product performs well but the CdA calculation is compromised by a brand of power meter you have found to be inaccurate.

kyzyl2 = John Hamann. I won't repeat that boring disclosure again on this thread :-)

Most certainly we have not tested every both-leg power meter brand on the market, but for those we have tried (5 different brands so far), we are confident these do well:

1) PowerTap, any model/configuration

2) Quarq

SRM is not on our list, not because it isn't a great PM (it is), but because SRM's very different method of zero-calibration is something that AeroPod can't handle right now with 100% confidence.

We do not recommend one-leg power meters for CdA measurement. AeroPod measures the total opposing forces of both legs, and when measuring small differences in CdA, you need the DFPM reporting accurately the applied forces of both legs.

We want as many DFPMs as possible to work with AeroPod, so as we qualify more we will add them to the list.

---

John Hamann

I'll echo your thoughts about some meters being better than others for this sort of application, although it comes down to expectations.

In my experience the CdA resolution I typically achieve varies by meter. It doesn't make lower resolution meters worthless, just there are limits to the resolution of what one can reliably test for. In general I get better quality aero test data from SRMs than from other meters. So I'd encourage sorting out that zero offset function.

There are some things I don't bother trying to test for with some meters that I might with an SRM.

The meters I've used in testing:

• SRM
• Powertap (G2 and G3 hubs)
• Garmin Vector (dual sided)
• Quarq
• Power2Max
• Verve Infocrank
• Stages (but with sizeable caveats as you mention, it's a reliability/resolution thing)

I'm not going to rate them because in reality you can get a good one one day and have great data, and then another day have one that's just not so as good.
That said, the SRM tests have all yielded excellent quality data.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au

It is great that John has come forward to discuss the project. I would love to hear from others with intimate knowledge of this sort of design such as SkippyKitten or others in this design space. I am certain this is not a trivial sort of thing to design and turn into a design that can be manufactured. I have had some considerable experience in alpha and beta testing electronic device designs through development and it can be a lot of 2 steps forward 1 step backwards and sometime 1 step forward and 2 back.... I give Velocomp kudos for having done what they have to date. Sure the different devices they make have limitations and have been iteratively better at each step. I have been a user and I understand the limitations and when one buys something there is a simple equation of utility (economics 101) so I realized it would not be an SRM but at the same time it provided data that was reasonable based upon the price differential and my needs. I think the same equation will hold true here as well. There are others entering this space as we speak but right now Velocomp is likely to be the first out of the gate and at a price that is fair for utility. So as with all things that rely on power, I believe precision trumps accuracy though it must be reasonably accurate if it is not precise then it serves no purpose. So data that shows how x number of random units from production work on the same situation on repeated measures would be very nice data to help those who really care about what the device is saying. This then allows the user to refine their aero aspects in real time in the real world, that is a good things for those who care about such things. The absolute number may no be 100% accurate but if trial x is 5% better than trial y you have usable information as long as you have proper resolution and high precision. Since the power meter is a key aspect of this equation then both must have the right level of resolution and precision or the final result will be degraded through propagation of errors. This is the one thing that makes this sort of device vulnerable.

Thanks for the mention. I’d rather not discuss another company’s product for obvious reasons. A few things worth noting in general:

Quarq chainsets and Powertap hubs are good in my experience.
Measuring airspeed is the easy bit, but you have similar issues to power meters (temperature, zero, slope) to solve.
Measuring airspeed where it’s supplying useful data and not affected by the rider or bike helps.
Measuring other data like yaw, speed and elevation change more accurately really helps.
A good gadget won’t make up for bad setup or test process.

---

Developing aero, fit and other fun stuff at Red is Faster

The PM will generally not be the weak link so long as you have a decent one and it's working properly. Measuring airspeed and direction is the real problem to solve. It's difficult to measure accurately and if the device is too close to the rider+bike its calibration will be effected by changes to the rider+bike, and its ability to measure crosswinds accurately will be compromised. Having the device tucked in close to the rider is pretty much a non-starter if you want to use it for evaluation of changes in position and equipment. Is there another viable use for these?

I don't know about your definitions of accuracy and precision, but if the device reading is off in a consistent way, then it can be fixed by a trivial matter of calibration. That isn't a problem. The main issue is repeatability, and reporting accurate values in widely varying conditions. For instance you might get it to report consistent numbers when there is no wind, but crosswind conditions will be off, and if the rider+bike is altered the calibration will be off as well. Or it will be accurate when airspeed is 30mph, but off when it's 20mph. Etc.

Hi Ruff, I suspect you are right about air issues, but again that fits with precision and accuracy. There is really only one definition of each in the world of measurement, accuracy assumes that the result of a measurement is in agreement with a considered true value, precision is the ability to get the same answer over and over again and NOT related to accuracy, precision can be great and accuracy can be the pits and vice versa... this is just the old picture of a target with an arrow in the centre being accurate and the next one with a nicely grouped set of arrows but all off to one side of the bulls eye. The golden situation is the ability to be both accurate and precise. So if we assume they will probably be close to true the key remains precision per the standard definition for measurement. It is of no use to have a device that is imprecise if you are trying to find small differences.. now we get into the world of measurement uncertainty with the 95% confidence interval as one definition of acceptable. Again this is where you see the value of precision and of propagation of errors. For instance you might get a reading of CdA of 0.315 but if the uncertainty is +/- 0.015 this suggests that 95% % of the time the correct reading will be between 0.300 and 0.330, so this is where measurements get fuzzy. If the precision is good then one would hope that repeated measurement of the same set up would generate results with a much tighter range than my example. This is where so many things get lost, I am not 100% certain what the algorithm is but I made a WAG that power would be a large influence in the equation and that may be incorrect. So a 3% error there plus a x% due to aerodynamic measures of wind pressure, air density, etc may add up to too much to make a meaningful measurement of CdA in the field, but obviously there are many who feel it is valid to try since there are at least 3-4 different groups (probably more) working on this problem. We all probably know that the CdA equation is non linear so as you increase the air speed the effect of aerodynamic drag increases very rapidly (ie proportional to the square of the speed so double the speed and it affects the drag by 4 times). I am a curious person and look forward to seeing how this and other real time CdA measuring devices work out in the long term. Time will tell if Aeropod is a viable solution, that will be judged on the data.

One may find precision of air velocity data is unaffected by changes to bike/position set up but accuracy is affected. That can result in drawing misleading conclusions.

e.g. a sensor used to collect air speed velocity data may well provide nice consistent data for each bike/position set up and as a result you have good precision, but the sensor may have different bias with each bike/position set up. IOW accuracy may be a function of bike position/set up.

That's the main concern when an air flow sensor is placed so close to the bike/rider, since it's in air flow affected by the bike/rider and measurements made at that location may not necessarily be representative of the entire system.

That's why rigorous validation against known standards is needed. It may be that for a lot of the time the change detected by a sensor is representative of the global change - it's understanding when it's not that is important.

---

http://www.cyclecoach.com
http://www.aerocoach.com.au