Introducing Arofly, “the world’s smallest power meter” + video

IanMunro wrote:

Griff500 wrote:

Almost, you are heading in the right direction, but as described above you are endangering Newton's laws.  

Overall force on the ground (total front and rear) can't change due to pedalling, but there is a weight shift between front and rear. Push the pedal, weight comes off the front of the bike and onto the rear. Push the pedal harder, more weight comes off the front of the bike. The extreme case being where the front wheel lifts altogether and you do a wheely. So there is  a  direct relationship between the power you are hitting the pedal with, and the reduced pressure in the front tyre. Measure this on a cyclical basis, and its easy to filter out short term transients such as bumps, and long term changes such as temperature.  

Nope, there is not a direct realationship,  there is a indeterminate coupling between the two, and that's why it could never work by measuring tyre pressure. 
Using your analogy of the front wheel lifting off the road if you put power in, that's only true if you  have your weight towards the rear of the bike, move your weight forward, you'll get a different reading, change the angle of your peak torque with respect to the ground and you'll  get a different result. Put different forces through the bars, you'll get different results. Do pretty much anything and you'll get different results. At best all you could do woud be determine cadence.
I can also say from my day job, which has in the past involved developing electronics and software to produce a complete solution for the optical measurement of road surface roughness, that it's far easier to say "it's easy to filter out transients", than to actually filter out those transients

Nowhere did I say this device works, and I even said in another post that I won't be at the head of the queue when the product launches in February. As a scientist by training however I don't draw conclusions without the facts, and on this product we don't have all the facts. I have only tried to explain how Arofly say it works, explain the science behind what they are trying to do, and dispel the red herrings you and others posted about the device measuring atmospheric pressure rather than tyre pressure. (Despite the fact that Arofly state in numerous places that it measures tyre pressure and in the manual state that the core of a Presta valve must be unscrewed.)

Yes there are big gaps in the Arofly explanation provided to date, and if it works, they clearly have a trick they are not telling us about. The lack of setup data such as weight or tyre size is mindboggling.  I contacted Arofly to highlight the level of skepticism and asked for more information (adding that I am a Physicist and Electronic Engineer to try and add some credibility). They responded immediately, stating that they understand the skepticism and will give me further information. My breath is baited!

asdfqwerty wrote:

I think the idea is that it calculates power based on air pressure changes within the tyre - that is, you apply pressure to the pedals and that causes a small but measurable change in pressure within the tyre at the valve. Sort of like standing on a balloon, causing the balloon to deform. That means it does more than measure its own air speed or the rate at which the wheel moves (as others are suggesting). I assume that's how they detect cadence even though it only attaches to the front wheel - the cyclic nature of pressure change as you press the pedals is presumably used to infer how fast you're pedalling.

It seems like it would be fairly unreliable though. Surely hitting a bump in the road will cause similar changes in air pressure. I like the idea of a £100 power meter, but it just seems like it would be too unreliable by not directly measuring the force you exert.

As you say, road deformations, moving in your saddle, standing, sitting would all create larger tyre deformations than those from cadence / power variations. 

I say it's BS. No better than a strava gestimation. The website provides no information at all, but patents have been applied for which means the IP is in the public domain. If it was new and ground breaking, it would be all over the engineering world. It's not. I say BS

These are my guesses:

wycombewheeler wrote:

Not sure how wheel speed is related to power. how does this take account of

1) gradient/elevation gain

This is easily achieved using barometers to within a few feet accuracy. However, I'd probably expect that the app pulls elevation data using a live mapping system

2) wind speed/direction

I could be wrong but this looks like it rotate and measures wind speed / direction

3) rider weight

Probably needs to be entered manually into the app

4) drafting

This relates to wind speed / direction. Drafting is effectively just a reduction in head wind

To summarise I think the app does a lot of the heavy calculations similar to Strava, and this device measures wind speed / direction to help make the data more accurate. This probably explains why it doesn't put out ANT+ signal, because a reading of wind speed alone will be useless to other devices.

I bet £50 this won't work on rollers.

Freddy56 wrote:

10grams  imbalance is enough to chuck you over a hedge on a descent. 

Its not you know.

Velomark wrote:

Without a contact point with the drive train I cant see how this is going to be more accurate than the free strava guesstimate based on ones weight speed etc

Exactly. It'll possibly be marginally more accurate as it might be able to obtain better information regarding wind speed and direction