Aero Sensor for Dummies Version II

Triathlon Forum
1

Seems that the original Aero Sensors for Dummies thread was lost in the migration to the new Forum, so here is a new place to discuss Aero Sensors.

I just received a new Gibli and in the coming months will be playing with and testing things. I have tested in wind tunnels three times and have done a ton of testing in indoor velodromes across the US, and in Mexico and Bolivia. I have never gotten satisfactory results in wind tunnel. Every time I have went, numbers would start to look out of line and I would go back to baseline and could not replicate baseline. At that point, you donā€™t know if any of the numbers are correct. In indoor velodrome testing, where all variables can be controlled, its a math problem. I have had excellent results in velodromes but renting them and get access is tough and expensive. Iā€™m real excited about testing anytime I want and honing in on those small gains.

20250714_173130
20250714_1731301920Ɨ2560 846 KB

1 Like
2

Here is a good video from Jim on the Gibli and why he is using it and a new service from him.

3

Can we get a run down of aero sensors for not dummies?
what is the state of the art sensor suite, analysis software and test procedures? Is there documentation somewhere?

Which wind tunnels offer cyclist testing?

4

I research three sensors; Gibli, Aerosensor and Notio before my purchase. It seemed to me Gibli had more tech in it and had a more user friendly app. Each sensor has its on software and test procedures. I have never use the Notio or Aerosensor and I am waiting for a mounting system (it doesnā€™t come with one) for the Gibli. Once I get it out and test it, I will give feedback on it.

A2 windtunnel in NC does a lot of Cyclist testing. In the past, someone would organize a Tunnel rental then sell off time slots to people who wanted to test and assist in the testing procedures. One of the people who has organized this in the past is now using an aero/CdA senor for testing people.

5

Background - I am a beta user of @marcag ā€œgizmoliteā€ device.

I started with chung testing, but never really found a venue or conditions that allowed any kind of consistent results to meaningful accuracy. Very rare to have no wind and venues with nice road surface and low car volume where i live.

Then over time I got to know and became friends with Marc online via ST (note consider this a disclaimer that the rest of my feedback cant be considered fully independent)

We collaborated on a very DIY ā€œsensorā€ hack - basically bluetooth weather sensor data feeding data to a garmin IQ app not disimilar to cdacrr app approach. On a good day it could produce good results - quite an incredible feat for what it was but seemed to struggle in some conditions, particularly more windy or situations with more yaw so things were inconsistent and could only ever reach a fairly high margin for error.

At the same time Marc started to develop the gizmolite and I graduated to being a beta user on that device and have gone through a few iterations and the accompanying app. I think me being in the opposite hemisphere gave him some data during long cold canadian winters :slight_smile:

I do most of my testing at an outdoor velodrome. The surface is in poor condition and really slow but overall consider myself fortunate to have access to it within 20 minutes of home. I have a few other venues but all have their own idiosynchrosies and overall seem to get best fidelity data at the velodrome.

Experience with device
All i can say is that testing with a reliable sensor vs nonsensor chung testing or DIY device to get actionable data has been like night and day. Id describe it more like probabilities of getting return on testing effort. With non sensor chung data it felt like i might get good data 1 out 10 outings, with the DIY sensor maybe 3 in 10 days. Now its like 8 or 9 out of 10 days. Bad days or data now seem to be down to protocol issues not data measurement quality or device limitation (e.g. not warming up long enough to get settle in to true cda or leaving the sensor out in the sun between runs causing bad temperature readings)

There are other practical things that have helped imrpove outcomes. Better mounting solutions and handlebars with good mount options have helped inprove calibration factor consistency in my testing etc.

App, and automated calcs have been straightforward to use, connect all sensors and get good results. I still like analsing data in Golden Cheetah post tests particularly for velodrome. For out and back testing app is much much easier as bo need to ā€œsnipā€ out the turnarounds

Would love to see some of the tools that Marc has built for himself in Golden Cheetah become public - my process flow to get data into GC is clunky.

Firmly believe when the device comes to market it will shake things up particularly if it can be delivered at more reasonable price point than the existing sensor market which i consider quite crazy and firmly believe will have a major recalibratuon in future

Its been a really satisfyning journey over the last few years, immensely enriched wuth Marc sharing knowledge and ideas - we are lucky to have him and others engaging here on ST. For anybody that is interested and prepared to invest the time id be a strong advocate of getting a sensor simply because of the enjoyment factor let alone the chance to optimise speed.

Happy to answer any questions on my experience or provide example data.

3 Likes
6

@marcag how far off are you from production of the Gizmo?
Iā€™m in the market for a sensor and Gibli and Aerosensor are front runners at the moment.
I did own a Notio but found it buggy and eventually moved it on
I did get some good Data from it on helmet and skinsuit choice but donā€™t think they were quite there with a finished product and app.

7

There is a batch going out in Sept as part of a broader Beta. I can definitely include you. Just DM me if you want to give it a whirl

8

Thanks Ryan,

Your testing, feedback and patience have been very much appreciated.

Just to clarify for the others, there are 2 sensors. The first ā€œthe Gizmoā€ is intended for aero testers. This has been available to testers for a while. The second, ā€œGizmoLiteā€, is intended for home use for people that want to self test.

The GizmoLite is as accurate as anything out there (in our opinion). And yes, I have tested them all :-). But the plan all along has been to make this affordable. Much, much more affordable than what the market has become. When we did Notio, the plan was for these things to get cheaper, not triple in price. But itā€™s a ā€œweird marketā€, as the little traffic on this thread shows.

1 Like
9

I would love to give it a whirl as well

I have a great place for testing

And many hours to do the testing

10

lmk if you have space for another beta tester. I have a great outdoor loop for testing and a lot of equipment & positions I need to test.

11

I still use the wind tunnel as well. If you watched Oside 70.3, the fastest bike split and leader into T2, Justin Riele was in the A2 wind tunnel in January with me.

FWIW Iā€™ve been using the Gizmo now almost 3 years and have >1000 hours of testing using them on well over 100 people. Marc and I did a 3d Aerocamp type testing session here in Tucson this past March.

So far this year my aero testing clients have won ~ 6 pro races iirc

1 Like
12

Late to the party but would love to help test @marcag . Iā€™ve been doing aero testing the weatherflow and the CDACRR app, and separately with MyWindSock

13

Hi Marc

This is very interesting. Can you give an update the Gizmo (Lite), and maybe explain how it differs from the Notio? :slight_smile:

If you still need beta testers I am very interested. Have done some aerotesting using the Notio and VE method in Goldencheetah to validate.

Best,
Morten

14

From a hardware perspective, there are a few things. We do yaw, which is one thing the Notio lacked. When we did Notio (2016) yaw was a large incremental cost. Today itā€™s peanuts. We also do a lot more with IMUs, which helps elevation/inclination a lot as well as vibration. We also have connectivity to external sensors used for monitoring people in training. When I left Notio they have a roadmap for 2.0. Too bad they didnā€™t do it.

I would say the software is orders of magnitude easier and there is some really cool stuff planned.

From a cost perspective we will come in less than Notio, which is already way less expensive than the competition. A big, big goal is to make this things more accessible, from a price and usability perspective

We were trying to do a broader beta in the fall. ā€œOther $hitā€, ie work for World Tour team and Ironman happened just as we were hit with a memory chip shortage which is now ironed out (I hope). So we pushed things to the spring (April ish). I just need to keep the pesky distractions away. The device is ready but preparing for support is something I over worry about

We wonā€™t release unless itā€™s easy to use. The problem with these devices is it doesnā€™t take much to get wonky numbers. There is one device out there that flags wonky numbers so doing the ā€œright thingā€ make it more complex. Another approach is just sweep all data under the cover, throw out a number and hope people donā€™t notice. Seems easy, but would you trust it ? Finding the balance is tough.

We plan to do a few things that are unique to help the beginner.

My plan was to msg back people that msgā€™d me to be part of the beta.

6 Likes
15

Everything you write makes great sense!

For the Notio the yaw is also what I learned after purchasing it. Trying to overcome the yaw ā€˜issueā€™ by testing on outdoor velodrome as much as possible. If that is the right ā€˜tacticā€™ I do not know completely. But at least that makes sure there is no traffic.

I will look forward to hearing from you again, and to follow the Gizmo in general! :slight_smile:

Best,

Morten

16

Knowing yaw is useful especially when conditions are changing and to compare things day to day. When we did the recent 20m drafting tests, yaw varied slowly as the day progressed. Yaw seems to vary more in Arizona than in Montreal. Maybe it is more location specific than I know.

Practically, where I live and do a lot of testing, yaw usually does not vary that much. Sometimes it does but rarely does it skew results. Day to day comparison are very useful but harder unless you know what you are doing and have the right software. So sometimes across 2 days, I will get a ā€œbetter numberā€ for the same configuration. I check, and the software can correct a few things and if they still donā€™t line up, I look at yaw and it can be the culprit.

But letā€™s say you are doing inter session comparisons rather than cross sessions. When I would test with my Notio ā€œinter dayā€, I would just repeat my baseline test every so often and if you are getting consistent results, you can assume yaw has not changed much.

Example : you are doing 12 test runs, and test run 1 is with helmet A, just run helmet A at test 4 and 8 for example, or even 6 and 12.

One thing I think would be fun to do is a seminar on how to aero test. I am thinking about this for April. Maybe a session in North America and one in Euope

4 Likes
17

Good point on checking the yaw angle by ā€˜re-testingā€™ the baseline often.

Something I should probably spend more time on doing.

Could be interesting with such seminar or webinar!

18

I struggle to properly assess how much interest there is for this. Iā€™d love to do it, because in case you didnā€™t notice I love this stuff :slight_smile:

Tell you all what. If there would be interest, just like this post. If we get more than just us two :joy: maybe we do it .

9 Likes
19

I have noticed :smiley:

Pretty sure we can get people in the Cycling Aero-Sensors and Aero Testing group on Facebook interested as well :smiley:

Personally, also fine with just me :innocent: :joy: But might be more worth it with others as well :smiley:

1 Like
20

I think this is because as things heat up and cool down you get more wind here. Plus if testing along the washes I can often find wind speed differences on my kestrel by hanging onto the railing leaning out into the wash, then at the path side of the railing and then again on the far side of the path. 3 different wind speeds within a 10m distance.

Hereā€™s an idea. since you and Iā€™ve talked about doing another aero camp here in Tucson, why not combine it with a how to aero test while we do the aero camp?