Login required to started new threads

Login required to post replies

Prev Next
Re: Aerolab 2.0 [trail] [ In reply to ]
Quote | Reply
trail wrote:
So to bridge the gap between a specialized tool for nerds, and something that could take off as a more mainstream performance tool, I think requires considerate UI design as well as a the right kind of documentation.

I don't believe DIY aero testing is ever going to be easy or mainstream... just because controlled testing of this sort is inherently tedious and time consuming. There is a long checklist of things to consider and many ways to screw it up. The pool of people who are willing to pay attention to details and put in the time isn't very large.

But you certainly want to make it as easy to accomplish and precise as you can. Lots of room for improvement there.
Quote Reply
Re: Aerolab 2.0 [rruff] [ In reply to ]
Quote | Reply
I might be naive and that is not so bad, I have been called worse, but other than use for aero determination one could use it to optimize tire pressure for instance couldn't you? So if the people who might use it for this sort of thing were aware of best practices then maybe that would be a nice benefit? Or is the issue that the CRR change is so small that it is hard to see without really good precision which is back to what has been said numerous times above, this is a tool for the person wiling to invest time in knowing how to do it and being careful.
Quote Reply
Re: Aerolab 2.0 [s5100e] [ In reply to ]
Quote | Reply
s5100e wrote:
I might be naive and that is not so bad, I have been called worse, but other than use for aero determination one could use it to optimize tire pressure for instance couldn't you? So if the people who might use it for this sort of thing were aware of best practices then maybe that would be a nice benefit? Or is the issue that the CRR change is so small that it is hard to see without really good precision which is back to what has been said numerous times above, this is a tool for the person wiling to invest time in knowing how to do it and being careful.

You *can* use it for Crr testing but you have to be very (very very) careful. Tom A. first identified the "impedance breakpoint" using VE. For regular testing at below the breakpoint, there are easier ways to measure Crr -- that's why BRR uses a drum in a laboratory. I'm reasonably sure that we wouldn't understand the impedance breakpoint to the extent that we do except for field testing.

So, you're right: Crr change is small so it's hard to see without really good experimental control. VE is like any measuring tool--you still need to know how to use it to get precise measurements. No matter how good the instrument, you can still use it carelessly and get crap results.

What we're talking about is both how to simplify the tool, and also how to teach people to use the tool.
Quote Reply
Re: Aerolab 2.0 [specialist] [ In reply to ]
Quote | Reply
specialist wrote:
eg given the variation between laps within each setup, what is the probability that setup A is actually better than setup B, or how many laps would I need to get that over 80% or 95% or whatever.

This is great insight - and also one of the reasons why I like the work per lap bootstrap methodology that I implemented when doing these sorts of things. The WLB method is a combination of Haile/Chung nuggets over the years.

The insight that it needs to be clear (for the lowest common denominator and not tea leaf readers) when you have unreliable data is great!

Best of luck on the project guys!

=================
Kraig Willett
http://www.biketechreview.com - check out our reduced report pricing
=================
Quote Reply
Re: Aerolab 2.0 [rruff] [ In reply to ]
Quote | Reply
rruff wrote:

I don't believe DIY aero testing is ever going to be easy or mainstream... just because controlled testing of this sort is inherently tedious and time consuming. There is a long checklist of things to consider and many ways to screw it up. The pool of people who are willing to pay attention to details and put in the time isn't very large.

But you certainly want to make it as easy to accomplish and precise as you can. Lots of room for improvement there.

Will it ever be as easy as power meter usage ? Maybe, but there's a long way to go.

With time, various sensors will make it less time consuming and error prone.
Easier to use apps will make it easier to attempt.
Better algorithms to detect errors, flag or even correct them plays a huge part in this.

Maybe today aero testing is for 10% of TTers. Maybe that can get to 80%. It won't be an overnight thing but it will become easier.

Who would have thought they could put a medical grade ECG in an apple watch ?
Quote Reply
Re: Aerolab 2.0 [marcag] [ In reply to ]
Quote | Reply
marcag wrote:
Who would have thought they could put a medical grade ECG in an apple watch ?

But we had ECGs used by doctors for a very long time before they were consumer devices. First we need a device, protocol, and software that actually works well... for anyone.

-can resolve a <1% CdA difference for A vs B with a high degree of certainty, with several swaps (quick to do like a helmet or wheel).
-using a good venue and no traffic.
-on a typical day with wind and varying conditions.
-in a reasonable time (2hr max).

Then see about removing the restriction of swaps and same-day testing... and try to educate the masses and make the process idiot proof... and improve the precision and accuracy.

Eventually, if the ability to resolve elevation is good enough to not add much error, useful information could be obtained while "just riding along". But this is still not something you'd want to leave on your bike since the airspeed probe needs to be located where it isn't effected by the rider/bike. So, another problem to solve.
Quote Reply
Re: Aerolab 2.0 [liversedge] [ In reply to ]
Quote | Reply
 

Hey folks,

Lots of good discussion for Aerolab 2.0 and there is a long list of things we wish to do.

To help prioritize it and get a better feel for the target audience, could we ask you to fill a 1 minute questionnaire ?

Questionnaire

Comments at the bottom of the questionnaire are welcome
Quote Reply
Re: Aerolab 2.0 [marcag] [ In reply to ]
Quote | Reply
Have you considered building from existing autocross telemetry tools?

They may help inspire some usability/layout solutions, and really the applications are fairly similar.

https://www.mclaren.com/...-acquisition-system/

https://www.aim-sportline.com/en/sw-fw-download.htm

https://github.com/autosportlabs (open source)
Quote Reply
Re: Aerolab 2.0 [codygo] [ In reply to ]
Quote | Reply
thanks for that ! I will take a look
Quote Reply
Re: Aerolab 2.0 [RChung] [ In reply to ]
Quote | Reply
per R. Chung: The rho calculator currently uses "station" pressure.I just discovered that Environment Canada historical weather data uses station pressure and as well indicates the weather station elevation which is good to know when using the Rho calculator. These little tidbits are so helpful. Again Thanks R. Chung for all of your generous assistance.
Quote Reply
Re: Aerolab 2.0 [liversedge] [ In reply to ]
Quote | Reply
Only just saw that GCv3.6 is out, downloaded it but no updates to aerolab. Been no updates on this thread for ages either, has Aerolab2.0 died, shame if it has as were a lot of key features talked about...?? Thanks Mark.
Quote Reply
Re: Aerolab 2.0 [Fastasasloth] [ In reply to ]
Quote | Reply
We are working on it right now, which is kind of mad, your timing is spookily prescient.
  • At the mo its looking at basic functionality to support newer devices, so there will be an "AERO" tab on raw data which is where second by second samples for Aerolab will live; estimated series for crr, rho etc and measured series for crr, rho etc. The idea being estimated and measured data can coexist and will be consumed in functionality without having to worry about a zillion ways to get the data prepared. Quite which series we choose to "standardise" is being debated.

  • We also just added functions to parse and store FIT session and lap data since some devices that are limited by connect IQ only recording second by second samples compute aggregates that cannot be recreated from the data recorded (they are computed using higher resolution data that can only be recorded once a second).

  • Additionally just added metadata for intervals/laps/runs; it is now possible to configure and record metadata for each run; so helmet, bike, shoes, position etc as well as notes and so on.

  • Next we need to add functions for creating "runs" that might be comprised of multiple laps etc. This might end up being just be another interval type. But will likely need to be able to fine tune things and record and import info like temp, air density measures.


After that it gets more ambiguous, but working on defining the sorts of analytic or application functionality people want. If anyone has thoughts they are welcome to comment here.

Mark

PS: reading back through this thread has been really useful too- thanks for the inadvertent reminder :)
Last edited by: liversedge: Nov 27, 23 11:46
Quote Reply
Re: Aerolab 2.0 [liversedge] [ In reply to ]
Quote | Reply
Great to read that work is still ongoing. When might we see a development build that incorporates (some of) the new features?

thanks again

Mark.
Quote Reply
Re: Aerolab 2.0 [Fastasasloth] [ In reply to ]
Quote | Reply
We usually do some heavy lifting development during the Christmas break- so would expect around the new year.
Quote Reply

Prev Next