AeroPod Missing Lap Markers

Post Reply
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

AeroPod Missing Lap Markers

Post by virtualdynamo »

I did several profile 4 tests yesterday. The AeroPod failed to catch one of the lap markers. In this Strava activity, the missing mark is between laps 6 and 7.
https://www.strava.com/activities/6368398286/laps
Attached is the IBR of concern.

This is not the first time the AeroPod has missed lap markers.
Attachments
Velocomp_12_10_2021_1440_12_km.ibr
(32.17 KiB) Downloaded 127 times
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

Oddly, while the IBR doesn't delimit on the lap marker, the exported CSV does. See row 62 at 2021-12-10T20:42:56Z
Attachments
Velocomp_12_10_2021_1440_12_km.csv
(47.76 KiB) Downloaded 120 times
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: AeroPod Missing Lap Markers

Post by Velocomp »

You only posted one .ibr ride file. You should have a whole bunch of them; please post all of them.
John Hamann
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

Here are the other IBR files.
Attachments
Velocomp_12_10_2021_1423_5_km.ibr
(27.85 KiB) Downloaded 120 times
Velocomp_12_10_2021_1416_2_km.ibr
(26.55 KiB) Downloaded 122 times
Velocomp_12_10_2021_1410_2_km.ibr
(28.2 KiB) Downloaded 126 times
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

Well gosh. I'm limited to 3 files at a time. Here's 3 more.
Attachments
Velocomp_12_10_2021_1437_9_km.ibr
(16.02 KiB) Downloaded 125 times
Velocomp_12_10_2021_1435_8_km.ibr
(15.73 KiB) Downloaded 128 times
Velocomp_12_10_2021_1429_7_km.ibr
(27.87 KiB) Downloaded 126 times
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

And here's the last 3.
Attachments
Velocomp_12_10_2021_1459_33_km.ibr
(199.92 KiB) Downloaded 120 times
Velocomp_12_10_2021_1454_17_km.ibr
(26.82 KiB) Downloaded 121 times
Velocomp_12_10_2021_1448_15_km.ibr
(27.24 KiB) Downloaded 133 times
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

Since I've submitted all the IBR files, I'll mention another error I've found. There are anomalies in the Timestamp.
In 1423_5, the time stamp jumps from 2021-12-10T20:29:15Z at 337 seconds to 2021-12-10T20:54:24Z at 338 seconds.
In 1448_15, the time stamp jumps from 2021-12-10T20:52:50Z at 69 seconds to 2021-12-10T21:50:01Z at 69 seconds.
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: AeroPod Missing Lap Markers

Post by Velocomp »

This is based on the first 3 ride files you sent; I will look at the others now.

Lots of things to comment on:

0) I combined your separate ride files into one file that covers the entire ride. I used the command "File/Combine-Split" to do this. Start with the earliest time file first, then add subsequent files, in time order. This results in a composite ride file with 6 laps (attached).

1) It appears that you are riding in conditions of large headwinds and tail winds. GPS data shows you're doing your riding in the shadow of Arrowhead Stadium; this may be responsible for the gusty wind conditions (you have strong tail winds when closest to the stadium; strong headwinds when away).You are not going to get good CdA measurements when you do this (the instructions say so). Wait for a day when wind conditions are much calmer, or find a place where buildings won't cause significant wind changes.
Screen Shot 2021-12-12 at 10.40.03 AM.png
Screen Shot 2021-12-12 at 10.40.03 AM.png (458.14 KiB) Viewed 4440 times
2) Lap 1 did not start and stop in the same physical location. So, its on-screen CdA lap data is not good, and furthermore, it means your second lap on-screen CdA data won't be good, either. As the instructions say, when using profile 4, during testing it's CRITICAL that the lap button be clicked at the same physical location.

3) In lap 2, DFPM and cadence readings dropped out between minute 8 and 11. You slowed down and apparently stopped just after the power dropout began. There is a 26 second time gap between stopping and starting. You slowed down again around minute 10:28. There are some intermittent speed readings at the second slow-down but no time gap. It's impossible to tell what caused these issues, but make sure your speed sensor is functioning properly (speed sensor problems would not cause DFPM/cadence problems).

4) The start and stop places of lap 2 are different; therefore, lap 2 data is bogus.

5) When a given lap's start-stop place is different, the following lap's on-screen CdA will be bad, too. In your case, this is lap 3 (minute 11-17).

6) Lap 3 starts and stops in the same place; this is what the instructions require for profile 4. However, the test route is full of hills and valleys (+/- 4%), causing your power output to be highly variable (100W-300W). The instructions say to test in a place where you can keep power output reasonably constant. Therefore, this is not the best route for testing. Lap CdA for lap 3 was 0.381 (HR= 211+180)

7) It looks like you rode off the route for lap 4, then back, for another test for lap 5. This is OK (very short laps are ignored by AP for CdA calculations)

8) Lap 5 duplicates the route of lap 3. Lap 5 CdA was .421.

9) Lap 3/5 are in the ballpark of each other, but wind and power variability are excessive.
Attachments
Toiga_Combined_12_10_2021_1410_10_km.ibr
(97.76 KiB) Downloaded 132 times
John Hamann
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: AeroPod Missing Lap Markers

Post by Velocomp »

I combined all your additional ride files. I deleted those that were post-testing.

For your multiple laps from mile 2.5-7.3, CdA is about .335

For multiple laps from mile 7.3 to 11.7, CdA was about .386

I don't know what changed between these tests.

The overall comment still applies; wind, slope and power variability conditions are not favorable for these tests.
Attachments
ArrowheadTest_Combined_12_10_2021_1410_21_km.ibr
(194.71 KiB) Downloaded 127 times
Screen Shot 2021-12-12 at 11.04.17 AM.png
Screen Shot 2021-12-12 at 11.04.17 AM.png (297.33 KiB) Viewed 4438 times
Screen Shot 2021-12-12 at 11.01.45 AM.png
Screen Shot 2021-12-12 at 11.01.45 AM.png (331.01 KiB) Viewed 4438 times
John Hamann
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

I'm well aware of issues with my "lab" and can understand how it's an overriding distraction, but that has nothing to do with this thread.

The only comment germane to my inquiry is the claim that my laps "did not start and stop in the same physical location." In point of fact, I couldn't have been more consistent in marking my laps. The most my lap marks deviated from the average is 2.6 meters. The slowest I was going at the lap marker point was 5.3 meters/second. With data recording once per second, it's not possible to mark the laps with any more consistency.

My calculations are based on using the FIT file from my GPS. That FIT file, the CSV created by Garmin's FitToCsv, and the Excel spreadsheet of my analysis are in the attached ZIP.

As to why one might think that I did otherwise, there are issues with Isaac's merge tool. I opine that it's challenged to synchronize properly with repeated riding of loops. I'm not aware of any way for the user to intervene in the synchronization.
Attachments
2021-12-10-12-43-27.zip
(3.59 MiB) Downloaded 123 times
2021-12-10-12-43-27.png
2021-12-10-12-43-27.png (210.83 KiB) Viewed 4433 times
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: AeroPod Missing Lap Markers

Post by Velocomp »

I guess I missed the point of your thread.

If you are marking your laps as you say that's great. That's exactly what AP needs to do its job of CdA measurement; in fact, I prefer manual clicking to the auto-lap feature in Garmin.

I have occasionally seen AP not "pick up" one of my lap marker clicks. How do I know that? When AP receives a lap message, the AeroPod CdA app shows a screen for 10 seconds with the last-lap information. If I don't see the last-lap screen, then AP hasn't received the lap message, and I click the lap button again.

I count laps from the combined file I posted earlier.

Your laps are consistent, except laps 1 and 2. Lap 1 appears to be where you started your ride, perhaps from the parking lot, to the loop. Lap 2 had problems with dropouts, so it's not surprising if its location data is not properly synchronized.

The other laps appear to be perfect, though I note that in some of them there are two rides around the circuit before the lap button is pushed (e.g. lap 4), and in other laps the lap button appears to be pushed after 1 ride around the circuit (e.g. lap 5).

One circuit of the route is 0.76 miles; this is a bit short for getting sufficient CdA data.

Since the purpose of AP is CdA measurement, I tailored my remarks about issues that would affect accuracy.
John Hamann
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

1. "When AP receives a lap message, the AeroPod CdA app shows a screen for 10 seconds with the last-lap information."
a. But in my case, the AP is receiving a lap message. How else could it appear in the CSV export of the IBR?
b. By "the AeroPod CdA app shows a screen for 10 seconds with the last-lap information."? do you really mean "the AeroPod Connect IQ CdA app displays the last-lap information for 10 seconds."? If so, am I supposed to watch for the values to jump from live to the frozen last-lap AND press the lap button at the same CRITICAL physical location? If not, please advise.
2. "If I don't see the last-lap screen, then AP hasn't received the lap message, and I click the lap button again." Isn't this contrary to, "As the instructions say, when using profile 4, during testing it's CRITICAL that the lap button be clicked at the same physical location."?
(3. "Lap 1 appears to be where you started your ride, perhaps from the parking lot, to the loop." Lap 1 is the ride from the point where the AeroPod indicated 70W during the calibration, to indicating 100W, and then to the start of testing.
4. "One circuit of the route is 0.76 miles; this is a bit short for getting sufficient CdA data." For the record, a given test is intended to be 2 circuits of the route. Indeed, on my 3rd "test", I lost track and had an intermediate lap press. I subsequently revised my secondary GPS to display lap distance since I have trouble counting to 2.)

Under the heading creating more questions than answers, note:
There are timestamp errors.
There is an issue with merge synchronization.
"Lap 2 had problems with dropouts," See: AeroPod Missing Data Acquisition
https://velocompforum.com/viewtopic.php?f=98&t=5826
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: AeroPod Missing Lap Markers

Post by Velocomp »

virtualdynamo wrote: Sun Dec 12, 2021 1:42 pm 1. "When AP receives a lap message, the AeroPod CdA app shows a screen for 10 seconds with the last-lap information."
a. But in my case, the AP is receiving a lap message. How else could it appear in the CSV export of the IBR?
b. By "the AeroPod CdA app shows a screen for 10 seconds with the last-lap information."? do you really mean "the AeroPod Connect IQ CdA app displays the last-lap information for 10 seconds."? If so, am I supposed to watch for the values to jump from live to the frozen last-lap AND press the lap button at the same CRITICAL physical location? If not, please advise.
2. "If I don't see the last-lap screen, then AP hasn't received the lap message, and I click the lap button again." Isn't this contrary to, "As the instructions say, when using profile 4, during testing it's CRITICAL that the lap button be clicked at the same physical location."?
(3. "Lap 1 appears to be where you started your ride, perhaps from the parking lot, to the loop." Lap 1 is the ride from the point where the AeroPod indicated 70W during the calibration, to indicating 100W, and then to the start of testing.
4. "One circuit of the route is 0.76 miles; this is a bit short for getting sufficient CdA data." For the record, a given test is intended to be 2 circuits of the route. Indeed, on my 3rd "test", I lost track and had an intermediate lap press. I subsequently revised my secondary GPS to display lap distance since I have trouble counting to 2.)

Under the heading creating more questions than answers, note:
There are timestamp errors.
There is an issue with merge synchronization.
"Lap 2 had problems with dropouts," See: AeroPod Missing Data Acquisition
https://velocompforum.com/viewtopic.php?f=98&t=5826
1) The .ibr ride file from 12/9, referenced in your earlier post, does NOT show any cadence dropouts. In fact, it appears to be a different file from the one you generated your graphs.

2) Your 12/9 ride file was recorded in profile 3. Profile 3 behaves completely differently from your 12/10 files, which were recorded in profile 4

3) In profile 3, you'll see lap markers recorded whenever the lap button is touched

4) In profile 4, whenever the lap button is hit, a new ride file is started about 10 seconds after the end-of-lap button is hit. There are 8 ride files for your 12/10 ride, and I don't see gaps in your data recording between laps, so it appears the profile 4 lap feature is working properly.

5) In profile 4 you should see lap markers only at the beginning and end of your ride files; you should not see lap markers in between.

6) As I reported previously, there appear to be some places on your 12/10 ride (1416 ride file) where DFPM power/cadence dropped to zero. I don't know why this happened. It looks like you stopped when cadence/power first went to zero; there is a 26 sec time delay between stopping and starting at the .55 mile mark. You started riding again and, about 2 minutes later, intermittent DFPM data started appearing. You slowed down to a near stop soon thereafter (it does not look like you actually stopped) and cadence then re-engaged. After this, none of your other ride files repeat this problem.

7) AP includes ride file pause feature which will affect time stamps. This does not indicate that something is wrong. Discrepancies between Garmin and AP timestamps can simply be a result of how each device records ride files.

8) You don't need to do a new cal ride prior to every test session.

In summary, apart from some unexplained behavior in your 1416 ride file I don't see anything wrong with data recording. There are significant issues with your ride route, which has been discussed in a previous post.
John Hamann
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

Indeed, I erred on my attachment for the AeroPod Missing Data Acquisition thread. I apologize and have addressed the matter in that thread.

As the purpose for this thread is only about the AP missing lap marks, I'll do my part to stick to that matter henceforth.

The issue of this thread is that in the original 1440_12 file I attached, I pressed the lap button, the AP knows it, and yet didn't create a new IBR.
Velocomp
Velocomp CEO
Posts: 7793
Joined: Fri Jan 11, 2008 8:43 am

Re: AeroPod Missing Lap Markers

Post by Velocomp »

There are two kinds of Aero test routes:

1) Closed Loops (click lap to start test, ride around in a closed loop, finish loop, click lap to finish loop, continue riding around another loop, etc)

Closed loop files have two lap markers in the .ibr ride file: start test and end test

2) Out-and-back rides (click lap to start test, ride out to half-way point, turn around, ride back to start, click lap to finish test, turn around, ride back to start, click lap to start another test)

Out and back files have 3 lap markers in the ride file: loop-around marker, start test, and end test.

In an out-and-back test, the turnaround marker and associated data during the loop-back is ignored in CdA calculations for the next test. However, the turnaround data is recorded in the same ride file as the subsequent test (data is located at the very start of the test)

Your test route is a closed loop, EXCEPT for two particular tests:

1) in your 1437 test, at the end of the test you did not continue riding the test loop; rather, you circled around in the adjacent parking lot to return to the start of the loop test. You can see in this screen shot where you started circling around to get back to your start point
1437 turnaround
1437 turnaround
Screen Shot 2021-12-13 at 9.34.47 AM.png (296.8 KiB) Viewed 4390 times
2) once you circled around back to the beginning of the test loop, you clicked the lap button to begin your 1440 test. You can see the rest of your circle-back in this screen shot
1440 turnaround
1440 turnaround
Screen Shot 2021-12-13 at 9.31.59 AM.png (496.88 KiB) Viewed 4390 times
3) You did your 1440 test

You'll find that your 1440 test has three lap markers, instead of the two markers of all your other ride files.
1440 lap markers
1440 lap markers
Screen Shot 2021-12-13 at 9.48.41 AM.png (227.56 KiB) Viewed 4390 times

Effectively, AP considered your 1440 test as the same form as an out-and-back test, and simply ignored the data of the first 0.21 miles of it in the first lap (it is the turnaround data).

There is nothing wrong with the way you conducted your test; AP simply figured out how you were doing the test, and adjusted its CdA measurements accordingly.

So, the lap button is working properly.
Attachments
1440 loop test
1440 loop test
Screen Shot 2021-12-13 at 9.33.35 AM.png (348.15 KiB) Viewed 4390 times
John Hamann
virtualdynamo
Posts: 25
Joined: Sat Nov 20, 2021 7:36 am

Re: AeroPod Missing Lap Markers

Post by virtualdynamo »

Thank you. That priceless piece of intrinsic knowledge is just what I needed to know. i.e.:
Closed loop files have two lap markers in the .ibr ride file: start test and end test
Out and back files have 3 lap markers in the ride file: loop-around marker, start test, and end test.
Post Reply