I have written up one of the weirdest GPS / GNSS related bugs that I've seen to date.

COROS have a nasty bug which results in position and speed data being out of sync with each other, sometimes offset by several minutes. I've seen this boost a friends alphas by 10 knots, during multiple sessions!

This issue should not be confused with the inability of COROS to calculate alphas correctly within their watch / app. This is solely an issue with the files being exported from the COROS app.

I'm just raising it here so that people might recognise the issue if it occurs in their data / friends data. It seems relatively rare, but I've seen this issue evident in tracks from the VERTIX 2 and APEX 2 Pro.

Full description at logiqx.github.io/gps-details/devices/coros/alpha/

Make sure you report these issues to Coros directly at -> https://support.coros.com/hc/en-us 

Hello Michael, do you have any feedback from Coros regarding this Alpha issue?  Just example, what I had today, after nice session, my  Coros watch and app shows Alpha500 48,59kmh, and after uploading 39.33kmh! This just creates distrust for the Coros watch.. 

Hello Michael, do you have any feedback from Coros regarding this Alpha issue?  Just example, what I had today, after nice session, my  Coros watch and app shows Alpha500 48,59kmh, and after uploading 39.33kmh! This just creates distrust for the Coros watch.. 

Hi Martin.

There are actually two distinct issues that relate to COROS watches and alphas.

The first is that they don't actually calculate alphas properly. So far as I can tell, they ignore the Doppler speeds and have implemented their own algorithm that doesn't work correctly. This is the issue you see on your watch and on your phone when your alpha is reporting higher than the GP3S website. I published a working algorithm and example Python code a couple of years ago, but COROS have not adopted it. My algorrithm has been used in Motion devices for at least a couple of years.

The second issue is related to how COROS watches are combining the NMEA data of the Airoha chip, resulting in positions and speeds that do not correspond to each other but with the same timestamp. This is the issue I have detailed in the link above and it causes third party software such as GP3S / GPS Results / GpsarPro / GPS Speedreader to generate incorrect alphas, often very much too high on GP3S. This is a frankly ridiculous data issue and full details were provided to COROS in April 2023.

There has been zero feedback for specific issues that I've reported to COROS.

Apparently the second issue described above (causing third party software such as GP3S / GPS Results / GpsarPro / GPS Speedreader to generate incorrect alphas) will be fixed in the firmware release of July 2024.

Fingers crossed it works...