Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

LKAS Fault #30571

Closed
Luft1950 opened this issue Dec 2, 2023 · 20 comments
Closed

LKAS Fault #30571

Luft1950 opened this issue Dec 2, 2023 · 20 comments
Labels
bug car vehicle-specific

Comments

@Luft1950
Copy link

Luft1950 commented Dec 2, 2023

Describe the bug

After the release of OpenPilot 0.9.5 I have been getting daily LKAS Faults on my 2023 Bolt EUV.

Which car does this affect?

2023 Bolt EUV

Provide a route where the issue occurs

I-5 Olympia to Highway 512 East and then to Puyallup.

openpilot version

0.9.5

Additional info

0.9.4 worked okay but it seems to occur for no reason on 0.9.5. Very unchill.

@Luft1950 Luft1950 added bug car vehicle-specific labels Dec 2, 2023
@adeebshihadeh
Copy link
Contributor

What's your dongle ID? Do you have a route where this happens?

@Luft1950
Copy link
Author

Luft1950 commented Dec 2, 2023 via email

@adeebshihadeh
Copy link
Contributor

adeebshihadeh commented Dec 2, 2023

Thanks, I'm looking into it. I found five events, but they're only from your car:

CHEVROLET BOLT EUV 2022                  c93a84c2e28979fe|2023-11-25--13-28-19--55
CHEVROLET BOLT EUV 2022                  c93a84c2e28979fe|2023-11-27--16-06-53--22
CHEVROLET BOLT EUV 2022                  c93a84c2e28979fe|2023-11-28--11-46-47--46
CHEVROLET BOLT EUV 2022                  c93a84c2e28979fe|2023-11-30--12-24-03--45
CHEVROLET BOLT EUV 2022                  c93a84c2e28979fe|2023-12-01--12-05-38--36

@sshane
Copy link
Contributor

sshane commented Dec 7, 2023

In the first case, right before the LKAS fault, I saw that an LKAS message/counter was missing on the 128 read back bus, so openpilot correctly re-sent the same counter, then it immediately faulted. Usually the EPS is robust to sending at a slower rate, it looked like the message might have actually got sent, but not added to bus 128 (thus duplicated counter)? pandaStates indicates no dropped messages though.

image

@Luft1950
Copy link
Author

Luft1950 commented Dec 7, 2023 via email

@nharada1
Copy link

For what it's worth I see the same issue on my 2022 Bolt EUV

@siddchou
Copy link

siddchou commented Feb 1, 2024

I am also getting similar issue in Mazda 2024 every time during my 1st trip each day .

e548ddc194075bfb|2024-01-31--16-12-06--0

@sshane
Copy link
Contributor

sshane commented Mar 20, 2024

@Luft1950 does this still happen?

@Luft1950
Copy link
Author

Luft1950 commented Mar 20, 2024 via email

@Luft1950
Copy link
Author

Luft1950 commented Mar 29, 2024 via email

@Luft1950
Copy link
Author

Luft1950 commented Apr 8, 2024 via email

@nharada1
Copy link

nharada1 commented Apr 8, 2024

I ended up simply returning my Comma, this issue was too prevalent to safely drive.

@sshane
Copy link
Contributor

sshane commented Apr 8, 2024

@Luft1950 I requested upload, please put the device on WiFi for 30 minutes to let the logs upload.

@Luft1950
Copy link
Author

Luft1950 commented Apr 9, 2024 via email

@sshane
Copy link
Contributor

sshane commented Apr 9, 2024

Looks like there is a problem with the panda reporting that the message was delivered. openpilot then re-sends the same counter as it thinks it wasn't delivered, but since it faulted here and there's no blocked or lost tx messages supposedly it did actually get sent.

image

@Luft1950
Copy link
Author

Luft1950 commented Apr 28, 2024 via email

@Luft1950
Copy link
Author

Luft1950 commented Apr 28, 2024 via email

@Luft1950
Copy link
Author

Luft1950 commented Apr 30, 2024 via email

@sshane
Copy link
Contributor

sshane commented Apr 30, 2024

Can you go in connect and send the link for each?

@sshane
Copy link
Contributor

sshane commented May 1, 2024

@Luft1950 You're the only one hitting this old issue. Currently this appears to be a software bug, but we don't currently have the time to root cause your specific case. We are going to allow you to return your device if you would like, just send this GitHub thread: https://shop.comma.ai/a/returns

Related to #28899, the CAN packet dropped out for a few frames where we were expecting the loopback bus to say that the LKAS message was transmitted, so we transmitted it again incorrectly.

image

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug car vehicle-specific
Projects
None yet
Development

No branches or pull requests

5 participants