Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

My solution to “fix” phantom braking

This site may earn commission on affiliate links.
To me, this is the single biggest issue that I have with FSD and it’s the reason that I am so reluctant to use it when there are cars behind me. I hit the camera icon on the screen whenever there is an event, but how does the development team know what the event was and why I’m sending them data? There really is no meaningful way for the beta testers (us) to communicate feedback to the dev team. My solution is to put a dedicated phantom braking icon at the top of the screen that we would hit every time the car brakes for no reason. That way, the developers would get a mountain of data that they could analyze all related to this serious problem and hopefully get to the root cause. This approach could also apply to other complaints like improper lane selection. If a row of dedicated icons is not desirable, maybe refine the camera icon so that when selected, a drop down list appears and you can tick the box next to what the event was. Thoughts?
 
How would the car know its braking for no reason? If it knew there was no reason to brake, couldn't it just.... I dunno... Not brake?
The OP is proposing a quick way for the driver to report phantom braking events. I kind of like the idea, and it’s not just applicable to FSD Beta, since phantom braking can occur when using basic AP. Maybe even easier than a new icon would be a “phantom braking“ voice command to report events.
 
It sounds like some commenters are misunderstanding the OP. They’re not saying the car would know it was phantom braking. They’re saying to have a dedicated feedback button to communicate it was phantom braking, whether by a one-touch icon or by a checkbox of options when hitting the camera.

@RedTesY’s suggestion could actually work. It just doesn’t in its current design. Voice commands aren’t sent to Tesla. They’re just pulled by a Tesla tech when needing to troubleshoot or investigate a car issue when it’s in service or remotely. So why not pair that bug logging feature with the camera button for FSD beta users? It would be easier for someone like me than sending that follow up email to their email address (which I rarely remember to do later anymore). Workflow would be:

1) FSD beta issue to report, user taps camera button
2) User activates voice command to file a bug report, this time giving more detail for why the autopilot clip was saved
3) FSD beta team pulls any bug reports from the car within 5 mins (or whatever) of any saved autopilot clip (i.e. - camera button was tapped)
 
Not sure that multiple buttons are needed to report an issue like PB. When the report button on FSDb is pushed I’m pretty sure they can see a massive deceleration and G-force to understand what the driver is reporting. Unless it’s one of those “it’s horrible PB“ 2 mph changes which are mute in my view.
 
  • Like
Reactions: jbadger
Unfortunately that doesn't help. Bug reports aren't sent to Tesla. They are available for service techs to work on a problem you've opened a ticket for.
I've heard that off and on for years. But what about this? Any voice command only works if there's network connectivity. Which means the actual voice recognition happens on some Tesla server somewhere. So Tesla does have (if they want) those bug reports. Or I'm missing the obvious...
 
I've heard that off and on for years. But what about this? Any voice command only works if there's network connectivity. Which means the actual voice recognition happens on some Tesla server somewhere. So Tesla does have (if they want) those bug reports. Or I'm missing the obvious...
Yeah, but there's a difference between the car sending the audio snippet to some cloud voice recognition processor (and then immediately deleting it because there is no need to store every voice command that ever gets sent to it), and storing a bug report in the developers' bug queue. Two entirely different systems/use cases.
 
Maybe something like the Waze hazard reporting might be more efficient?

Press the cameras icon, present a sub-menu with 5 or 6 categories {i.e., PB, wrong lane, running stop signal, etc. ). Possibly even a 2nd sub-menu to further categorize the events. Like on Waze you can report it as a hazard, identify it as law enforcement, then specify visible, hidden, other side, etc.

Ideally leave the completion of the user input details to be added once the drive completes instead of forcing it while driving.
 
Yeah, but there's a difference between the car sending the audio snippet to some cloud voice recognition processor (and then immediately deleting it because there is no need to store every voice command that ever gets sent to it), and storing a bug report in the developers' bug queue. Two entirely different systems/use cases.
Of course. Just saying it's an opportunity where they could spin those off into some log if the S2T sees "bug report". Though of course it wouldn't be accompanied by any status data they might want...