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

FSD Beta Button in the UK?

This site may earn commission on affiliate links.
watching the beta videos, I can’t help thinking that there is a need for a pair of extra cameras on the front wings.
as the car creeps out of a turning into traffic, in some situations the necessary field of view isn’t always covered until the car is too far out.
hope I’m wrong!
Even when it sees it it's pulling out onto oncoming traffic and requiring intervention (see kim paquettes latest tests). Very worrying for something that Elon is claiming will be in wide release in a couple of weeks.
 
The shame is that side facing ‘nose’ cameras on the front wings would have nailed the issue of creeping forward at junctions. they would have line of sight far beyond our own.
I imagine at some point Tesla will redesign the wing camera housing to have two cameras that point behind (as currently) and outwards, replacing the B-pillar ones (which already suffer from condensation as an added "bonus").

Whilst I imagine the B pillar cameras are reasonably effective, they are further back than most people's heads are when driving, and when at a junction you can move your head forwards for even more visibility. A wide FOV can only compensate so much.
 
the B pillar cameras are reasonably effective
It seems so obvious to have side facing cameras as far forward as possible - especially since you have the rear facing side repeaters - that I can only think there was some good reason they didn't go that route. If the b-pillar condensation issue was evident from the outset, then you'd think the reason(s) for not having forward mounted side facing cameras must also have been significant.
 
It seems so obvious to have side facing cameras as far forward as possible - especially since you have the rear facing side repeaters - that I can only think there was some good reason they didn't go that route. If the b-pillar condensation issue was evident from the outset, then you'd think the reason(s) for not having forward mounted side facing cameras must also have been significant.
The only thing that makes sense in my head was that they didn't want to have potentially significantly expensive parts on the outside of the car, which could potentially get damaged in the event of an accident, or stolen, or vandalised, etc.

I don't know how much the current wing cameras are, but if you had two cameras in one unit then it would be pretty expensive I think.

I had a ranger out late last year for condensation in my B-pillar cameras, as it basically meant that every journey I made I was getting warnings the whole way. He took one off the car, checked for the Goretex patch, put it back on and basically said for me to come back to them again if it persisted. As soon as it got warmer the problem went away, but it does seem to mean that for several months in a year the B-pillar cameras are degraded, which makes them unlikely to be much use for creeping out at junctions either!
 
Will be capable to manage this kind of situations?
 

Attachments

  • Screenshot_20210914-115157_Gallery.jpg
    Screenshot_20210914-115157_Gallery.jpg
    150.4 KB · Views: 40
  • 20210914_100455.jpg
    20210914_100455.jpg
    249.7 KB · Views: 37
I imagine at some point Tesla will redesign the wing camera housing to have two cameras that point behind (as currently) and outwards, replacing the B-pillar ones (which already suffer from condensation as an added "bonus").
but this would introduce a new problem, as the camera would be much lower than the current B pillar camera - as such, if you have traffic to the side of you at a junction then the camera would not be able to see over it.

So it would work in some situations but be potentially worse than the current B Pillar camera in others.

+1 on condensation fix requirement though.
 
  • Like
Reactions: Durzel
but this would introduce a new problem, as the camera would be much lower than the current B pillar camera - as such, if you have traffic to the side of you at a junction then the camera would not be able to see over it.

So it would work in some situations but be potentially worse than the current B Pillar camera in others.

+1 on condensation fix requirement though.
Good point. I would suggest the wing mirrors but they're already the thing that always gets whacked by owners or other people, so that probably wouldn't be a good idea either. Perhaps the base of the wing mirror where it meets the car could have a camera in, I dunno.

EDIT: Although the wing mirror isn't much higher than the wing camera, and lower than the drivers eyeline.. argh!
 
The only thing that makes sense in my head was that they didn't want to have potentially significantly expensive parts on the outside of the car, which could potentially get damaged in the event of an accident, or stolen, or vandalised, etc.

I don't know how much the current wing cameras are, but if you had two cameras in one unit then it would be pretty expensive I think.

I had a ranger out late last year for condensation in my B-pillar cameras, as it basically meant that every journey I made I was getting warnings the whole way. He took one off the car, checked for the Goretex patch, put it back on and basically said for me to come back to them again if it persisted. As soon as it got warmer the problem went away, but it does seem to mean that for several months in a year the B-pillar cameras are degraded, which makes them unlikely to be much use for creeping out at junctions either!
Leeds SC told me they had ordered new parts for both b-pillars a couple of years ago. When I pushed them as to why they thought the replacement part would help, they acknowledged that it probably wouldn't! So I said 'don’t bother!'.
 
but this would introduce a new problem, as the camera would be much lower than the current B pillar camera - as such, if you have traffic to the side of you at a junction then the camera would not be able to see over it.

So it would work in some situations but be potentially worse than the current B Pillar camera in others.

+1 on condensation fix requirement though.

I think protection for edging out with reduced visibility would be a huge help and retain b-pillar cam's too.

There was talk somewhere on TMC about feasibility of multiplexing in more video for retro-fitting, or using some kind of wireless solution.

I think the consensus was that any wired solution would be a big headache and wireless prone to interference, especially from adjacent cars using similar technology.
 
Tesla will clearly reach an optimal point beyond which the system cannot improve sufficient to encompass some blind junctions. In truth, driver caution will win every time.
i have one locally where a building is right on the left corner with no pavement, both roads are narrow and to get out you have block traffic from the right. Visibility to the right is good but your bonnet is well Out into the road before you can see traffic from the left. Oh and there’s a bend to the left as well!
I have visibility well before the b pillar. a nose camera would absolutely nail that.
 
Leeds SC told me they had ordered new parts for both b-pillars a couple of years ago. When I pushed them as to why they thought the replacement part would help, they acknowledged that it probably wouldn't! So I said 'don’t bother!'.
When it gets colder no amount of preconditioning would help clear mine, and I'd pretty much get constant warnings for the duration of a journey. Only after leaving the car parked up, and online (sentry) did it usually clear itself, but it was unreliable.

I wonder if anyone (including Tesla!) have ever thought of installing something like this inside the B-pillar assembly: 8.0US $ |PI film heating sheet kapton PI Polyimide heater 3D Printer Electrothermal Film heating ring 5v 6v 12v 24v|Gaskets| - AliExpress

Seems like it would probably cure the problem?
 
  • Like
Reactions: Adopado
The angles covered by the B pillar cameras are pretty good - essentially the same as looking left and right.

iirc I don't think the angles covered are so much the problem, more about how far can an object be detected in relation to its speed. I think with chucks test scenario, you can detect a car at a certain distance, but detecting it at a distance far enough away where it can be determined that it is safe to proceed is a highly dependent on what sets of cameras are being used - forward facing windscreen mounted cameras give much better results than when b-post camera is being used as windscreen cameras have a much better range.

I'm sure someone will have the real-world camera distances vs car speed vs safe manoeuvre timings.
 
When it gets colder no amount of preconditioning would help clear mine, and I'd pretty much get constant warnings for the duration of a journey. Only after leaving the car parked up, and online (sentry) did it usually clear itself, but it was unreliable.

I wonder if anyone (including Tesla!) have ever thought of installing something like this inside the B-pillar assembly: 8.0US $ |PI film heating sheet kapton PI Polyimide heater 3D Printer Electrothermal Film heating ring 5v 6v 12v 24v|Gaskets| - AliExpress

Seems like it would probably cure the problem?

It would most likely fix the condensation issue. I did some tests with very gentle warming of the camera window from the outside and it hardly took anything to clear the glass.

I don't know about compatibility with camera imaging.

Again, even minor retrospective changes that include wiring / power supply changes can be a big headache.
 
i have one locally where a building is right on the left corner with no pavement, both roads are narrow and to get out you have block traffic from the right. Visibility to the right is good but your bonnet is well Out into the road before you can see traffic from the left. Oh and there’s a bend to the left as well!
I have visibility well before the b pillar. a nose camera would absolutely nail that.
There's a junction like that coming out of a car park where I worked pre-covid... in the winter you could see car headlights which helped, however in the summer it really was a case of sticking the nose of the car out bit by bit and hoping that no one was there or that if there was they saw you...

a few of my colleagues had accidents at that exit, I guess the problem here is as much to do with road planning, which should improve in many instances irrespective of self driving where it will inevitably evolve further.
 
The camera images need to overlap so the forward facing side cameras can't really be ahead of the rear facing side cameras otherwise you get a dead spot. I would have thought that for level 5 (ie driverless) you'd need everything to be dual anyway so single points of failure (wich could be dirt) can be handled They've now said they're trying to master the seemless stiching together of the images so more cameras could be added to deal with that. It has crossed my mind that FSD2 in the cybertruck might be more to do with changes to cameras and sensors than it is computing power.