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

FSD Beta Videos (and questions for FSD Beta drivers)

This site may earn commission on affiliate links.
jury is still out on the reliability of wifi (if that's what you mean) for high bandwidth mission critical sensors.

personally, I would not ask for this, on my design. I would not want to trust the unreliability of wireless for this.
jury is still out on the reliability of wifi (if that's what you mean) for high bandwidth mission critical sensors.

personally, I would not ask for this, on my design. I would not want to trust the unreliability of wireless for this.

in fact, I keep arguing for fiber, not copper. but I always get shot down. (but I keep trying, one vendor will take the idea, I'm sure..)
wireless short range is very reliable
 
Well, I logged about 20 or so miles on 9.1 this evening. My neighbor has it, she let me take her car out.

-failed the first left turn I tried. Put the car in the wrong lane.

- speed limited to 25-33mph on roads with no lane markings? Also, poor lane keeping on roads with no markings, car “hunted” a lot, kinda like a drunk driver.

-random turn signal usage for no particular reason

-some of turns at intersections (right or left) are pretty slow.

-it handled a lot of other things very well

-performance on turns where lane markings were present were handled very well. “Human like” I’d say.

It was interesting to say the least. We in a pretty rural area, a lot of the turns I was trying were pretty tough, a lot of visibly issues. Maybe that’s why there were issues.
 
Not on the nose, where the other cameras are now.
That would help a bit over the B pillar cams. Did you watch Chuck Cook's latest ULT video? The present mix of cameras do not see far enough left and right to consistently allow safe ULTs onto busy highways. I had estimated 100 m side range for the B pillar cameras but Chuck actually measured it with a drone and it was 80 m. At 100 km/hr, less than three seconds isn't going to allow FSD to make such a manuever. In this video, the FSD beta basically gave up on making a ULT most of the time. When it didn't, it made unsafe turns that required intervention.
 
  • Helpful
Reactions: diplomat33
wireless short range is very reliable
I'm sorry, but I have to disagree.

hundreds - no thousands of hours of automated tests show otherwise.

at least not wifi and ip. if you run a pure rf protocol (like the 2.4ghz uses in RC toys) then it can be very good but I still think wired is what is called for, when life and limb are at stake. in fact, as I often harp about, redundant wired and not just single-linked.

I like the weight loss of wireless, but so far, I've not found that its at the reliability I would demand for automotive use. on a feature that isn't critical (maybe a backup camera) it could be ok, but not when the control of the car, at high speed, would depend on that video stream.

finally, wireless can be attacked easier than wired. we dont want that in automotive. we *REALLY* do not. and we think of those things when we design a transport for data that just cannot afford drops or errors or attacks by bad actors.
 
I'm sorry, but I have to disagree.

it’s ok to be wrong 👍

at least not wifi and ip. if you run a pure rf protocol (like the 2.4ghz uses in RC toys) then it can be very good

yes, I know. There is no reason to limit yourself to IP cameras. This isn’t home brew.

but I still think wired is what is called for,

you are entitled to your OPINION

I like the weight loss of wireless, but so far, I've not found that its at the reliability I would demand for automotive use. on a feature that isn't critical (maybe a backup camera) it could be ok, but not when the control of the car, at high speed, would depend on that video stream.

finally, wireless can be attacked easier than wired. we dont want that in automotive. we *REALLY* do not. and we think of those things when we design a transport for data that just cannot afford drops or errors or attacks by bad actors.

This is fluffery.

A cord can be damaged just as easy as a wireless link can be lost. There’s no redundancy in either system.

Your attack worry is mostly unfounded. It would require that the system be operation before and while autopilot was engaged, then, once it was engaged, at speed, hacked, the system being unable to detect it was being hacked, and then???? What? What happens then? It looses signal would be about the worst case.

In that case I suppose it would just fall back to the wired set of cameras to safely pull over, since we aren’t talking about all the cameras being wireless.

the real issue is latency, which you havnt brought up, which is how I know you aren’t a pro. But I think over the short distances involved based on commercial products I have seen, the four cameras I have proposed would work fine wirelessly, even with a little latency (since we are mainly trying to determine traffic crossing the car and identifying gaps large enough for the car to safety cross an intersection, so the reaction time wouldn’t really be miliseconds in those cases.
 
Apparently in 8.2, you can adjust the speed to anything you'd like, but I'm not sure if that's still the case with 9.1, and if it saw a sign, I think it resets back to the limit.
Eh, I don’t know. The unmarked roads I tried had speed limit signs, (mostly all 45mph). It would always display them as 25mph, but in certain cases the car would get up to like 33mph. On marked roads, no issues. Weird.
 
if they need to retrofit cameras in then they’re screwed with current cars. It would absolutely not be easy to retrofit that stuff in unless they happened to put in wires already in anticipation of that, which is doubtful. Even if the wire harnesses are already installed and waiting, cutting Holes in the bumpers or replacing them entirely with new ones that have holes for the cameras in would not be trivial.
Actually I think the easiest change would be to add dual cameras to the repeaters (would just need a swap). But as you say the wiring harness is an issue. There are probably ways that can compress the feed and send it through the existing wiring, but fitting a controller that does this along with cramming another camera might be impractical.
 
Actually I think the easiest change would be to add dual cameras to the repeaters (would just need a swap). But as you say the wiring harness is an issue. There are probably ways that can compress the feed and send it through the existing wiring, but fitting a controller that does this along with cramming another camera might be impractical.
what are the repeaters? Do you mean the side cameras?
 
it’s ok to be wrong 👍
for someone who is not working in this field, you sure are sure of yourself.

I'll just end it here, for my part. I have other things to do with my time than argue with those that have not done this before.

my actual years of testing wireless in an automotive field means nothing to you. ok. so I wont offer my insights anymore.

you win.
 
Sad, Tesla has really painted themselves up in the corner with all those promises the last 5 years.
Going to retrofit a million or more cars with front bumper radars or higher resolution cameras will be very expensive.
I dont think they will. as long as elon is there, his damned ego wont let them admit what the tech really needs.

he's the problem with tesla. get rid of him and tesla can actually move forward. he's not the brains - but he interferes like he is. wrong job for him.

when I heard about them REMOVING a sensor (radar) I knew that was it - beginning of the end. unless they have a come-to-jesus moment and re-add what they removed, they wont ever achieve a safety level that many of us INSIST on if we are to give the car more and more self-control.

its like you are losing a tennis match with an opponent, so you blind one eye thinking it might help. what kind of thinking is that?
 
for someone who is not working in this field, you sure are sure of yourself.

I'll just end it here, for my part. I have other things to do with my time than argue with those that have not done this before.

my actual years of testing wireless in an automotive field means nothing to you. ok. so I wont offer my insights anymore.

you win.
You are offering you opinion. But you are also being disingenuous. Why would tesla limit themselves to WiFi when, as you pointed out, there are alternatives that are quite good. Then you say there’s no redundancy, which is another disingenuous statement because nothing we are talking about has redundancy. Then you lay down you last attack by saying wireless cameras could be hacked, without laying out the “so what” or consequence, hoping to use the fear train to win the argument.

It’s OK to have an opinion, it doesn’t make you right. It’s ok to be wrong 👍

A lot of “smart” people have been wrong before lol. But you are right, I’ve destroyed your feeble argument enough that there isn’t a need for you to reply.
 

Just my opinion of course, but I think Tesla could benefit by adding front bumper cameras pointing left and right. It would give the car a better view of cross traffic when approaching intersections. It would be especially helpful because there are instances when the view is partially blocked like when tall bushes line the road making it hard to see cars coming down the road or when there is traffic in the adjacent lane that blocks the B pillar from getting a good view.
 
Thanks, great video.

Seems car identifies other cars around 50 meters out, that is 2 seconds at 60 mph/26 m/s or way to little to safely know when to drive. Chuck though has 150 meter of vision, meaning 5-6 seconds to act.

Sad, Tesla has really painted themselves up in the corner with all those promises the last 5 years.
Going to retrofit a million or more cars with front bumper radars or higher resolution cameras will be very expensive.

When the UI is actually directly facing oncoming traffic and the car has pulled out into the cross walk at least a little bit, it's definitely over 100 m .

I mean, the main issue is that the B-pillar cameras are behind where our eyes sit in the car (and we definitely lean forward in these scenarios). So the car has to poke out even more than we would poke out too see well enough.
 
Why does FSD Beta just swerve into incoming traffic? Seriously, WTF?!?

Note the color of the broken stripes separating opposing traffic when the car wanted to move left. They're white. White lane markings are only used for same-direction travel.

That's the US standard for separating lanes moving in the same direction. Yellow stripes are to be used to separate different directions of traffic. See, for example, Helpful Driving Info | Markings: Colors, Patterns, Meaning

More evidence: many locales in CA have replaced the double double YELLOW lines separating HOV from mainlines (both with traffic in the same direction) with double WHITE lines.

So the city has made an error in road markings, which could confuse anyone depending on those lane markings that imply one-way traffic.

This is a corner case due to improper lane markings. Nothing more.
 
Note the color of the broken stripes separating opposing traffic when the car wanted to move left. They're white. White lane markings are only used for same-direction travel.

That's the US standard for separating lanes moving in the same direction. Yellow stripes are to be used to separate different directions of traffic. See, for example, Helpful Driving Info | Markings: Colors, Patterns, Meaning

More evidence: many locales in CA have replaced the double double YELLOW lines separating HOV from mainlines (both with traffic in the same direction) with double WHITE lanes.

So the city has made an error in road markings, which could confuse anyone depending on those lane markings that imply one-way traffic.

This is a corner case due to improper lane markings. Nothing more.

Which would easily be solved with HD maps. ;)
 
I am being a bit tongue and cheek but I wonder what Tesla's visualizations would show if it saw this? LOL


E8Xt3pDXEAE64Lt
 
  • Like
Reactions: Thp3