Tesla Autopilot vs. Cirrus Jet

Apparently the back end of an SF50 is too high to be considered an obstacle, until it is.
 
When Tesla used Lidar, did it recognize an obstacle, even if it was something uncommon?
 
I’m sure the Tesla lawyers would be able to win without irrefutable proof that you stopped summoning and the car ran amok on its own, even if that actually happened, how would you prove it?
 
DQYZizUVAAEH74k
 
Salty said:
I’m sure the Tesla lawyers would be able to win without irrefutable proof that you stopped summoning and the car ran amok on its own, even if that actually happened, how would you prove it?
Click to expand...
Tesla’s lawyers probably already convinced the engineers to keep detailed logs of user input for precisely this reason.
 
iamtheari said:
Tesla’s lawyers probably already convinced the engineers to keep detailed logs of user input for precisely this reason.
Click to expand...
You're saying the opposite I am. If I am the user and use summon and let go of the button, but the car keeps going, there's virtually no way I could prove that I let go of the button. Logs could only work against Tesla. And there aren't going to be logs that prove the phone didn't glitch and give the app bad data.

Frankly, the technology isn't there yet. There's a reason why most of this type of thing is done with locked down devices not iOS apps.

As a user, you really are taking a lot of risk using summon. Tesla has very little liability, whether or not it was their fault. The user agreements make that pretty clear. Unless there's a huge rash of Tesla's crashing into stuff on summon, they ain't gonna be paying.
 
Felts Field, Spokane.

"Self crashing vehicle" lol
 
Salty said:
I’m sure the Tesla lawyers would be able to win without irrefutable proof that you stopped summoning and the car ran amok on its own, even if that actually happened, how would you prove it?
Click to expand...

The time data that records when and for how long the summons was active, overlaid on the GPS and accelerometer indications of when the impact occurred. Tesla knew that info not long after the robot car assault happened.
 
3393RP said:
The time data that records when and for how long the summons was active, overlaid on the GPS and accelerometer indications of when the impact occurred. Tesla knew that info not long after the robot car assault happened.
Click to expand...
That does not prove when you lifted your finger off your phone.
 
Salty said:
That does not prove when you lifted your finger off your phone.
Click to expand...

Perhaps you didn't read my response correctly. The vehicle's data log records when and for long the summons was active. That information is available to Tesla via data link to their cars. The impact was similarly recorded. A time log makes coincidence of the summons and impact quite simple to interpret.
 
FastEddieB said:
Exactly. Lots of limitations and caveats noted here:

https://www.tesla.com/ownersmanual/modely/is_is/GUID-6B9A1AEA-579C-400E-A7A6-E4916BCD5DED.html
Click to expand...

I took the liberty of doing an edit on one in particular:

Warning
Smart Summon may not stop for all objects (especially very low objects such as some curbs, or very high dollar objects such as a multi-million dollar private jet) and may not react to all traffic. Smart Summon does not recognize the direction of traffic, does not navigate around empty parking spaces, and may not anticipate crossing traffic.
Click to expand...
 
3393RP said:
Perhaps you didn't read my response correctly. The vehicle's data log records when and for long the summons was active. That information is available to Tesla via data link to their cars. The impact was similarly recorded. A time log makes coincidence of the summons and impact quite simple to interpret.
Click to expand...
You’re missing something. It’s the users fault if he’s still holding the summon button. It’s teslas fault if the user is not holding the button. In either case the summon would be active from the cars logs. Tesla never claims to assure avoiding striking an obstruction. Quite the contrary. They make it clear the cameras can’t see everything and that you are responsible and must be in line of sight.
 
3393RP said:
Perhaps you didn't read my response correctly. The vehicle's data log records when and for long the summons was active. That information is available to Tesla via data link to their cars. The impact was similarly recorded. A time log makes coincidence of the summons and impact quite simple to interpret.
Click to expand...
You're assuming that the logs are telling the truth, which may or may not be the case.
 
Was the Tesla owner and vision jet owner one in the same??
 
I'm thinking it's Tesla's fault, unless proven otherwise. To mis-quote Cartman, if it's so d**n safe, why does it keep getting caught in fishing nets all the time?
 
Salty said:
You’re missing something. It’s the users fault if he’s still holding the summon button. It’s teslas fault if the user is not holding the button. In either case the summon would be active from the cars logs. Tesla never claims to assure avoiding striking an obstruction. Quite the contrary. They make it clear the cameras can’t see everything and that you are responsible and must be in line of sight.
Click to expand...

It is interesting that the aircraft fuselage is mostly white, just like the multiple eighteen wheeler trailers that Teslas have struck while on autopilot.
 
Last edited:
asicer said:
You're assuming that the logs are telling the truth, which may or may not be the case.
Click to expand...

:D

I didn't know AI has progressed to the point where computers can decide to tell lies. That Elon guy is almost like Jesus.
 
3393RP said:
just like the multiple eighteen wheeler trailers that Teslas have struck while on autopilot.
Click to expand...
Multiple? I only know of one which involved an older Model S. That bug was fixed many years ago. What were the others?

In my view, the fault lies with the details of the incident. If the Summon feature was working as designed then it was the operator, who was holding the summon button on the app, who is at fault. If the operator released the summon button, but the car continued to move into the airplane, then it's Tesla's fault.
 
Larry in TN said:
Multiple? I only know of one which involved an older Model S. That bug was fixed many years ago. What were the others?

In my view, the fault lies with the details of the incident. If the Summon feature was working as designed then it was the operator, who was holding the summon button on the app, who is at fault. If the operator released the summon button, but the car continued to move into the airplane, then it's Tesla's fault.
Click to expand...
Except that there is no way for a user to prove when they released the button, so no way to prove its Teslas fault. This is what I’ve been trying to say.
 
Salty said:
...It’s the users fault if he’s still holding the summon button. It’s teslas fault if the user is not holding the button. In either case the summon would be active from the cars logs...
Click to expand...

Salty said:
Except that there is no way for a user to prove when they released the button, so no way to prove its Teslas fault.
Click to expand...
Is the summon button state recorded in the logs or is it not?

Nauga,
and a little onboard data
 
Salty said:
You're saying the opposite I am. If I am the user and use summon and let go of the button, but the car keeps going, there's virtually no way I could prove that I let go of the button. Logs could only work against Tesla. And there aren't going to be logs that prove the phone didn't glitch and give the app bad data.

Frankly, the technology isn't there yet. There's a reason why most of this type of thing is done with locked down devices not iOS apps.

As a user, you really are taking a lot of risk using summon. Tesla has very little liability, whether or not it was their fault. The user agreements make that pretty clear. Unless there's a huge rash of Tesla's crashing into stuff on summon, they ain't gonna be paying.
Click to expand...

I'm sorry, but I don't understand the highlighted comment in light of how Smart Summon purportedly works.
https://www.tesla.com/ownersmanual/modely/is_is/GUID-6B9A1AEA-579C-400E-A7A6-E4916BCD5DED.html
Smart Summon works with the Tesla mobile app when your phone is located within approximately 6 meters of Model Y.
Click to expand...
This seems to imply that a phone app is actually used (iOS or Android not specified).

Salty said:
Except that there is no way for a user to prove when they released the button, so no way to prove its Teslas fault. This is what I’ve been trying to say.
Click to expand...
Reading through your posts, I think I understand what you are trying to say- that the car logs can only record when it receives the "push" or "release", not whether the phone actually sent the commands in a timely fashion.

You are making assumptions in what is or isn't logged. For example, there's no reason that the button press and release commands can't also include a time stamp that can be logged with the time the car actually received and acted on the commands. There's no reason the app itself doesn't have a log file. As the citation above indicates the feature is still in beta testing, it wouldn't surprise me the app keeps a log as well, which is compared to the car's log.

Quoted for the citation above, this may well be what happened:
Always anticipate when you need to stop Model Y. Depending on the quality of the connectivity between the phone and Model Y, there may be a slight delay between when you release the button and when the car stops.
Click to expand...
 
nauga said:
Is the summon button state recorded in the logs or is it not?

Nauga,
and a little onboard data
Click to expand...
Probably, but if the app locked up or ios sent bad data to it they would be “lying”
 
Cap'n Jack said:
I'm sorry, but I don't understand the highlighted comment in light of how Smart Summon purportedly works.
https://www.tesla.com/ownersmanual/modely/is_is/GUID-6B9A1AEA-579C-400E-A7A6-E4916BCD5DED.html

This seems to imply that a phone app is actually used (iOS or Android not specified).


Reading through your posts, I think I understand what you are trying to say- that the car logs can only record when it receives the "push" or "release", not whether the phone actually sent the commands in a timely fashion.

You are making assumptions in what is or isn't logged. For example, there's no reason that the button press and release commands can't also include a time stamp that can be logged with the time the car actually received and acted on the commands. There's no reason the app itself doesn't have a log file. As the citation above indicates the feature is still in beta testing, it wouldn't surprise me the app keeps a log as well, which is compared to the car's log.

Quoted for the citation above, this may well be what happened:
Click to expand...
See my last post.

the “lag” is because of latency in wireless transmission as well as computation limitations.
 
Salty said:
Probably, but if the app locked up or ios sent bad data to it they would be “lying”
Click to expand...
Does every potential failure mode where the car continues to move include the app locking or ios sending bad data? While I agree it's *possible* that the logs would be in error, I do not agree that it's a certainty.

Nauga,
planting a fault tree
 
Salty said:
Except that there is no way for a user to prove when they released the button, so no way to prove its Teslas fault. This is what I’ve been trying to say.
Click to expand...
If there's a bug which allows the car to continue moving without receiving a constant signal from the app, that can be determined and fixed. The logs (presumably) kept by the car will show if it was receiving a "go" signal. If it stopped receiving that signal, but continued moving, that's a software bug (design or implementation) and would be the cause.

I would think that the most likely explanation is that the operator wasn't paying enough attention but a software issue is also possible.
 
3393RP said:
I didn't know AI has progressed to the point where computers can decide to tell lies. That Elon guy is almost like Jesus.
Click to expand...
Incorrect data recorded in log files has long preceded the invention of AI.
 
Larry in TN said:
If there's a bug which allows the car to continue moving without receiving a constant signal from the app, that can be determined and fixed. The logs (presumably) kept by the car will show if it was receiving a "go" signal. If it stopped receiving that signal, but continued moving, that's a software bug (design or implementation) and would be the cause.

I would think that the most likely explanation is that the operator wasn't paying enough attention but a software issue is also possible.
Click to expand...
I think so too. All I’m trying to say is that as a user, I can’t prove I let go of the button at x point of time. The touch screen can fail, the os can fail, without Tesla being at fault. Unless you try to claim the system is inherently unsafe because they are using commodity hardware and os, which have the described issues. But that argument wouldn’t go far unless there is a rash of incidents.
 
Salty said:
See my last post.

the “lag” is because of latency in wireless transmission as well as computation limitations.
Click to expand...
I'm sorry, but you still aren't making yourself clear. I've written simple phone apps, as well as much more complicated programs on computers, so I don't see the "computation limitations" related to whether a control is pressed or not.

As for "latency in wireless transmission"- I assume this mean the latency of the network that gets the data from the phone to the car. I can see that happening.

This really sums up things well, assuming they aren't logging their app.
Salty said:
I think so too. All I’m trying to say is that as a user, I can’t prove I let go of the button at x point of time. The touch screen can fail, the os can fail, without Tesla being at fault. Unless you try to claim the system is inherently unsafe because they are using commodity hardware and os, which have the described issues. But that argument wouldn’t go far unless there is a rash of incidents.
Click to expand...
 
Cap'n Jack said:
I'm sorry, but you still aren't making yourself clear. I've written simple phone apps, as well as much more complicated programs on computers, so I don't see the "computation limitations" related to whether a control is pressed or not.
Click to expand...
Computers always have limitations. They don’t scale infinitely. I only added that to cover other possibilities. The physical OSI layer is certainly the most likely place to introduce latency, but the other layers that run on the phone and the car can also run out of memory and cpu resulting in latency.
 
Cap'n Jack said:
As for "latency in wireless transmission"- I assume this mean the latency of the network that gets the data from the phone to the car. I can see that happening.
Click to expand...
I can't remember if summon works through the phone and car's cellular data connection or if it's through Bluetooth. I know that one of the complaints about the feature is that you have to be relatively close to the car but that still seems like it might be out of Bluetooth range.
 
Larry in TN said:
I can't remember if summon works through the phone and car's cellular data connection or if it's through Bluetooth. I know that one of the complaints about the feature is that you have to be relatively close to the car but that still seems like it might be out of Bluetooth range.
Click to expand...
Yeah. BT spec is minimum ten meters, though I've experienced situations where it was effectively more or effectively less. So some other wireless tech. Hey, there's latency in controlling my home's thermostats.
 
Salty said:
Computers always have limitations. They don’t scale infinitely. I only added that to cover other possibilities. The physical OSI layer is certainly the most likely place to introduce latency, but the other layers that run on the phone and the car can also run out of memory and cpu resulting in latency.
Click to expand...
Very simply, too many apps running or too much data stored. More likely the phone than the car, unless there's a "memory leak" on one of the car's programs.
 
You must log in or register to reply here.
Back
Top