Short-Term TSLA Price Movements - 2016

[ggr]

It is common that People lie to avoid responsibility. The wife did not want her husband to think it's her fault and thought she could lie her way out.
This happened more often than you think. I saw with my own eyes that my inlaw drove up to front of the garage, stopped for a couple seconds, then hit the gas pedal by mistake taking down her garage door.

In the 80s or 90s, Audi had a huge PR hit with "unintended acceleration". At some point Audi instrumented some cars (IIRC) and actually had a video of someone pushing hard on the accelerator while screaming "It won't stop!". The eventual conclusion was that all of these instances were people pushing the wrong pedal and panicking. (Toyota had a real problem, much later.) Audi actually moved the pedals in response, but were never officially found responsible.

 

[EinSV]

As I mentioned in the thread about this accident, my ex-mother-in-law drove her car right into the same 7-11 not once but twice. Needless to say, she no longer drives.

I am sure if she were driving a Model X or Model S it would have been all Tesla's fault. At least until they checked the logs.

 

[electracity]

As I mentioned in the thread about this accident, my ex-mother-in-law drove her car right into the same 7-11 not once but twice. Needless to say, she no longer drives.

My in laws each took out my garage door about a month apart. Accelerator accidents are apparently not uncommon. I can see foot confusion especially for a new Tesla driver.

 

[Krugerrand]

Cadillac adopting a page from Tesla's playbook.

Cadillac Bets on Virtual Dealerships

 

[Causalien]

I dont doubt if she got offered free lawyer from the oil brothers or GM...however There is no chance tesla will lose this case if it was as they said that log showed autopilot was never even turned on. Log data on thousands of other autopilot teslas can back that up in terms of logging accuracy. Free repair not good idea, will encourage others to lie and sue when they cause future accidents.

Yeah, which is pretty sad that this will go to court. Because both side cannot back down.

 

[SteveG3]

reports of Tesla in discussion with Samsung, and a substantial shipment of Samsung batteries sent to Tesla for testing,

Samsung SDI Climbs on Talks With Tesla to Supply Batteries

perhaps a reflection of energy storage business demand. article talks about this as possibly a strategic move for leverage with Panasonic. that may be. I'm wondering if it may also be something of a strategic move on Samsung's part in regard to Apple.

 

[Xpert]

Cadillac adopting a page from Tesla's playbook.

Cadillac Bets on Virtual Dealerships

So finally The copycat game has started. This might mark the beginning of the end of the century old dealership model as we know it.

 

[Xpert]

If it were me in this situation, I would insist that an independent third party examine that logs to see if they come to the same conclusion as Tesla. And if the logs aren't cryptographically signed to prove that the data came from my car at the appropriate time, then I'd say the evidence is meaningless in a court of law. There could also be issues around providing the code that created the logs, as there is really no way for Tesla to insist that there are no bugs that could cause anomalous behavior.

I'm guessing that Tesla has thought this through from the beginning and that the logs are properly time-stamped and signed, but it's just a guess. I think it's very unlikely that Tesla is getting itself in a hole here, but if they were sloppy they could be telling the truth and still have trouble, either because the logs are modifiable (and thus not useful evidence) or there are relevant bugs.

They wont ask for any review as they know they lied. And if you understand how programming works it wont be necessary.

 

[Causalien]

They wont ask for any review as they know they lied. And if you understand how programming works it wont be necessary.

No, there is still the possibility that the sensor for the gas pedal gave the wrong output. Even with optop triggers, a slight delay in the capacitor of the receiver can cause a missed reading.

I think what tesla can do is to show a profile of tesla's autopilot acceleration curve. Since it is programmed, the accel curve will be the same with slight deviation while a human's will be all over the place.

 

[CButterK]

reports of Tesla in discussion with Samsung, and a substantial shipment of Samsung batteries sent to Tesla for testing,

Samsung SDI Climbs on Talks With Tesla to Supply Batteries

perhaps a reflection of energy storage business demand. article talks about this as possibly a strategic move for leverage with Panasonic. that may be. I'm wondering if it may also be something of a strategic move on Samsung's part in regard to Apple.

That large shipment from Samsung were low capacity, low density 18650 cells, which WILL NOT work for cars, so it has to be for TE. So that said, I'm very curious regarding this move. Considering Tesla just announced that they can produce up to 3X their original cap, I wonder why they need Samsung's cells. Is it possible that demand for Tesla cars & Energy is so high, that Panasonic can't produce enough? Or perhaps Panasonic just can't produce low end cells for TE at a price that Samsung is offering? Either way, this has to be a positive move for Tesla, no?

 

[NicoV]

That large shipment from Samsung were low capacity, low density 18650 cells, which WILL NOT work for cars, so it has to be for TE. So that said, I'm very curious regarding this move. Considering Tesla just announced that they can produce up to 3X their original cap, I wonder why they need Samsung's cells. Is it possible that demand for Tesla cars & Energy is so high, that Panasonic can't produce enough? Or perhaps Panasonic just can't produce low end cells for TE at a price that Samsung is offering? Either way, this has to be a positive move for Tesla, no?

It could be for something else than TE or for propulsion battery: E.g. maybe they want to replace the lead-acid battery with something thats better suited for the continuous recharge cycles of the vampire drain.

 

[Xpert]

No, there is still the possibility that the sensor for the gas pedal gave the wrong output. Even with optop triggers, a slight delay in the capacitor of the receiver can cause a missed reading.

I think what tesla can do is to show a profile of tesla's autopilot acceleration curve. Since it is programmed, the accel curve will be the same with slight deviation while a human's will be all over the place.

while I can understand hardware failure is a very remote possibility and a totally different issue, manual acceleration by pressing the pedal would show in the log as a separate input vs. autopilot acceleration. Capacitor or sensor failure can easily be tested after the crash, and I highly doubt that is a problem. Again, autopilot was never engaged as the wife claimed.

Example codes would be like:
if (autopilot = on)
{
write in log ("autopilot engaged")
accelerate at given speed
...
}else{
if(gas pedal pressed)
{
write in log ("manual acceleration", "date/time", "speed")
accelerate according to pedal pressed percentage
}

“We analyzed the vehicle logs which confirm that this Model X was operating correctly under manual control and was never in Autopilot or cruise control at the time of the incident or in the minutes before. Data shows that the vehicle was traveling at 6 mph when the accelerator pedal was abruptly increased to 100%. Consistent with the driver’s actions, the vehicle applied torque and accelerated as instructed. Safety is the top priority at Tesla and we engineer and build our cars with this foremost in mind. We are pleased that the driver is ok and ask our customers to exercise safe behavior when using our vehicles.”

 

[Fallenone]

That large shipment from Samsung were low capacity, low density 18650 cells, which WILL NOT work for cars, so it has to be for TE. So that said, I'm very curious regarding this move. Considering Tesla just announced that they can produce up to 3X their original cap, I wonder why they need Samsung's cells. Is it possible that demand for Tesla cars & Energy is so high, that Panasonic can't produce enough? Or perhaps Panasonic just can't produce low end cells for TE at a price that Samsung is offering? Either way, this has to be a positive move for Tesla, no?

The 3X is not a definite, like the 2k per week, they can, but may not be wise to really do that (overtime high wages, stress on equipment, etc.). Also they didn't specify when they can reach this potential 3X, for all we know, it could be happening by the end of this year with the phase 1 14%*3, or can be 2020 when the original plan is completed, or can be even longer when they utilized the additional land they purchased.

 

[Fallenone]

Did some speculation on what the percentage does China reservation for Model 3 account for the total. My conclusion is about 7.1%. Details in this thread
China Market situation and outlook

 

[Causalien]

while I can understand hardware failure is a very remote possibility and a totally different issue, manual acceleration by pressing the pedal would show in the log as a separate input vs. autopilot acceleration. Capacitor or sensor failure can easily be tested after the crash, and I highly doubt that is a problem. Again, autopilot was never engaged as the wife claimed.

Example codes would be like:
if (autopilot = on)
{
write in log ("autopilot engaged")
accelerate at given speed
...
}else{
if(gas pedal pressed)
{
write in log ("manual acceleration", "date/time", "speed")
accelerate according to pedal pressed percentage
}

“We analyzed the vehicle logs which confirm that this Model X was operating correctly under manual control and was never in Autopilot or cruise control at the time of the incident or in the minutes before. Data shows that the vehicle was traveling at 6 mph when the accelerator pedal was abruptly increased to 100%. Consistent with the driver’s actions, the vehicle applied torque and accelerated as instructed. Safety is the top priority at Tesla and we engineer and build our cars with this foremost in mind. We are pleased that the driver is ok and ask our customers to exercise safe behavior when using our vehicles.”

There's not much needed to say about code except if it runs into unknow state with no catchall to gracefully fail. This requires tesla to open up the source.

The sensor problem is the harder to prove part in the lawsuit. As it is very possible that standard tests cannot detect a 1 in a billion runaway case. Being in engineering, U am sure you've met a few of these. This it is the easiest path to sue for.

The logic for defense is easy though, if you applied the breaks and the sensors misses it, the worst the car will do is just coast at the current apeed and sliw down with regenerative breaking. They mentioned acceleration. So the only possibily left is the sensors for accel and break got shorted together. Or autopilot got engated. But this is easy to prove by looking at logs of the Finite state machine state.

Can't prove anything without the source. So tesla wins. That guy seems pretty important though. So tesla might win my purposefuly losing.

 

[Gerardf]

while I can understand hardware failure is a very remote possibility and a totally different issue, manual acceleration by pressing the pedal would show in the log as a separate input vs. autopilot acceleration. Capacitor or sensor failure can easily be tested after the crash, and I highly doubt that is a problem.”

Some might argue that the gas pedal sensor might have an error, I see this suggested in several places.

The lady said she did push the breaking pedal. In case the logs show the gas pedal was sending a signal, and the same log does NOT show a break signal, proof is that she made a mistake and pressed the wrong pedal. Case closed.
So the log of the breaking pedal is at least as important in this case as the accelerator pedal log.
I hope Tesla can put this case to rest soonest.

 

[NOLA_Mike]

In the 80s or 90s, Audi had a huge PR hit with "unintended acceleration". At some point Audi instrumented some cars (IIRC) and actually had a video of someone pushing hard on the accelerator while screaming "It won't stop!". The eventual conclusion was that all of these instances were people pushing the wrong pedal and panicking. (Toyota had a real problem, much later.) Audi actually moved the pedals in response, but were never officially found responsible.

But, in the meantime the press was so bad that Audi essentially went out of business in the USA for a decade or so...

 

[Hogfighter]

Not that it matters much, but 2 different errors would have had to occur with the software for this accident to happen.

1) The vehicle would have to commanded the acceleration
2) The vehicle would have had to ignore the application of the brakes

So which is more likely, both of those to occur simultaneously, or that the lady hit the accelerator instead of the brake?

 

[brantse]

Not that it matters much, but 2 different errors would have had to occur with the software for this accident to happen.

1) The vehicle would have to commanded the acceleration
2) The vehicle would have had to ignore the application of the brakes

So which is more likely, both of those to occur simultaneously, or that the lady hit the accelerator instead of the brake?

+1 Also, I think the primary reason that this is a very likely scenario, when compared to an ICE vehicle, is the fact that full acceleration is nearly silent. The loud roar of an engine is likely enough feedback for most people that they may have hit the wrong pedal and instinctively release it.

 

[vgrinshpun]

Not that it matters much, but 2 different errors would have had to occur with the software for this accident to happen.

1) The vehicle would have to commanded the acceleration
2) The vehicle would have had to ignore the application of the brakes

So which is more likely, both of those to occur simultaneously, or that the lady hit the accelerator instead of the brake?

Exactly. For the reference, such simultaneous failures are termed as "double contingency" in the nuclear industry. Double contingency is considered so unlikely, that nuclear power plants are NOT required to be designed for it.