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

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

This site may earn commission on affiliate links.
It seems my Model 3 can recognize stopped vehicles.
Sometimes, sure, using the vision subsystem. But the Doppler radar has to filter out non-moving responses (else it sees the ground, the curb, soda cans, etc.) so, for example, if the car in front of you suddenly moves out of the the lane, due to a stopped car (or firetruck!) in front if it, your car is less likely to determine that stopped car is really there, as the radar says no, and the vision hasn't had enough time to make a true determination.
 
This was Tweeted out about 15 minutes ago. Dana Hull and some of our favorite FUDsters are pushing it out:

Senator Markey Announces Recommended Safety Fixes for Tesla Autopilot

:rolleyes::rolleyes::rolleyes:

Both are stupid. Rebranding won’t do anything, whatsoever, as autopilot already implies that the system requires human oversight, and anyone who’s driven the car for more than 2 minutes knows that it tries to force you to do so.

And there’s no security measure that can withstand complete access to the hardware. Trying to prevent the owner from intentionally defeating the driver monitoring system is a hopeless enterprise.

Stupid, stupid, stupid.
 
Honestly this is bullish. Could have gone much worse. Those recommendations are by no means an existential threat to autopilot for Tesla.

Edit to add: I do think Tesla needs to do a better job of making people aware that autopilot often does not recognize objects in the road that are 100% stationary, like fully stopped vehicles. That's what most of these accidents are, and they need to let people know more explicitly that stationary objects in the roadway are the biggest risk of using autopilot.

I've never once had to take over to stop my Model 3 when a stationary car was ahead of me. And this happens on my commute everyday, where a stretch of road with a 90km/h limit has a stop light. I often go 100+km/h and my car ALWAYS detects stationary cars at the stop light.
 
I've never once had to take over to stop my Model 3 when a stationary car was ahead of me. And this happens on my commute everyday, where a stretch of road with a 90km/h limit has a stop light. I often go 100+km/h and my car ALWAYS detects stationary cars at the stop light.

I've gone past mail trucks and utility trucks that were parked off the side of the road but sticking into the travel lane, where I've taken over because it wasn't apparent to me that autosteer was dodging out of the way to avoid a side-scrape or actual collision. (I don't KNOW that it would have hit anything, it just wasn't clear to me that it wouldn't have.)

I've also taken over to avoid debris in the road.

Of course, that while using autosteer on city streets where it's not yet supposed to be used, so I don't really hold any shortcomings against it.

But I do think it has room to improve before fully autonomous driving is enabled. (No surprise there.)
 
Last edited:
I've never once had to take over to stop my Model 3 when a stationary car was ahead of me. And this happens on my commute everyday, where a stretch of road with a 90km/h limit has a stop light. I often go 100+km/h and my car ALWAYS detects stationary cars at the stop light.

Sure but most of the time it recognized the car prior to it coming to a full stop right? In other words the other car was moving and then stopped? Obviously there are times when autopilot does not recognize stopped vehicles. That is indisputable. How often that is an issue is something I cannot say.
 
Both are stupid. Rebranding won’t do anything, whatsoever, as autopilot already implies that the system requires human oversight, and anyone who’s driven the car for more than 2 minutes knows that it tries to force you to do so.

Especially when the media reports that Teslas are self driving anyway.
Which do people read more, headlines or the owner's manual?
 
  • Funny
Reactions: Lessmog
Recent Electrek article regarding production of the new Fiat 500e:

The return of the Fiat 500e, and its outsized meaning for EVs - Electrek

“In 2019, Fiat said it would invest nearly $800 million on the new 500e, including a new production line in Mirafiori, Italy. Fiat said it would make 80,000 units a year of the 500e, starting in the second quarter of 2020.” (Emphases mine).

This is the first time I have seen any projections of FCA 2020 EV production. So I plugged these numbers into my FCA-Tesla Pool model to assess the implication for 2020 CO2 emission penalty reduction. All the previous assumptions and references in the following link are used (I’ll update when 2019 EU sales and emission figures are released) unless called out below.

Tesla, TSLA & the Investment World: the 2019-2020 Investors' Roundtable

New Assumptions:
I assumed Fiat can produce and sell 40,000 500e’s in the EU in 2020. The 500 is a popular model – around 188,000 Fiat 500/Abarth 500’s were sold in Europe in 2018 (this does NOT include the 4-door 500L and 500X), so replacing 40,000 of these with 500e’s represents a 21% model substitution.

http://carsalesbase.com/european-car-sales-data/fiat/fiat-500/

The 2018 total FCA fleet size of ~700,000 was retained and the average fleet emissions (124 g/km) and emission target (92 g/km) were applied to the postulated remaining 660,000 ICE vehicles.

Results:
The 40,000 500e’s would absorb all of the available 2020 “Super-Credits” - 20,700 ZEV’s @ €19,500 penalty reduction per ZEV, plus 19,300 “regular” ZEV credits of €9,000 per ZEV for a total penalty reduction of around €577 million.

This still leaves €1,138 million penalty to be offset by Teslas in the pool – around 126,000 would be required. If we retain our working assumption of a 50% penalty reduction payment to Tesla, that would be €569 million for 2020.

Even if FCA "only" sells 20,000 500e's in 2020, that will still scoop up nearly all the Super-Credits for a penalty reduction of around €390 million. Not a bad strategy.
Can they sell a lot of ~100 mile range small cars compared roomy more capable, 220+ range model 3 costing only ~5k more and becoming more available in that market?

It may be valid to assume there is a limit to sales well below the aspirational run rate of 1500/week by end of Q2.
 
  • Like
Reactions: rsm287
Sometimes, sure, using the vision subsystem. But the Doppler radar has to filter out non-moving responses (else it sees the ground, the curb, soda cans, etc.) so, for example, if the car in front of you suddenly moves out of the the lane, due to a stopped car (or firetruck!) in front if it, your car is less likely to determine that stopped car is really there, as the radar says no, and the vision hasn't had enough time to make a true determination.

I think the combination of HW3 and unreleased FSD software makes this a temporary situation that will likely improve dramatically in the next few months. You can tell from the Autonomy day video that the new software has much better detection of 'drivable space' and the drivable space excludes parked vehicles on the side of the road. The new software renders both the stationary vehicles and the drivable space, and that was almost a year ago (April 2019, 'Autonomy day').
 
This is a bit weird though. If there's such high demand for US made Model 3s in China, why wouldn't they just dedicate some more vehicles to China in Q1 or Q2?

Do they have to book ships/containers far in advance?

The only thing I can think of is that MiC M3 margins are higher, so they might want to encourage people to buy those instead.

Also implies to me that there's enough demand elsewhere to sell all of US production, so let them stick with their home-made stuff for the moment. If this is correct then it really drives a stake through the heart of the "no-demand" arguments.
 
  • Like
Reactions: FrankSG
Can they sell a lot of ~100 mile range small cars compared roomy more capable, 220+ range model 3 costing only ~5k more and becoming more available in that market?

It may be valid to assume there is a limit to sales well below the aspirational run rate of 1500/week by end of Q2.
I don’t claim to understand European car preferences. Other forum members can chime in.
 
  • Helpful
Reactions: Foghat
Yup, I’m hoping you’re right, I want to keep accumulating, so I just made my 2020 IRA contribution, and put in a limit order just above max pain that I hope will trigger.
... and it did! At 561.25 - this is why I like limit orders (even if the big guys make a few cents on my transaction, I saved about $36 versus a market order first thing this morning).
Not an advice, but every time I feel all chuff about saving a few pennies before I expect the stock to skyrocket, it crashes instead (happened before a few ERs that I thought were going to be killer and during the funding-secured roller coaster). So this is all part of my master plan to pick up a lot more shares when my annual bonus gets paid next month ;-).
 
OT:
Don’t be ridiculous, Otto is a registered trademark already.
How about OtterPilot, because, they are cute aren’t they?

Otto is also a name, so why not Mr. Otto Pilot? Then it becomes a robotic assistant.

Or Sir Otter Pilot. Fancy, but yet still adorable.

And the best thing is, Overlord Musk might actually change it to something similar if the Senator gets any traction with his complaints. God, I love Tesla.
 
I feel pretty good about this. I was worried.
1) Rename Autopilot - how about Otto Pilot?
2) More steering wheel sensors: OK.
More steering wheel sensors? Everyone informed. Using cheat devices (weights for instance) tells you they know what to do. There are laws about running red lights and stop signs. When drivers do that they are punished and sued for the damage they do. We don’t blame the city/town for placing these. Or how about people using electric hair dryer in the bathtub? Do we blame the manufacturer or ban hair dryers?