Electric Vehicles - Models Discussion

[folivier]

Auto wipers are a nightmare during love bug season. Conventional wipers can be a nightmare also if you make the mistake to use them when you hit a flock of those devil bugs. Best thing is to wait until the morning dew has softened them up then use your washer and wipers copiously.

 

[audreyh1]

You can certainly set auto wipers to off.

 

[ken830]

I've read complaints that Tesla's rain-sensing wipers through software (and computer vision) is inferior to the old analog system that other cars use. It's a commodity part from companies like Bosch.

Or Tesla gets rid of the ultrasonic sensors used for parking to save some bucks but it's better to have actual sensors.

Same to a larger extent for Lidar that other companies use for self-driving instead of relying only on cameras.

Vision-only (or vision-primarily) is the correct approach for first-gen autonomous vehicles. And it may be the correct approach overall. Lidar is pretty much useless. It doesn't provide any necessary data with which to drive a vehicle.

 

[jabbahop]

Vision-only (or vision-primarily) is the correct approach for first-gen autonomous vehicles. And it may be the correct approach overall. Lidar is pretty much useless. It doesn't provide any necessary data with which to drive a vehicle.

I wonder why Waymo is spending all the money on Lidar on its self driving fleet if it is useless?

 

[EastWest Gal]

I just took my first ride in a Tesla. The driver was a friend in his 70s; he, his wife and I were riding to a memorial service together. He only has had the car two months, so he is in a learning curve. One thing I observed is that he tended to drive too fast. I think he is so used to the sound of an ICE that he did not realize how fast he was going, since the electric car is relatively quiet. It was a scary ride and there were a few close calls in the 25 minute ride. I have his wife's number and may talk to her about it.

Have any of you observed the tendency to drive too fast in an EV?

 

[DayDreaming]

I just took my first ride in a Tesla. The driver was a friend in his 70s; he, his wife and I were riding to a memorial service together. He only has had the car two months, so he is in a learning curve. One thing I observed is that he tended to drive too fast. I think he is so used to the sound of an ICE that he did not realize how fast he was going, since the electric car is relatively quiet. It was a scary ride and there were a few close calls in the 25 minute ride. I have his wife's number and may talk to her about it.

Have any of you observed the tendency to drive too fast in an EV?

With my last car, I really had to floor it to accelerate, like when merging onto a highway. When I got my Tesla, flooring it was a hard habit to break and I often found myself speeding before I realized it. You can set the car to sound a chime when you go more than x mph over the posted speed limit. I set mine to chime at 12 mph over the limit. I think your friend needs to do this!

 

[joesxm3]

I noticed that when using FSD I was able to comfortably go 10 mph than I felt comfortable going on the same roads with me driving. That is, I tended to drive 10 mph slower on manual.

 

[audreyh1]

I just took my first ride in a Tesla. The driver was a friend in his 70s; he, his wife and I were riding to a memorial service together. He only has had the car two months, so he is in a learning curve. One thing I observed is that he tended to drive too fast. I think he is so used to the sound of an ICE that he did not realize how fast he was going, since the electric car is relatively quiet. It was a scary ride and there were a few close calls in the 25 minute ride. I have his wife's number and may talk to her about it.

Have any of you observed the tendency to drive too fast in an EV?

Not me.

And I have speed limit chimes turned on at 4 miles above the current limit. The car does a good job of reading most speed limit signs.

I pretty much use FSD these days and it drives exactly the speed limit unless I override it.

It does accelerate faster to that speed limit than I normally do, but perfectly safely.

 

[explanade]

I wonder why Waymo is spending all the money on Lidar on its self driving fleet if it is useless?

Yeah I believe Waymo also has a better record than FSD as far as needed interventions and accidents per million miles.

Also I think most other self-driving systems integrate Lidar and other sensors. FSD is out on a limb as far as vision-only.

Considering the limitations of vision-only systems in reduced visibility conditions due to weather, it appears every other company in this space is using Lidar.

 

[Midpack]

I just took my first ride in a Tesla. The driver was a friend in his 70s; he, his wife and I were riding to a memorial service together. He only has had the car two months, so he is in a learning curve. One thing I observed is that he tended to drive too fast. I think he is so used to the sound of an ICE that he did not realize how fast he was going, since the electric car is relatively quiet. It was a scary ride and there were a few close calls in the 25 minute ride. I have his wife's number and may talk to her about it.

Have any of you observed the tendency to drive too fast in an EV?

Maybe for a few days at first, but I pay attention to my speedometer so I quickly got a feel for reasonable speeds. That and I use TACC a lot. (Traffic Aware Cruise Control).

 

[TrvlBug]

I just took my first ride in a Tesla. The driver was a friend in his 70s; he, his wife and I were riding to a memorial service together. He only has had the car two months, so he is in a learning curve. One thing I observed is that he tended to drive too fast. I think he is so used to the sound of an ICE that he did not realize how fast he was going, since the electric car is relatively quiet. It was a scary ride and there were a few close calls in the 25 minute ride. I have his wife's number and may talk to her about it.

Have any of you observed the tendency to drive too fast in an EV?

My EV won't let me forget if I'm speeding. The reading is huge, in my face, in the instrument cluster w/ a smaller one that blinks when I'm over the speed limit. Or I can look at my speed and blinking in my HUD. Or if DH is with me, he can see it (it's that big) and tells me 'do you know how fast you're going' or something to that effect .

 

[homestead]

I don't drive any faster with my EV but I sure get up to speed quicker than before especially going through traffic signals.

 

[audreyh1]

I don't drive any faster with my EV but I sure get up to speed quicker than before especially going through traffic signals.

Yeah, that’s where FSD leaves the other cars in the dust, ha ha.

 

[ken830]

I wonder why Waymo is spending all the money on Lidar on its self driving fleet if it is useless?

Yeah I believe Waymo also has a better record than FSD as far as needed interventions and accidents per million miles.

Also I think most other self-driving systems integrate Lidar and other sensors. FSD is out on a limb as far as vision-only.

Considering the limitations of vision-only systems in reduced visibility conditions due to weather, it appears every other company in this space is using Lidar.

You guys can wonder all you want, it doesn't change the fundamental fact that driving a vehicle requires solving vision. Think about what you need to drive a vehicle. Think about what Lidar gives you. Lidar can only give you a depth map in one plane. And only if it's mounted above the car. Otherwise, you need it all around the car. And then what do you do with the depth map? If it shows an obstacle, you can't make a driving decision based on that because you need to identify it. And the only way to do that is visually with a camera and a neural net. Lidar data isn't enough to tell you if that's a solid object or just a plastic bag or dense fog/water vapor from a manhole cover. And another nearby lidar shooting out photons will just confuse it. Imagine a roadway full of lidar cars. They are constantly shining lasers out causing interference. So you need to solve vision. And if you solve vision, then where does that leave Lidar? What does it give you? Absolutely nothing.

And the depth information you get from Lidar, you can get that (and much more) from vision. It's all encoded in video. The proof? You and I. And every living creature with a neural network in it's brain that can navigate it's environment by vision. That's the proof that it's technically possible.

Waymo only operates in a tiny fraction of the globe. It's a highly controlled environment that's fully mapped. And Lidar can give you precise SLAM in a very easy manner. But this is a local maximum. It's a crutch. And they have remote operators that help the vehicles when they encounter unusual situations. They started service at the end of 2018. Now, 6 years later, they have a total of just 700 vehicles in three markets. They are bleeding money. If they were profitable, they would scale up. But they can't. Because they would lose too much money too quickly.

Tesla is not ready either, but their solution is scalable. The evidence is in the fact that they sell 2 million vehicles each year. And every single one of those come with all the hardware for self driving whether people pay for it or not. And they are profitable, not just on per unit gross profit level, but net profitability. The problem they need to solve is a software problem.

And it works incredibly well. Well enough that any random Joe can just use it anywhere in the world with the disclaimer that they just supervise it. And the evidence it works well (not good enough yet, but well) is the fact you don't have reports of hundreds or thousands of people crashing and dying left and right while on autopilot or FSD. Even in regions where Tesla doesn't offer FSD, people have imported vehicles that have the feature and it just mostly works even if it isn't optimized for certain regional road regulations.

Waymo has a hardware scaling problem. They have a software problem too, but the bigger issue is the scaling. They have the wrong approach, but let's say they have the right approach. They can't deploy it because they wouldn't be able to afford it. And so, by definition, they can't have the right approach.

I've always been super optimistic about how advanced AI is, but working in the AI industry has really shocked me with how much more advanced it really is. It's the most influential technology ever. And solving driving is just a small piece. But it'll be so advanced, it can drive a car with just camera steams. It can't drive a car with just lidar. Or any other set of sensors. So the solution is to use the most scalable sensor suite. Show less

 

[RetMD21]

IDK if y'all have seen the Waymo cars honking at each other at 4 AM

 

[audreyh1]

Yeah - that shows up on social media often.

 

[rembrandt]

3 killed, 1 seriously injured in Tesla Cybertruck crash in Northern California.​

By AP​

Wednesday, Nov 27, 2024 2:07 PM Updated Wednesday, Nov. 27, 2024 4:02 PM
PIEDMONT, Calif. (AP) — Three people were killed and one seriously injured early Wednesday when a Tesla Cybertruck crashed and caught fire in Northern California, officials said.
Piedmont Police Chief Jeremy Bowers said speed was likely a factor in the single-vehicle collision, but he added that authorities would continue to investigate the circumstances of the tragedy.
The Cybertruck's occupants had been driving back from a “function” around 3:10 a.m. Wednesday, but Bowers would not give additional details. Someone who also had been at the function, but was driving in a separate car, was able to pull the surviving victim out of the Cybertruck.
The chief would not release any identifying details of the victims, citing family privacy.
Although the police officers who responded were unable to douse the flames with their fire extinguishers, the fire department put the blaze out quickly.
Piedmont Fire Chief Dave Brannigan said the speed in extinguishing the flames meant it was unlikely that the Cybertruck's large lithium-ion battery had caught fire. He called the incident “more along the lines of a typical car fire.”

Bowers, addressing a question about recent Cybertruck recalls, said “there’s no indication that there were mechanical effects that were the primary cause for the collision.”
The Cybertruck, which has been on sale for about a year, has been recalled six times for safety problems, most recently on Nov. 5 because a fault in an electric inverter can cause the drive wheels to lose power.
Other recalls covered rearview camera images that don’t activate immediately after shifting into reverse, trim pieces that can fall off, windshield wipers that can fail, and an incorrect font size on instrument panel warning lights. In April, the futuristic trucks were recalled to fix gas pedals that can get stuck in the interior trim.
The city of Piedmont is about 2.5 miles (4 km) northeast of Oakland.

 

[Midpack]

^^ There are fatal crashes every day, on average 1 every 4.5 hours in the USA, why highlight that one? Speeding at 3:10am, the make/model vehicle probably didn't have anything to do with it...

 

[JoeWras]

Agree. Crashes happen with all kinds of vehicles. The saying is "Nothing good ever happens after 2 AM" and this is a prime example.

For us, we try to stay off the roads after 9 PM. Too many shenanigans out there.

 

[rembrandt]

Deadliest Tesla Cybertuck Crash Yet Highlights Car’s ‘Horrifying’ Safety Concerns.​

A fatal Tesla Cybertruck crash in the San Francisco Bay Area early Wednesday morning killed three people and left another injured.​

Tesla’s giant “refrigerator on wheels” car model quickly became the butt of jokes after the company finally began delivering the futuristic trucks in November of 2023. The polarizing design made the trucks become darlings of social media and online video engagement overnight. But the truck’s ascent into the cultural zeitgeist helped reveal an almost immeasurable amount of flaws in the final product that shipped out.
The cool-looking aerodynamic wheel covers caused damage to the tires; the stainless-steel body panels became pitted and rusted in the rain; the massive lightbar had alignment issues; the body panels were not rounded out — and their edges were so sharp, you could shave vegetables against them; the “frunk” panel had no sensor to detect whether or not an object would be in the way of it closing … which, combined with those razor-sharp edges meant a finger could easily be cut off. These are only the ones that came to mind.

 

[Texas Proud]

Just want to comment on this... had to copy...

And the depth information you get from Lidar, you can get that (and much more) from vision. It's all encoded in video. The proof? You and I. And every living creature with a neural network in it's brain that can navigate it's environment by vision. That's the proof that it's technically possible.


Well, bad analogy... I have seen people on their phones run into others... I have seen people run into parked cars... I have seen people run a red light or a stop sign... vision is important but not the only thing that is needed IMO..

Not sure, but wasn't it only vision that ran over the lady pushing her bike across the street? The car missed it but if you say the video it was clear...

 

[Midpack]

Not sure, but wasn't it only vision that ran over the lady pushing her bike across the street? The car missed it but if you say the video it was clear...

It was an Uber Volvo XC90 with vision, radar and LiDaR with their own software. The sensors weren’t the problem anyway, it was the software and the Uber driver, and the pedestrian. It’s easy to search first Death of Elaine Herzberg - Wikipedia

 

[Zathras]

Deadliest Tesla Cybertuck Crash Yet Highlights Car’s ‘Horrifying’ Safety Concerns.​

A fatal Tesla Cybertruck crash in the San Francisco Bay Area early Wednesday morning killed three people and left another injured.​

Tesla’s giant “refrigerator on wheels” car model quickly became the butt of jokes after the company finally began delivering the futuristic trucks in November of 2023. The polarizing design made the trucks become darlings of social media and online video engagement overnight. But the truck’s ascent into the cultural zeitgeist helped reveal an almost immeasurable amount of flaws in the final product that shipped out.
The cool-looking aerodynamic wheel covers caused damage to the tires; the stainless-steel body panels became pitted and rusted in the rain; the massive lightbar had alignment issues; the body panels were not rounded out — and their edges were so sharp, you could shave vegetables against them; the “frunk” panel had no sensor to detect whether or not an object would be in the way of it closing … which, combined with those razor-sharp edges meant a finger could easily be cut off. These are only the ones that came to mind.

This is a blatant hit piece.
It is loaded with adjectives meant to elicit fear.

We know nothing about the crash other than:
A) It happened shortly after 3am
B) Excessive speed was involved.

The truck ended up wedged between a sturdy tree and retaining wall.

Pedaling fear is not helpful. Let’s wait to hear what the investigation finds out.

 

[RetMD21]

Deadliest Tesla Cybertuck Crash Yet Highlights Car’s ‘Horrifying’ Safety Concerns.​

A fatal Tesla Cybertruck crash in the San Francisco Bay Area early Wednesday morning killed three people and left another injured.​

Tesla’s giant “refrigerator on wheels” car model quickly became the butt of jokes after the company finally began delivering the futuristic trucks in November of 2023. The polarizing design made the trucks become darlings of social media and online video engagement overnight. But the truck’s ascent into the cultural zeitgeist helped reveal an almost immeasurable amount of flaws in the final product that shipped out.

Although there wasn't a link posted I was able to find it. Deadliest Tesla Cybertuck Crash Yet Highlights Car’s ‘Horrifying’ Safety Concerns

If you read the article, it cites the horrifying crash test results but the link actually goes to a video debunking those claims.

The article further states:

According to Piedmont police, the fiery crash occurred just after 3 a.m. on Hampton Road in the Bay Area city. First responders arrived on the scene to find a Cybertruck — the front of the vehicle wrapped around a tree and the vehicle completely charred. It was reported that an emergency notification recorded by an iPhone pinged a nearby emergency call center and prompted first responders to the scene within minutes of the crash. However, firefighters noted the blaze was “too intense” to appropriately put out when they reached the crash; the blaze was completely snuffed out at 3:16 a.m. by Piedmont firefighters.
In other articles it looks like the police failed to put out the fire but it was no problem for the fire department. The timeline supports that.

Also in the article:
What’s also undeniable is how the Cybertruck’s lack of crumple zones and overly rigid body structure aided in the crash’s deadly nature.

A bold assertion with no factual basis

I'm not a fan of the Cybertruck's looks and It isn't something I would buy but it is easy to be taken in by the anti Tesla media.

Sorry I was composing and missed the post by Zathras making the same points

 

[ken830]

There were several fatal crashes in the past few days in the SF Bay Area that were bad enough to make the news. Not a single one mentioned the make or model of the vehicle in the headline and only one mentioned the make and model in the article. If a Tesla was involved, you can bet your entire net worth that it will be prominently featured in the headline. If you can find someone to take the other end of that bet, you can definitely retire early.