I have zero experience with any self-driving cars.
I actually suspect that a self-driving car that needs an intervention, even a minor one, every 5-ish minutes is far safer than one that needs an intervention once hourly or worse yet, once every 100 hours.
If a driver needs to keep fiddling with it often, they’ll learn to keep paying attention adequately. If they don’t need to fiddle with it but once every month or so, real soon they’ll learn not to pay attention. Then shit will happen.
Yeah. At least if said required interventions don’t come with very ample warning. There is a sort of middle ground where the car might turn on its hazards and come to a halt (the Mercedes does this, I believe), which is certainly safer than plowing into the unknown object ahead, but still not really safe (like stopping for a ladder in the fast lane of a highway).
Well, it’s gonna be fun to see how this plays out further. FSD v12 is around the corner and uses end-to-end neural nets. Which should work out much better than the current spaghetti code, but be even less inspectable for bugs.
Sure, in its current iteration. But I’m not really talking about safety. I think Autopilot and FSD, despite their faults, are both much safer than human drivers when used properly.
The question was, is Musk honest and accurate in his naming of the features, his description of their capabilities, and his estimates of when certain things will be released? And I think the answer in unequivocally no. Calculating the yardage required to change lanes when you’re talking about the % of time on Autopilot might be literally true, but it is highly deceptive.
It’s akin to me saying “I’m not touching you” when I held my finger in front of my sibling’s face. Technically true, but it’s ignoring what people are really talking about.
There’s certainly several convos going on in this thread and other threads about the hype-reality disconnect out there regarding all forms of self-driving cars, not just Teslas.
My overall attitude is the Tesla features are misleadingly named and have been overpromised and underdelivered since Day One and still are both. And there seems to be plenty of anecdotal evidence that many people are misusing the features they do have.
How much the hype-bro propaganda causes or influences that bad behavior above and beyond the simple “people are dumb and lazy” reason that explains so much of human behavior in bulk is not a question I’m competent to answer. Some? Sure. Enough to move the statistical needle or attach civil liability for crashes or criminal liability for fraud? I bet not (much).
YMMV of course.
Tesla Autopilot will change lanes on its own, therefore your whole analogy is broken! (Kidding). Seriously, when I used it this summer for a long road trip it was remarkably good at changing lanes to pass people, and then getting back in the right lane.
Here is a list of common interventions roughly in order that I experience them, from most common to least:
The first 3 do not require disengaging autopilot. 4 and 5 will disengage autopilot.
Exactly how that breaks down into 90%, 98%, etc. is going to be open to lots of interpretation.
Speaking of which:
I don’t do off-road stuff so I couldn’t say how impressive that is. More than my Model 3 can handle, though 
.
Looks like something Quigley might do. I have a friend with a Ford Econoline 4x4(6?) dualie passenger van that they did, somewhat similar to this & carries a lot of weight behind the rear axle.
AWD was an option in the Astro in 1990.
As for the Tesla’s lack of a frame, it remains to be seen. The advantages of body-on-frame include high towing capacity and ruggedness. A unibody generally can’t tow as much as it can’t distribute the weight as well. A body on frame vehicle can withstand more damage qithout destroying structural integrity. Generally even rust won’t kill a body on frame vehicle, whereas rust can cause a unibody vehicle to be scrapped. I know, I scrapped a 240-Z because of unibody rust.
However, the Cybertruck isn’t really a unibody. It’s got a huge stamping that’s very strong, and the body panels are structural. It seems to have no trouble towing huge amounts of weight (other than battery life), and the stainless construction won’t rust. It’s really a semi-monococque design like light aircraft, which have load bearing structural members reinforced with stressed skin.
How it will hold up over time and against the rigors of truck use, time will tell.
Now, let’s talk about the disadvantages of body on frame. A big one is rigidity. Anyone who has driven a pickup for any length of time can tell you about the banging, squeaking, and juddering that happens in a truck when you go over bumps or ride offroad. Modern suspensions control this somewhat, at a cost. But no one would ever confuse a truck with a unibody car in normal driving.
Cybertruck is stiff. As stiff as a McLaren. That translates into better handling, the ability to tune the suspension better, a car-like ride, and less wear and tear on components. I would personally put a significant value on that, having driven pickups quite a bit.
When I first heard of Cybertruck, I was really down on it. Not because of looks, but because it seemed about as useful of a truck as a Honda Ridgeline. But that was before they added drive by wire, 4-wheel steering, air suspension, etc. IMO these features are real game-changers, not just gimmicks.
Watch the Hagerty video where he takes the Cyberteuck around a go-kart track. In a conventiinal truck, getting around the 180 curve requires a 12-point turn. The cybertruck just did it without having to reverse, no problem.
But more to the point, the comment later in the video was that if you took a regular truck around the course you’d be a sweaty wreck at the end because of all the steering work you’d be doing. A regular truck has typically 3-4 turns of the wheel lock-to-lock. In the Cybertruck it was effortless, and you can do max turns with a 180 degree turn of the steering wheel. You can go lock to lock without taking your hands off the wheel. If you’ve ever had to manoever around a tight space in a pickup, you can sympathize.
The Jury is still out. For $100,000, it’s a niche truck. It’ll probably do fine as a niche truck for outdoorsy adventure types with money. But to reach fleet sales and make a dent in the work/farm truck market, they need a cheaper vehicle. For example, even though Ford makes some ridiculous trim levels for the F-150, the biggest seller trim level by far is the XLT, their second-lowest trim level. In the upper trims, the Lariat is most popular, and it’s far from the kost expensive.
If Cybertruck has to compete in a Raptor-sized market, it’ll fail. Ford sells between 700,000 and 900,000 F-150’s per year. Of that, around 10,000 are Raptors. so, maybe 1-2% of production. Add in the other high end >100k trims and you’re probably still only 10% of the market or so as a guess. The big sales are to fleets and workmen/farmers.
We get the idea that there are a lot more of these silly trim level trucks probably because they are what you find in the cities in grocery parking lots and such.
That actually seems worse for sustaining damage. At least with a unibody chassis, the fairly soft body panels are only transferring their out of shape problems to the interior panels that are generally thought of as the structure. In this case, the body panels are the structure, and the fact that they’re really rigid seems they’ll transfer their problems to each other more efficiently. Still seems like a likely write-off once you bend up the body.
And if you roll an aircraft along the ground on something other than its wheels, what do you usually do with it?
I’d require a cite for this. The comfort, rigidity and noise seem to be completely independent of the construction methods. You can make a body-on-frame ride just as quiet and comfy as a unibody. Otherwise I’d think Rolls-Royce would have probably dropped body-on-frame for the Phantom before 1990. It’s just that the vast majority of passenger cars made since the 60s are unibody, and no one is really concentrating on making really comfy pickups. Unibody and monocoque are relatively light and cheap, and I haven’t seen any other convincing advantages.
You know, that’d be interesting if I was buying a truck to negotiate a go-kart track. It’s the wrong tool for the job. I’m fairly confident my BRZ would top it in that test, and a go-kart would beat my BRZ.
The rigidity is basic physics. A ladder frame has a small cross section compared to a unibody, and thus a smaller moment of inertia (for the same mass cross-section). The rigidity of a beam is directly related to the moment of inertia. Same reason why a thin-wall tube is far more rigid than a solid rod for the same mass.
Yeah, but that really is only a consideration when you’re worried about weight. Even then, lots of sports cars with tube/ladder frames are plenty rigid for their weight, since the body doesn’t really have to weigh much at that point (they’re only rigid where it counts). I’d be really surprised to find out that a F-350 chassis is more flexible than that of a Honda Fit.
And deform that tube just a little bit, and its structural rigidity is generally gone.
Well, sure. Though if you spend a little of the mass on cross-bracing, it still comes out way ahead.
I’d take that bet on F-350 vs. Honda Fit in an apples-to-apples test. I.e., normalize for mass and measure deflection from a specific load. Pickup trucks are floppy. Check out this video, for instance. Some trucks do better than others, but they’re all pretty bad.
Surely, and you can get similar benefits from expanding out a ladder chassis into a box frame.
Ehh, that’s probably a bad example, and it’s really still a question of function. Seems like they’re flexting where they’d need to flex, honestly. I’m sure they could engineer that out if they cared and it was cost effective. The important part as far as a truck goes is that is returns to where it’s supposed to be afterward. A unibody that flexed that much would probably be done.
Yeah, but a unibody wouldn’t flex that much in the first place. Truck beds are separated from the cabs because they’d be ripped to shreds if they were. But a unibody/monocoque/exoskeleton/whatever design wouldn’t flex nearly as much for the same applied force, and likely remain undamaged even with far higher forces.
Assuming it’s designed right, of course. So we don’t yet know for sure how it’ll play out with the CT, but neither do we have any reason to believe it’ll have any problems. I’m sure someone will do the tests someday.
Another aspect to rigidity is that it’s harder to get into bad resonance modes. In the video I linked, that Toyota was looking seriously close to twisting at its resonant frequency. Increase the overall stiffness by 10x or 100x and that basically can’t happen.
thats actually quite impressive - assuming a stock vehicle … you can tell, its an experienced driver w/ lots of control (as opposed to gunning it)…
my off-the-pantsometer-call: at least on par with other stock 4x4 monster-trucks, possibly better
does the CT have torque-vectoring (sending more power to a wheel with grip than one already slipping?)
I’ve done a lot of off-roading. Mostly in short-bed trucks and Jeeps.
I don’t do it much anymore but I would think that my current 4Runner could do that. It has low range, so plenty of power. Also the traction control is really great. Also has a locking rear differential (which I really don’t use because the traction control is so good).
I have the vehicle for snow. And it’s just a good size for what I do.
Nah, we have lots of reason to think the CT will be generally be a write off once you bend it up beyond a fender bender: every unibody vehicle that preceded it. I haven’t seen any reason to doubt that will be the case.
Bend up the bed of your Tacoma or F150 while playing excite truck? Take it off and drive without one if it offends you, since it’s not structural and is just a box attached to the frame. That’s not an option with the Cybertruck.
In my experience, Teslas (maybe all EVs?) have phenomenal traction control, be it on a dragstrip, dirt, or snow. You can floor the accelerator all you want but the car will only allow a dribble of power if traction is very limited.
The CT driver in the video is probably one of their test drivers but they could have achieved the same climb using any assembly line worker that has a right foot.
EVs, especially multi-motor EVs, have a slight intrinsic advantage when it comes to traction control, but I’d say that “phenomenal traction control” is standard on all modern vehicles regardless of powertrain. I’m not into off-roading but I do amateur auto racing and instruct at track events, and the progress I’ve seen in traction control systems over the last 10 years, even on pedestrian vehicles, has been staggering.
I’m assuming that translates to off-roading as well.
Right?
Like, Tesla famously doesn’t do [much] market research, and they definitely don’t do focus groups. That’s the whole reason something like the Cybertruck exists, it’s focus-group poison that no conventional automaker would greenlight. And in some ways, that’s great! The world gets some cool stuff that way.
What that means, though, is that while Ford spends a ton of money making sure that their engineering efforts will equate to sales (what do truck buyers actually care about), Tesla, sorta like Apple, just lets their engineers design stuff that the engineers (or designers) think is cool. Or, in the case of the Cybertruck as a whole, that Elon thought was cool during an earnings call.
Like Apple, though, that means when your product hits the market, rather than consumers rejoicing that they’re getting what they’ve been asking for, it’s the job of the manufacturer to convince consumers that you actually wanted this all along. And so we get nonsense like putting auto journalists in a quarter ton pickup truck on a go kart track to convince them of how much they wanted 4 wheel steering. “Isn’t this thing nobody was clamoring for GREAT?!?!”
And then like good evangelicals, they’ll go out into the world and proselytize. As I said a few months ago:
I really wish Mazda or some other small automaker had come out with the Cybertruck. It would have sold a few magazines and garnered lots of youtube views, but wouldn’t see many actual examples. Then everyone would forget about it until you saw one at cars in coffee in 20 years and went, “Oh cool, I remember these!” And all of the weird design decisions would be quirky and fun, and stories about how Mazda put journalists on a go-kart track to show off their slick steering system would be funny instead of annoying.