-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
- Status
One thing to consider about the rapid disruption assumptions is that cars are not phones, cameras or software, they are massive objects which need a much much larger infrastructure (beginning with mining). So I'm unsure how fast the electric car supply chain will be able to grow.
ValueAnalyst
Closed
ATTENTION: This is the third time Elon is referring to tires.
I know some here disagree, but Tesla may be working on long-range high-performance tires, which would make sense for Tesla Network.
I know some here disagree, but Tesla may be working on long-range high-performance tires, which would make sense for Tesla Network.
jhm
Well-Known Member
Musk seems to be suggesting that 0.1Wh/m or 160 Wh/mile is possible. This would be down from 320 Wh/mile or so. This seems incredible that the right set of tire could double the efficiency. Even a 10% improvement would be a real victory.
ValueAnalyst
Closed
This makes more sense, but this is also not "consistent with what they said before" which was "5,000 per week at some point in 2017."
With this hire at least he has the right person in terms of ai computer vision.
Now applying that to autonomous driving is so trivial anybody can do it,
I presume , ok just kidding.
The prior hires never made sense and it quickly revealed itself,
What lead Elon to hire them is a mystery.
Last edited:
jhm
Well-Known Member
Right, that was the context of the quote. Hypermile record at 900km on 100kWh could be bumped up to 1000km or more. So a 10% increase due to tires.
A different kind of run flat that is essentially a spoked plastic lattice/tire inside the pneumatic casing could do that and help with tire noise.
Current best tires might do it, too.
Hopefully the new hire as head of the Autopilot program is an improvement. The program is clearly a mess. Misjudging the actions of Mobileye after prematurely betting on their own developments, losing key people to competition (both established and startup), changing leadership seemingly every 6 months and missing self-imposed deadlines one after another. The next one is coming up fast : it's been 5 months since Elon tweeted that the first FSD feature diverging from EAP should arrive in maybe 3 but definitely 6 months. At this point I am willing to bet that the SF->NYC fully autonomous drive is not going to happen before the end of the year.
Meanwhile, the competition has gotten serious about self-driving. Much more so than about electrifying their offerings. Including investing in silicon valley style startups (which is the way to go) and backing those groups with the billions in R&D money needed to get somewhere. Tesla is front of the pack with AP1, but at the speed it is wasting that lead, things are not looking good. We all know computing hardware has a limited shelf-life. The longer it takes for FSD features to materialise, to higher the chances are that Tesla will have to replace their computing hardware on cars out there because it can't keep up with newer hardware iterations in younger car models. Or worse from a financial point of view : that Model 3 buyers will opt to forego the FSD option which is key in bringing up gross margins as quickly as possible.
Meanwhile, the competition has gotten serious about self-driving. Much more so than about electrifying their offerings. Including investing in silicon valley style startups (which is the way to go) and backing those groups with the billions in R&D money needed to get somewhere. Tesla is front of the pack with AP1, but at the speed it is wasting that lead, things are not looking good. We all know computing hardware has a limited shelf-life. The longer it takes for FSD features to materialise, to higher the chances are that Tesla will have to replace their computing hardware on cars out there because it can't keep up with newer hardware iterations in younger car models. Or worse from a financial point of view : that Model 3 buyers will opt to forego the FSD option which is key in bringing up gross margins as quickly as possible.
"Next stop 270" (Not an advice)
Dynastar
Member
Agreed. At this point I wouldn't suggest paying for FSD to a model 3 buyer who might be stretching to get a $35k+ car. I realize there's been a lot of furious paddiling over the last 12 months but the duck isn't moving much.
DurandalAI
Member
That kind of *sugar* ticks me off. Yes, it's a cool development, but they make no mention of the efficiency. Why would you convert sunlight to hydrogen if the efficiency is say, 5%? (Speculation, we don't know the numbers.) Maybe if the cost is very low. If solar PV is 20% efficient, and electrolysis is 33% efficient, then solar PV to hydrogen is 6.66_ efficient. Still better than the paint, but what is solar PV to BEV efficiency? 14-16% efficient? (Again, all just hypothetical numbers.) But when no reference is given to the efficiency, you can bet it's not even worth mentioning (because it wasn't.)
Well, that would be great but it would be a first for Tesla/EM when it comes to exceeding a production/delivery timeline and number.
I hope I am pleasantly surprised (wrong, if you like) but though I am not from Missouri I will say, *show me*
DurandalAI
Member
I hope that a subscription service, or possibly Tesla Mobility subsidized FSD will be offered. (Tesla Mobility 10:1 ratio, for every 1 hour your car does Tesla Mobility, you get 10 hours of FSD for yourself!) Something like that, but more "simple" as Tesla doesn't seem to like complicated calculations to be applied to customers. An optional $100/mo for FSD until your $5,000 is paid of would be reasonable, the balance of paid/unpaid is transferred to the next owner as well for residual value.
T3slaTulips
Member
Yes, saw that. I was talking about the over the air "wide" release. On 11th Musk said it will go out next weekend with additional improvements. That would be 17/18th. Still didn't happen afaik. My guess is that their shadow mode data wasn't as good as hoped. So they are still tweaking... Coming from the software world, you can't really put a deadline on algorithmic tweaks as a lot of times it's more of a trial and error. So I can understand.
- Status
Similar threads
- Sticky
