DanCar
Active Member
The tear down report I saw said a high end iphone cost less than $200 to manufacture.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Autonomous Car Progress
- Thread starter stopcrazypp
- Start date
-
- Tags
- Autonomous Vehicles
mspohr
Well-Known Member
I think the guy that I saw must have been referring to the Foxcon assembly cost only (without the parts).
The IPhone 15 cost $423 to make according to this tear down.
iPhone 15 to iPhone 15 Pro Max, here is how much they cost Apple to make
A new report reveals the amount it cost Apple to make each individual iPhone 15 series handset - iPhone 15 to iPhone 15 Pro Max. It also highlights that the iPhones have become more expensive to manufacture compared to the iPhone 14 series.
www.google.com
Doggydogworld
Active Member
100x, because deep pockets liability awards can easily run 100x as much as individual human driver liability.
That's not an issue in China, obviously, so they can deploy with much lower safety margin.
EVNow
Well-Known Member
Exactly. This is the reason we shouldn't expect Tesla to take liability anytime soon.
I wonder whether we'll see any states enact legislation to limit liability .....
Doggydogworld
Active Member
EVNow
Well-Known Member
Don't want to get political here - but I can see this happen in some states.
We already have limitations imposed by congress on suing a whole host of corrupt businesses - including credit rating agencies that lost all our information to hackers.
OxBrew
Active Member
Show me the raw data, I'm mot convinced.Good article advocating for autonomous cars. He makes the case that we should not wait for AVs to be perfect and that AVs can save lives now:
We Are All Bad Drivers
Anyone rooting against self-driving cars is cheering for tens of thousands of deaths, year after year. We shouldn’t be burning self-driving cars in the streets. We should be celebrating them.www.thefp.com
It will take 40-50 years to build enough of them to make any dent. The article does not even come close to discussing any of the hurdles to implementation, or describe in any detail what benefit will exist prior to a critical mass of safe AV fleets on the road. There will be zero benefit to anyone but the robo taxi firms until a huge chunk of the whole population has access. Until then, they will INCREASE congestion, just like Uber and Lyft do now.
In that time we could have trains and light rail, if we just decided to, actually solving safety, access and congestion. Hyderabad India built out enough light rail in 4 years to actually get traffic moving again on surface streets.
Its wasted resources, a distraction by big auto to keep us from cutting into their bottom line by building real mass transit that works.
Last edited:
Reading the offence that posters in TMC have to leaving following space with the car in front, the upset at FSDb limiting the set speed at 85mph (when actual limits on those roads are much lower), and general entitlement and expectations for all other vehicles to yield to them (regardless of whether or not they are in the right) makes me think that AVs following the same set of road rules would be safer.
But every one of your other arguments introduce reality into the equation. Especially increased congestion, increased effect on the environment (from the manufacturing process due to probable shorter life-span, tire wear pollutants, stress on power grids), and pushing people away from mass transit.
Government decisions/sabotage of mass transit really upsets me. At least in my city, it is a visible 'thing' where they have done everything wrong and destroyed any confidence in mass transit.
I moved to the edge of the downtown from a rural village in order to have access to mass transit and further reduce my environmental footprint. We used to take the bus to the theatre in order to avoid parking hassles. It was longer, but not much longer, my record from standing ovation to walking in my door was 26 minutes.
The bus route at my door took me useful places, and walking 2 long blocks got me to a transit station with a bus that in 45 minutes took me to one block my daughter (least driving time possible for that trip is 22 minutes.) Since she lives car free, she could come visit with the kids in tow any time she wanted. Now that trip is almost an hour, requires a bus that runs less frequently to take her to an LRT station, then transfer by elevator (since she has a stroller for the kids) to the train, then transfer to an every 20 minute bus at my nearest LRT station. Of course, during Covid, she stopped coming altogether since it wasn't allowed and I visited her instead using the car.
During Covid, I bought my mobile isolation unit (tesla) and now drive almost everywhere. Since Covid, due to work-from-home, the buses became less frequent and more expensive so we seldom use buses now. If we are both going downtown and expect to be there more than 90 minutes, it is cheaper to take the car and pay for parking. And WAY faster (it is 10 minute drive vs taking a bus that runs every 30 - 60 minutes.) We are a single car household and if we both need to go different places at the same time, I have signed up for a car share service with a car in my complex. (Which is what I'll use this week when we take the Tesla in for an all-day service appointment.)
The transit decisions in my city have actually driven me off the buses. And as the LRT system expands (now 4 years late and counting but I wonder if I'll live long enough to see it functioning.) The word 'functioning' is loosy-goosy here, the system fails on a regular basis and has been out of service for multiple weeks at a time since the first phase opened. And even when phase 2 opens, it will serve me but not my daughter so the commute to here will still involve a transfer to a bus that runs infrequently.
Our city was served by a bus transitway and both us, with plans of living car-free (me due to age and possibly not wanting or being able to drive any more) moved to parts of town within a few blocks of a transitway. It was a good 4.5 years of everything working to plan until the bus system was destroyed in favour of the LRT. And despite 5 years since that happened, the response have been to further destroy the bus system despite the fact that, due to the long downtimes of the LRT, the next phase has no way of having buses cover for the system when it is down.
As a result, traffic in our city has become insane despite work-from-home, as people switch to personal cars because jobs and schools and daycare pickup are all time sensitive and the bus system is only good for people who don't have to be somewhere at a specific time or have the time to set their arrival to be an hour early in case of cancellations of buses.
This is sadly par for the course for Canada & the US. In some ways I didn’t realize how bad it was until I spent some time in East Asia and Europe. The quality of transit has also gotten noticeably worse post COVID. Canada caters less to cars above all else than the US, but when you’ve travelled, you realize it’s still not great. For some reason we can’t build cost-effective transit either, with the US being the worst for this. It’s not really just about transit but more fundamentally how cities are built and whether historic areas are preserved or razed, and this since the 1950s.
The best hope might be services like Waymo reducing demand for parking and lanes (with pooling) so that space can go back to making cities more livable again. I don’t think we’ll ever have Asian or European quality transit simply because we don’t have the same urban form or even the same construction and operating costs, so the best bet might be autonomous EVs. Reduce the volume of traffic by pooling trips and reducing parking requirements, and you can set aside more space for bike paths, wider sidewalks, pedestrianized areas, etc… The key will be to do this before induced demand fills the roads back up again.
OxBrew
Active Member
Agree woth the concept, but it will take decades, and a change in USA people's relationship to car ownership.
We have MAYBE 10 years to fix this, or literally make large swaths of the planet uninhabitable in our kid's lifetime, or sooner.
Not hyperbole, actual reality.
What can we actually do in 10 years(really, 6-7, but we need some sliver of hope)
Guys, we're f----d.
diplomat33
Average guy who loves autonomous vehicles
1 hr long deep dive by Wayve about their end to end, embodied AI approach:
He gives a ton of interesting examples of situations that the Wayve AV can do.
Here is the chart that illustrates their e2e approach:
I would push back when he says that he hopes the data driven nature of e2e means there can be cross country learning (ie learning to drive in Britain will automatically help model learn in US). There are some big differences between driving in US an Britain. There could be some crossover but I imagine you would need a lot of data training too. I don't think it would be quite as easy as just train e2e to drive in one country and it can drive in another country.
And I feel he is misleading in his intro when describing the "Waymo approach" because he seems to ignore the advances that they have made and describes things that are outdated now. For example, he says the old way is too expensive and time consuming because it requires manually labeling all your perception data, manually creating HD maps that need to be constantly updated and writing heuristic rules. He ignores that Waymo now does auto labeling of perception data and unsupervised learning, HD maps are largely automated and they use ML-first for planner, not heuristic code. But I get that Wayve will downplay the competition in order to prop up their approach as better. Everybody does that to make themselves look better.
But the talk is still well worth a watch for all the info and examples of e2e.
He gives a ton of interesting examples of situations that the Wayve AV can do.
Here is the chart that illustrates their e2e approach:
I would push back when he says that he hopes the data driven nature of e2e means there can be cross country learning (ie learning to drive in Britain will automatically help model learn in US). There are some big differences between driving in US an Britain. There could be some crossover but I imagine you would need a lot of data training too. I don't think it would be quite as easy as just train e2e to drive in one country and it can drive in another country.
And I feel he is misleading in his intro when describing the "Waymo approach" because he seems to ignore the advances that they have made and describes things that are outdated now. For example, he says the old way is too expensive and time consuming because it requires manually labeling all your perception data, manually creating HD maps that need to be constantly updated and writing heuristic rules. He ignores that Waymo now does auto labeling of perception data and unsupervised learning, HD maps are largely automated and they use ML-first for planner, not heuristic code. But I get that Wayve will downplay the competition in order to prop up their approach as better. Everybody does that to make themselves look better.
But the talk is still well worth a watch for all the info and examples of e2e.
Last edited:
Just a Waymo selfdriving example of how what’s relevant can surprise.
I took a Waymo in Scottsdale for a two mile trip, four in the car, out to brunch, to a popular place with minimal parking. It seemed like a good idea.
The trip was straightforward and the car drove well in all situations. No driving issues whatsoever.
But, entry was iffy. It was not simple to figure out where the car would stop and let you in. So we walked quickly, almost ran, 100+ feet to the car to get in when it stopped to wait.
Then we all had to put seatbelts on because the car complained loudly and often when we didn’t do it fast enough, harder with three in the back seat.
Then in the beginning it mentioned that if we wanted to we could ask it to stop and pull over. We ignored that, unwisely, since we were captivated with the driving.
At the restaurant with cars everywhere dropping people off, looking for parking spots, it slowed, I thought it was going to stop, but then it drove on looking for a better place. Then it slowed for another possible spot, and we sadly just waited for it to fully stop. It did not, but drove to the back of the restaurant where it could safely stop without traffic. We exited complaining to each other about the smelly trash cans. Old people, what can I do…
We should have pushed the button to tell it to stop at the front when we had two chances to do that. I realized this belatedly as I was walking to the front of the restaurant.
So entry and exiting are what I’ll pay attention to with future selfdriving trips. The selfdriving is anticlimactic.
My two cents.
I took a Waymo in Scottsdale for a two mile trip, four in the car, out to brunch, to a popular place with minimal parking. It seemed like a good idea.
The trip was straightforward and the car drove well in all situations. No driving issues whatsoever.
But, entry was iffy. It was not simple to figure out where the car would stop and let you in. So we walked quickly, almost ran, 100+ feet to the car to get in when it stopped to wait.
Then we all had to put seatbelts on because the car complained loudly and often when we didn’t do it fast enough, harder with three in the back seat.
Then in the beginning it mentioned that if we wanted to we could ask it to stop and pull over. We ignored that, unwisely, since we were captivated with the driving.
At the restaurant with cars everywhere dropping people off, looking for parking spots, it slowed, I thought it was going to stop, but then it drove on looking for a better place. Then it slowed for another possible spot, and we sadly just waited for it to fully stop. It did not, but drove to the back of the restaurant where it could safely stop without traffic. We exited complaining to each other about the smelly trash cans. Old people, what can I do…
We should have pushed the button to tell it to stop at the front when we had two chances to do that. I realized this belatedly as I was walking to the front of the restaurant.
So entry and exiting are what I’ll pay attention to with future selfdriving trips. The selfdriving is anticlimactic.
My two cents.
diplomat33
Average guy who loves autonomous vehicles
Doggydogworld
Active Member
Thanks for sharing. I don't think the Pull Over button works like you think, it just starts looking for a place to pull over. Which your car was already doing.....
Waymo definitely struggles with pick up and drop off. Uber/Lyft offer much more convenience by violating traffic law, sometimes to the point of being unsafe. I'm not sure what the answer for this is.
Waymo definitely struggles with pick up and drop off. Uber/Lyft offer much more convenience by violating traffic law, sometimes to the point of being unsafe. I'm not sure what the answer for this is.
Just for completeness in the story, the return trip was fine. The Waymo picked us up about where we expected at the restaurant entrance. It recognized the phone that called it and unlocked to let us in. We quickly put on seatbelts, as the car slowly pulled away.
It drove through a lower traffic route, maybe a minute or two longer, and dropped us off where we were picked up during the first trip. Anticlimactic trip.
So we had time to watch the neighborhoods as the car drove and comment on the route. All in all, it was a pleasant trip and I recommend Waymo in Phoenix.
It drove through a lower traffic route, maybe a minute or two longer, and dropped us off where we were picked up during the first trip. Anticlimactic trip.
So we had time to watch the neighborhoods as the car drove and comment on the route. All in all, it was a pleasant trip and I recommend Waymo in Phoenix.
How many people would pay 400 dollars for a round trip ticket from California to Las Vegas on High Speed Rail?
www.thecentersquare.com
LA-Vegas high-speed rail tickets could be $400 round-trip
(The Center Square) - A round trip ticket from Los Angeles to Las Vegas on a planned high-speed rail line could eventually cost up to $400, the project’s developer says.
Speaking as a robotics engineer, when I saw the first video of Tesla's V12 software, I immediately thought that it would be our best hope of handling the nuances of real-world driving. Plus eliminating the need for LIDAR is the best hope for making self-driving cars affordable for the average person (though a number of companies are forking on solid-state LIDAR which would be way less expensive than Waymo's hardware). Interesting to see that Tesla isn't the only company pursuing that line of research.
And as for Waymo: Even if they can mostly automate the creation of their fancy HD maps, the need for them (both making them and keeping them up-to-date) is going to substantially limit the rate at which they can expand.
Here’s something I was thinking about recently: Is Volvo positioned to leapfrog most/all in ADAS and autonomous driving in the next couple years?
Chinese companies such as XPENG and BYD are dedicating thousands of people to work on FSD-like systems and appear that they could end up getting there first.
However these companies will have a tough time coming to US (probably won’t) and will have a tough time entering some other markets. And it may be hard for many US based companies to acquire Chinese made tech.
Volvo on the other hand is already in the US but is owned by Geely from China. So in theory, they could get Chinese tech into EV cars that will be manufactured in the US in the future and they appear to be one of the only prominent car companies positioned to do this.
I’m not holding firm to this. Just thought it was an interesting line of thinking.
Chinese companies such as XPENG and BYD are dedicating thousands of people to work on FSD-like systems and appear that they could end up getting there first.
However these companies will have a tough time coming to US (probably won’t) and will have a tough time entering some other markets. And it may be hard for many US based companies to acquire Chinese made tech.
Volvo on the other hand is already in the US but is owned by Geely from China. So in theory, they could get Chinese tech into EV cars that will be manufactured in the US in the future and they appear to be one of the only prominent car companies positioned to do this.
I’m not holding firm to this. Just thought it was an interesting line of thinking.
Similar threads
- Article
