Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
It can be X going to Z+Y where Z is new improved technology ('better' X) which seems very likely (doesn't prove me right though mind you).
I don't remember but it might be UPS (or another big delivery fleet in the USA), they actually monitor their car fleet for break downs (they have a lot of sensors here and there in the car IIRC) so they can fix it in a garage instead of a costly recovery in a street somewhere, I'm sure they can't predict every breakdown but it works economically apparently.
Isn't this a separate matter from vehicle autonomy?
As in, you can have X, X+Y, Z, or Z+Y ?
Quote:
Originally Posted by Valmond
So more sensors can actually make it less costly anyway.
The time frame comes mainly from that article and other observations I have done, Nissan promises self driving car 2020, GM thinks it seems likely by 2020, Sweden tries out highway self driving today (Volvo), England too test next year, it is not only Google !
I feel we are at a tipping point here, everyone start to race (no pun intended) toward the goal of a self driving car.
Self-driving cars available to the public in 2020, sure, I'll give you that much.
Self-driving cars cheaper than human operated in 2030? I'll give you that as well.
But self-driving vehicles economical by 2020? I don't think so. It takes at least years for something to go from "merely available" to economical. It doesn't happen overnight.
Quote:
Originally Posted by Valmond
The tech exists and legislation starts to adjust, they are just rolling it out slowly so we adapt and 5-7 years from now seems reasonable according to all those statements from the manufacturers but also from plummeting sensor prices and obvious tests and so.
Actually, I think all those new electric cars that have come out since like a year will help push this too, but that could be my imagination.
How much do you expect a complete LIDAR system to cost in 2020, and what maintenance do you expect it to need?
Self-driving cars available to the public in 2020, sure, I'll give you that much.
Self-driving cars cheaper than human operated in 2030? I'll give you that as well.
But self-driving vehicles economical by 2020? I don't think so. It takes at least years for something to go from "merely available" to economical. It doesn't happen overnight.
Once they enter the market it will be interesting to see how fast the price comes down. I suspect it will be faster then you think, especially when you factor in insurance.
Once they enter the market it will be interesting to see how fast the price comes down. I suspect it will be faster then you think, especially when you factor in insurance.
Insurance will never go below $0/month, so that limits how much cost it can offset, which was the whole point of my earlier argument about a maximum cost differential of $7,000.
Isn't this a separate matter from vehicle autonomy?
As in, you can have X, X+Y, Z, or Z+Y ?
Sorry I'm not following you here, please say if there is something I'm not getting.
I got that you thought that 'today car electronics' (X) + tomorrows electronics (Y = sensors and stuff)
is greater than 'tomorrows car electronics' + Y
I think it is reasonable to think that the 'X' part have evolved a lot which means it is not possible to prove X < X'+Y
Quote:
Originally Posted by ncole1
Self-driving cars available to the public in 2020, sure, I'll give you that much.
Self-driving cars cheaper than human operated in 2030? I'll give you that as well.
But self-driving vehicles economical by 2020? I don't think so. It takes at least years for something to go from "merely available" to economical. It doesn't happen overnight.
First: yeah, they are already here
2020: I'm not 100% sure this will happen myself of course but I'm quite sure it will, and in say one or two years I'm kind of confident. Here we are, predicting the future, lets see what happens!
For 2025 probably yes I say because at that time, there will be only computer assisted and computer driven cars to buy. Most probably human driven cars will be forbidden where ever there is heavy traffic. I mean OK 'ol farmer can drive his truck but where there is millions of people transiting, a human driver will just block circulation.
2030: well if it is touch'n go at 2020, then at 2030 it seems obvious if it isn't already too old tech, we think cars is some kind of obvious thing we need but it really isn't.
Quote:
Originally Posted by ncole1
How much do you expect a complete LIDAR system to cost in 2020, and what maintenance do you expect it to need?
I'll try to look that up, it was in one of those books, 80.000$ at some point when DARPA started their autonomous car challenge and 200$ from somewhere but I can't prove that.
My bet though is that at 2020 a LIDAR will cost about 10$ to 1$. The maintenance will be about nothing, it will function for tens of years (say between 10-20) and it will 'tell you' when its functioning is declining.
A car will have several of course.
Insurance will never go below $0/month, so that limits how much cost it can offset, which was the whole point of my earlier argument about a maximum cost differential of $7,000.
It will be interesting to see how much driverless cars cost then.
Quick-change materials break the silicon speed limit for computers
This is yet another example of computers getting smaller and faster. Honestly as much as I study it sometimes it sill amazes me at the kind of changes we are now seeing.
This is from Phys.org:
(Phys.org) —Faster, smaller, greener computers, capable of processing information up to 1,000 times faster than currently available models, could be made possible by replacing silicon with materials that can switch back and forth between different electrical states.The present size and speed limitations of computer processors and memory could be overcome by replacing silicon with 'phase-change materials' (PCMs), which are capable of reversibly switching between two structural phases with different electrical states – one crystalline and conducting and the other glassy and insulating – in billionths of a second. Modelling and tests of PCM-based devices have shown that logic-processing operations can be performed in non-volatile memory cells using particular combinations of ultra-short voltage pulses, which is not possible with silicon-based devices.
This is yet another example of computers getting smaller and faster. Honestly as much as I study it sometimes it sill amazes me at the kind of changes we are now seeing.
This is from Phys.org:
(Phys.org) —Faster, smaller, greener computers, capable of processing information up to 1,000 times faster than currently available models, could be made possible by replacing silicon with materials that can switch back and forth between different electrical states.The present size and speed limitations of computer processors and memory could be overcome by replacing silicon with 'phase-change materials' (PCMs), which are capable of reversibly switching between two structural phases with different electrical states – one crystalline and conducting and the other glassy and insulating – in billionths of a second. Modelling and tests of PCM-based devices have shown that logic-processing operations can be performed in non-volatile memory cells using particular combinations of ultra-short voltage pulses, which is not possible with silicon-based devices.
What would the heat dissipation needs of such a computer be? Until we know this, we aren't in a position to simply assume that the heat dissipation barrier present in silicon-based computer chips will be circumvented.
Sorry I'm not following you here, please say if there is something I'm not getting.
I got that you thought that 'today car electronics' (X) + tomorrows electronics (Y = sensors and stuff)
is greater than 'tomorrows car electronics' + Y
I think it is reasonable to think that the 'X' part have evolved a lot which means it is not possible to prove X < X'+Y
First: yeah, they are already here
2020: I'm not 100% sure this will happen myself of course but I'm quite sure it will, and in say one or two years I'm kind of confident. Here we are, predicting the future, lets see what happens!
For 2025 probably yes I say because at that time, there will be only computer assisted and computer driven cars to buy. Most probably human driven cars will be forbidden where ever there is heavy traffic. I mean OK 'ol farmer can drive his truck but where there is millions of people transiting, a human driver will just block circulation.
2030: well if it is touch'n go at 2020, then at 2030 it seems obvious if it isn't already too old tech, we think cars is some kind of obvious thing we need but it really isn't.
I'll try to look that up, it was in one of those books, 80.000$ at some point when DARPA started their autonomous car challenge and 200$ from somewhere but I can't prove that.
My bet though is that at 2020 a LIDAR will cost about 10$ to 1$. The maintenance will be about nothing, it will function for tens of years (say between 10-20) and it will 'tell you' when its functioning is declining.
A car will have several of course.
X = today's electronics in cars
Z = tomorrow's electronics of cars, exclusive of system components needed for autonomous driving (such as a "beefed up" version of what UPS has that you mentioned).
Y = system components needed for autonomous driving
What would the heat dissipation needs of such a computer be? Until we know this, we aren't in a position to simply assume that the heat dissipation barrier present in silicon-based computer chips will be circumvented.
What would the heat dissipation needs of such a computer be? Until we know this, we aren't in a position to simply assume that the heat dissipation barrier present in silicon-based computer chips will be circumvented.
Quote from the link:
These materials could eventually enable processing speeds between 500 and 1,000 times faster than the current average laptop computer, while using less energy.
In labs they can usually make transistors and such run at blazing speed (400+GHz transistors exist since several years for example) but moving it to production is not always 'possible' because of cost-efficiency or it might demand crazy setups like removing a lot of heat or very cold temperatures.
We'll see which one will actually make it to the foundry process and be integrated in a chip
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.