The Coming Zombie Robot Driving Apocalypse of You

antrobus

cells wrote: »

That clearly is wrong

A human car doing 10k miles a year would take 100-200 years to hit 1 million miles. A car just like a building or even a person, degrades with time as well as use...

A robot car doing 10k miles a year would similarly take 100-200 years to hit 1 million miles.

cells wrote: »

...So doing a million miles in 3 years is more material friendly and possible than doing a million miles in 100 years...

Of course it is. But that would be true irrespective of who is driving.

cells wrote: »

..also usage in humans clearly makes a difference so the driver is not trivial. Plus these robo cars will be serviced correctly and on time...

Why? I can see no reason why an owner of a fleet of vehicles would be more likely to have them serviced correctly and on time because they were robotised.

N1AK

cells wrote: »

Why pay £100 for a train ticket from London to Birmingham when a robo taxi will take you door to door for £25 and allow upto 5 passengers

There is literally no chance of the journey costing £25. It might get sub-£70 but even that is optimistic in my opinion. It'd still be a massive game changer if it happened though.

That is a market of billions of passengers just in the UK

Billions of journeys perhaps...

I actually agree with most of the points, but the hyperbolic numbers aren't helpful

N1AK

antrobus wrote: »

A robot car doing 10k miles a year would similarly take 100-200 years to hit 1 million miles.

I assume most of his calculations here are based on the assumption that low use drivers, or all drivers, would rent autonomous cars as required. It's a reasonable assumption for that scenario.

I think a fairly strong case can also be made for autonomous vehicles averaging higher lifetime mileages on average (though I think 1 million very optimistic).

I'm a massive believer in autonomous vehicles. I think they are going to make a huge change to society and do it soon. I don't think there's any need to overstate the benefits to explain why.

cells

N1AK wrote: »

You clearly don't.

It's got nothing to do with processing power, or the affordability of it, thus Moore's is not the reason. The computer power required is quite limited, and comically cheap compared to the sensor technology in things like Google's test cars. When the first self-driving car is being sold the processing power will, almost certainly, be lower than the typical desktop PC of the time.

I've written image analysis software (a decade ago now) and even then the issue was always knowing what to look for, not having the processing power to do it.

look at how amazing google is

https://www.youtube.com/watch?v=3BNOsxMZD14

3D mapping on a mobile phone!


in ten years time this stuff will be much higher resolution much higher frame rate and much more accurate. with only two extremely cheap digital cameras as eyes

cells

N1AK wrote: »

There is literally no chance of the journey costing £25. It might get sub-£70 but even that is optimistic in my opinion. It'd still be a massive game changer if it happened though.

we can try to guess how much a self drive taxi business might cost

lets say £15k to buy the vehicle and it is in use for 2 years and does 400,000 miles and is then scrapped. 70mpg engine.

Capital cost is 3.75p a mile
Fuel cost is 7.15p a mile
Insurance maybe 0.1p a mile
Maintenance maybe 1p a mile

Total maybe = 12p a mile


London to Birmingham is 125 miles so = £15 cost


Now that is the cost to the service provider so there needs to be some overheads on top of that but the margins will be extremely thin as the barriers to entrance will be so small that the industry will be ultra competitive.

Even apple only doubles up so lets say £30 as an upper figure and that is door to door not station to station

cells

antrobus wrote: »

A robot car doing 10k miles a year would similarly take 100-200 years to hit 1 million miles.

Sure but my guess is that the UK will go from the current ~35m human cars to perhaps just 10m robo cars.

So the average yearly mileage per car should go up by at least 3.5x. If they also take business from trains buses planes coaches etc that might means 5x

antrobus wrote: »

Why? I can see no reason why an owner of a fleet of vehicles would be more likely to have them serviced correctly and on time because they were robotised.

the current fleet owners, tom !!!!!! and harry, know so little about cars and just see maintenance as a partly avoidable expensive which gets them nothing in return and as such many tom dicks and harrys don't maintain their vehicles properly and even do stupid things like fake mots

N1AK

cells wrote: »

we can make up unrealistic numbers and try to guess how much a self drive taxi business might cost

You can but it doesn't mean you should.

I'm not sure which assumption is the least realistic. A fully autonomous vehicle fit for use as a Taxi at £15k or insurance to carry passengers without a driver present for 0.1p a mile. Not that the other's aren't pretty far up the wouldn't survive contact with reality scale.

I especially like how you assumed current petrol prices are a new norm AND that their would be no mileage whatsoever where customers aren't present. Those were especially good signs that you'd really thought this through :rotfl:

Even apple only doubles up so lets say £30 as an upper figure and that is door to door not station to station

Apple doesn't only double. It's average is closer to 25% and it has released various products with margins in excess of 70%. Do you have some compulsive lying disorder that forces you to spout blatantly false numbers?

cells

N1AK wrote: »

You can but it doesn't mean you should.

I'm not sure which assumption is the least realistic. A fully autonomous vehicle fit for use as a Taxi at £15k or insurance to carry passengers without a driver present for 0.1p a mile. Not that the other's aren't pretty far up the wouldn't survive contact with reality scale.

I especially like how you assumed current petrol prices are a new norm AND that their would be no mileage whatsoever where customers aren't present. Those were especially good signs that you'd really thought this through :rotfl:



Apple doesn't only double. It's average is closer to 25% and it has released various products with margins in excess of 70%. Do you have some compulsive lying disorder that forces you to spout blatantly false numbers?

Why are you trying so hard to be an !!!!

neither of us know what the cost will be or even if they will be at all within our lifetimes hence why I post with such things ad 'we can try to guess' or maybe this or that

If you dont like my guesses then fek off and don't reply to them

globalds

N1AK wrote: »

I'm not sure which assumption is the least realistic. A fully autonomous vehicle fit for use as a Taxi at £15k or insurance to carry passengers without a driver present for 0.1p a mile. Not that the other's aren't pretty far up the wouldn't survive contact with reality scale.

So we are talking about a vehicle that will be sold all over the world ..That could as easily be on the streets of Beijing as Birmingham ...I think these ..The simplest ones could be very very cheap.

Insurance ..well seeing as premiums are based on statistical risk ..I would expect ..regardless of the Luddite's doubts ..That the premiums are going to be very low too.

These vehicles are going to have to be better ,Far better than the things we presently have ,other wise they won't replace the present set up.
failure rates and fail safes far in excess of anything we demand of human systems .The likes of Google and apple and now I guess every motor manufacture on the planet knows this and So it will be a slow but sure approach ...That and the necessity of stopping dumb humans doing stuff like filling one full of explosives and asking it to driving next to something they don't like.

antrobus

cells wrote: »

Sure but my guess is that the UK will go from the current ~35m human cars to perhaps just 10m robo cars.....

There is a problem with your 'guesses'. They have little connection with commercial reality.

Like your claim that "robo-taxis might get 1,000,000 miles but over a 3 year period". Do the math. A vehicle would need to cover 38 miles each and every hour, of each and every day, in order to get that kind of mileage over three years. Your guess is wrong by about a factor of ten.

Like your guess about the size of the world taxi market was wrong by about a factor of ten.:)