RIP the internal combustion engine!

slow wing · Post by **slow wing** » Sat Oct 20, 2012 2:55 am

US sales of plug-in vehicles tripled in going from 2011 to 2012 model years...

Link

This is consistent with the rate of growth needed for the EV takeover predicted in the OP...

slow wing · Post by **slow wing** » Mon Oct 22, 2012 9:15 am

This post is slightly off-topic, but it explains why Tesla is doing so much better at EVs than other car companies. That reason is Elon Musk. Here he is explaining, in very simple terms, how he designed and built rockets for another of his companies - SpaceX - that are very much cheaper than what anyone else has done before...

...
Anderson: So what have all your creative people come up with, then? What’s different in your basic technology versus 50 years ago?

Musk: I can’t tell you much. We have essentially no patents in SpaceX. Our primary long-term competition is in China—if we published patents, it would be farcical, because the Chinese would just use them as a recipe book. But I can give you one example.

Anderson: What is it?

Musk: It involves the design of the airframe. If you think about it, a rocket is really just a container for the liquid oxygen and fuel—it’s a combination propellant tank and primary airframe. Traditionally, a rocket airframe is made by taking an aluminum plate perhaps a couple of inches thick and machining deep pockets into it. Then you’ll roll or form what’s left into the shape you want—usually sections of a cylinder, since rockets tend to be primarily cylindrical in shape. That’s how Boeing and Lockheed’s rockets are made, and most other rockets too. But it’s a pretty expensive way to do it, because you’re left with a tiny fraction of the plate’s original mass. You’re starting with a huge slab of material and then milling off what isn’t needed, so you get a huge loss of material. Plus, machining away all that metal takes a lot of time, and it’s very expensive.

Anderson: What’s the alternative?

Musk: It’s similar to the way that most airplanes are made: The stiffness is provided by ribs and hoops that are added on.

Anderson: It’s basically aluminum origami—you’re cutting very precise grooves into it so it folds together into a stiff shape.

Musk: But there’s a catch, because you can’t rivet a rocket like you can an airplane. The pressure differential of an airplane—the difference between the internal and external pressure during flight—is perhaps 7 to 10 psi. But in the case of a rocket, it’s likely to be 80 psi. It’s a lot harder for rivets to withstand that pressure with no leaks.

Anderson: Right.

Musk: So the approach used for aircraft is not exactly feasible for rockets. But there’s another way to do it, which is to use an advanced welding technology called stir welding. Instead of riveting the ribs and hoops, you use a special machine that softens the metal on both sides of the joint without penetrating it or melting it. Unlike traditional welding, which melts and potentially compromises some metals, this process works well with high-strength aluminum alloys. You wind up with a stiffer, lighter structure than was possible before. And your material loss is maybe 10 percent, just for trimming the edges. Instead of a ratio of purchased to flown material—what they call the “buy to fly” ratio—of maybe 10 to 20, you have a ratio of 1.1, 1.2 tops.

Anderson: Wow. Why can you tell us about that?

Musk: The reason I can talk about it is that nobody else knows how to build a rocket this way. [Laughs.]
...

wired.com - Elon Musk’s Mission to Mars

Wow!

Fat Albert · Post by **Fat Albert** » Mon Oct 22, 2012 11:51 am

DOB · Post by **DOB** » Mon Oct 22, 2012 6:56 pm

FA, 100-110 years ago, the same graph would have shown Horse-powered vehicles in the place of cars, and cars in the place of EVs.

TheBouncer · Post by **TheBouncer** » Mon Oct 22, 2012 6:59 pm

Bloomberg did a great doc on Musk a while ago, part of a series on serial entrepreneurs...

Phenomenal talent.

bimboman · Post by **bimboman** » Mon Oct 22, 2012 7:17 pm

DOB wrote:FA, 100-110 years ago, the same graph would have shown Horse-powered vehicles in the place of cars, and cars in the place of EVs.

Except the analogy would make more sense if the Horse drawn vehicle was replaced by one drawn by Donkeys....

The Man Without Fear · Mon Oct 22, 2012 7:22 pm

bimboman wrote:
DOB wrote:FA, 100-110 years ago, the same graph would have shown Horse-powered vehicles in the place of cars, and cars in the place of EVs.

Except the analogy would make more sense if the Horse drawn vehicle was replaced by one drawn by Donkeys....

Although the donkeys were about a thousand times more reliable, luxurious and just as fast as early motor cars, which made Heath Robinson look like a paragon of efficient design.

bimboman · Post by **bimboman** » Mon Oct 22, 2012 7:44 pm

The Man Without Fear wrote:
bimboman wrote:
DOB wrote:FA, 100-110 years ago, the same graph would have shown Horse-powered vehicles in the place of cars, and cars in the place of EVs.

Except the analogy would make more sense if the Horse drawn vehicle was replaced by one drawn by Donkeys....
Although the donkeys were about a thousand times more reliable, luxurious and just as fast as early motor cars, which made Heath Robinson look like a paragon of efficient design.

And the Donkey would be twice the price of a horse as well, and this is why the sales figures would have looked similar, although the car was quite a revolution on the horse wasn't it, rather than an Electric car well just being another crappier car.

slow wing · Post by **slow wing** » Tue Oct 23, 2012 9:34 am

Opportune timing to be having this discussion now on the historical importance of electric cars, as Motor Trend Magazine has just offered this opinion on the Tesla Model S...

Motor Trend Magazine wrote:It may very well be the most important new car since the Model T

Here's the video that quote is from...
2013 Tesla Model S - The Quickest Sedan Built in America - Ignition Episode 38 - YouTube, 7m55s

slow wing · Post by **slow wing** » Wed Oct 24, 2012 12:54 am

TheBouncer wrote:Bloomberg did a great doc on Musk a while ago, part of a series on serial entrepreneurs...

Phenomenal talent.

Cheers for the tip-off, Bouncer - found it!

Elon Musk Profiled: Bloomberg Risk Takers - Video, 52m05s.

It's his life story, chock full of interesting interviews and anecdotes about his personal and professional life. It's exceptionally well done! Rivetting viewing...

6roucho · Post by **6roucho** » Wed Oct 24, 2012 4:38 am

slow wing wrote:Opportune timing to be having this discussion now on the historical importance of electric cars, as Motor Trend Magazine has just offered this opinion on the Tesla Model S...
Motor Trend Magazine wrote:It may very well be the most important new car since the Model T

Here's the video that quote is from...
2013 Tesla Model S - The Quickest Sedan Built in America - Ignition Episode 38 - YouTube, 7m55s

Ignition is the best videozine for petrolheads - and now electroheads.

merlin the happy pig · Wed Oct 24, 2012 9:01 pm

See here for wireless charging.

http://news.bbc.co.uk/2/hi/programmes/c ... 708468.stm

Though the video doesn't state it explicitly, using a little imagination you could install these at traffic lights.
As you approach an electronic handshake could tell the charger your details so that you can be debited while charging.
Efficiency of 90% which is as good as plugging a battery into the mains

Another possibility is for motorways and main trunk roads to include inductive charging while moving,
The induction efficiency is lower for a moving vehicle but a mitigating factor is that it can directly power the motor avoiding the losses in the charge/discharge cycle.

Geek · Post by **Geek** » Wed Oct 24, 2012 10:16 pm

Maybe there's been a dramatic breakthrough in inductive charging in the last year but I doubt it. I saw some presentations from researchers working at the cutting edge of this technology in late 2010 and the efficiencies they were discussing for charging over these kinds of distances were certainly not in the 90% range. I wonder what power per unit area is being transfered at those efficiencies and how small the transmitter is relative to the receiver.

bobbybobson · Post by **bobbybobson** » Wed Oct 24, 2012 10:50 pm

merlin the happy pig wrote:See here for wireless charging.

http://news.bbc.co.uk/2/hi/programmes/c ... 708468.stm

Though the video doesn't state it explicitly, using a little imagination you could install these at traffic lights.
As you approach an electronic handshake could tell the charger your details so that you can be debited while charging.
Efficiency of 90% which is as good as plugging a battery into the mains

Another possibility is for motorways and main trunk roads to include inductive charging while moving,
The induction efficiency is lower for a moving vehicle but a mitigating factor is that it can directly power the motor avoiding the losses in the charge/discharge cycle.

Not quite, a wireless charging system will have all of the losses of a system with a physical connection as well as the magnetic losses associated with transferring power across the air gap between the primary and secondary coils (basically like a transformer without a core). But the magnetic losses are not particularly substantial so the difference isn't all that great. I certainly think that wireless charging is something which will become a commercial reality in residential settings (i.e. instead of plugging your car into the mains, you park it over a charging pad), although roadway powered IPT systems are probably a lot further off for a number of reasons.

bobbybobson · Post by **bobbybobson** » Wed Oct 24, 2012 11:11 pm

Geek wrote:Maybe there's been a dramatic breakthrough in inductive charging in the last year but I doubt it. I saw some presentations from researchers working at the cutting edge of this technology in late 2010 and the efficiencies they were discussing for charging over these kinds of distances were certainly not in the 90% range. I wonder what power per unit area is being transfered at those efficiencies and how small the transmitter is relative to the receiver.

The problem is not efficiency but the fact that the power which can be transferred across an air-gap (especially a relatively wide one as is the case with EVs) is relatively low. The challenge is not to improve efficiency (wireless charging systems are actually very efficient) but to improve coupling between the primary and secondary.

slow wing · Post by **slow wing** » Wed Oct 24, 2012 11:20 pm

The potential for wireless charging is an interesting topic. It certainly isn't needed but it might potentially make EVs even better.

It has actually already been used previously for EVs...

Magne Charge is a largely obsolete inductive charging system, also known as J1773, used to charge battery electric vehicles (BEV) formerly made by General Motors, for vehicles such as the EV1, Chevy S10 EV, and other electric vehicles.[citation needed] It was produced by the General Motors subsidiary Delco Electronics.[1] It is still used by a few hundred first generation Toyota RAV4 EV electric vehicles. ...

http://en.wikipedia.org/wiki/Magne_Charge

Nissan is now looking at using it for its Leaf...

And a Korean company is already looking at induction charging while travelling along a road...

Researchers in Korea reportedly are already working on an electric bus powered by a continuous inductive connection with a cable buried in the road.

The scenario of induction highways that allow EVs to make intercity journeys without stopping to recharge may no longer be the stuff of science fiction.

Link for both Nissan and the Korean company: Nissan Leaf goes wireless - charged by electromagnetic induction from floor pad

slow wing · Post by **slow wing** » Wed Oct 24, 2012 11:40 pm

P.S. that last item looks dodgy to me. A buried cable isn't going to do the trick with any sort of efficiency, unless I am missing something.

bobbybobson · Post by **bobbybobson** » Wed Oct 24, 2012 11:43 pm

slow wing wrote:P.S. that last item looks dodgy to me. A buried cable isn't going to do the trick with any sort of efficiency, unless I am missing something.

Will take a lot more than a cable to transfer any real power..

slow wing · Post by **slow wing** » Wed Oct 24, 2012 11:52 pm

Bobby Bobson, so what would you use for highway charging? How much power could it transfer and at what efficiency?

The trouble with a row of 2-dimensional transmitters is that each one is only going to be aligned for a small fraction of a second. Insurmountable?

bobbybobson · Post by **bobbybobson** » Thu Oct 25, 2012 12:16 am

slow wing wrote:Bobby Bobson, so what would you use for highway charging? How much power could it transfer and at what efficiency?

The trouble with a 2-dimensional transmitter is that it is only going to be aligned for a small fraction of a second. Insurmountable?

I know that inductive power transfer systems have been developed for operation at ~100 kW, however that is with a significantly smaller airgap than for EVs. I believe that the systems currently being developed (for both stationary battery charging and roadway powered EVs) are in the range of 10s of kW. Because most of the losses occur in the power conversion rather than the power transfer I (roughly estimate) that efficiencies of 80 - 90% are realistic. Rather than using a 'cable', the systems being developed at the moment are based on a series of transmitter 'pads' in the roadway, which offer a far more even flux distribution (and one which can be better directed towards a similarly desinged receiver pad) than that of a cable. Moreover, they seem to show reasonable resilience to misalignment which is inevitable in a roadway setting.

Edit: Turns out that the systems I am thinking of (based on pads) are in the range of < 10 kW..

merlin the happy pig · Thu Oct 25, 2012 1:18 am

bobbybobson wrote:
merlin the happy pig wrote:See here for wireless charging.

http://news.bbc.co.uk/2/hi/programmes/c ... 708468.stm

Though the video doesn't state it explicitly, using a little imagination you could install these at traffic lights.
As you approach an electronic handshake could tell the charger your details so that you can be debited while charging.
Efficiency of 90% which is as good as plugging a battery into the mains

Another possibility is for motorways and main trunk roads to include inductive charging while moving,
The induction efficiency is lower for a moving vehicle but a mitigating factor is that it can directly power the motor avoiding the losses in the charge/discharge cycle.
Not quite, a wireless charging system will have all of the losses of a system with a physical connection as well as the magnetic losses associated with transferring power across the air gap between the primary and secondary coils (basically like a transformer without a core). But the magnetic losses are not particularly substantial so the difference isn't all that great. I certainly think that wireless charging is something which will become a commercial reality in residential settings (i.e. instead of plugging your car into the mains, you park it over a charging pad), although roadway powered IPT systems are probably a lot further off for a number of reasons.

My apology, near as good as.

As for mobile recharging, there has been quite a bit of research into powered roads in Korea.
http://www.reuters.com/article/2009/05/ ... 7U20090517

This is the only one I could find at short notice.
The efficiency doesn't approach that of stationary charging at this point (I heard 75%) and I'm not sure what the theoretical limits would be.
However as I stated before, one advantage of drawing power directly from the road is that you don't have to charge/recharge the battery which itself is maybe 90% efficient resulting in .9 x .9 = .81 for charge/discharge.

The other upside of course would be smaller batteries, lighter cars.

Agree with your comment, this is a lot further in the future and may not occur at all.

bobbybobson · Post by **bobbybobson** » Thu Oct 25, 2012 1:27 am

merlin the happy pig wrote:
bobbybobson wrote:
merlin the happy pig wrote:See here for wireless charging.

http://news.bbc.co.uk/2/hi/programmes/c ... 708468.stm

Though the video doesn't state it explicitly, using a little imagination you could install these at traffic lights.
As you approach an electronic handshake could tell the charger your details so that you can be debited while charging.
Efficiency of 90% which is as good as plugging a battery into the mains

Another possibility is for motorways and main trunk roads to include inductive charging while moving,
The induction efficiency is lower for a moving vehicle but a mitigating factor is that it can directly power the motor avoiding the losses in the charge/discharge cycle.
Not quite, a wireless charging system will have all of the losses of a system with a physical connection as well as the magnetic losses associated with transferring power across the air gap between the primary and secondary coils (basically like a transformer without a core). But the magnetic losses are not particularly substantial so the difference isn't all that great. I certainly think that wireless charging is something which will become a commercial reality in residential settings (i.e. instead of plugging your car into the mains, you park it over a charging pad), although roadway powered IPT systems are probably a lot further off for a number of reasons.

My apology, near as good as.

As for mobile recharging, there has been quite a bit of research into powered roads in Korea.
http://www.reuters.com/article/2009/05/ ... 7U20090517

This is the only one I could find at short notice.
The efficiency doesn't approach that of stationary charging at this point (I heard 75%) and I'm not sure what the theoretical limits would be.
However as I stated before, one advantage of drawing power directly from the road is that you don't have to charge/recharge the battery which itself is maybe 90% efficient resulting in .9 x .9 = .81 for charge/discharge.

The other upside of course would be smaller batteries, lighter cars.

Agree with your comment, this is a lot further in the future and may not occur at all.

There is actually a lot of research happening in NZ with regard to wireless power transfer for both EV and industrial applications - http://www.qualcommhalo.com/ - a spin-off company from the Univeristy of Auckland, was recently acquired by Qualcomm. I think they are looking into both stationary charging and roadway charged/powered vehicles..

merlin the happy pig · Thu Oct 25, 2012 1:29 am

bobbybobson wrote:
slow wing wrote:Bobby Bobson, so what would you use for highway charging? How much power could it transfer and at what efficiency?

The trouble with a 2-dimensional transmitter is that it is only going to be aligned for a small fraction of a second. Insurmountable?
I know that inductive power transfer systems have been developed for operation at ~100 kW, however that is with a significantly smaller airgap than for EVs. I believe that the systems currently being developed (for both stationary battery charging and roadway powered EVs) are in the range of 10s of kW. Because most of the losses occur in the power conversion rather than the power transfer I (roughly estimate) that efficiencies of 80 - 90% are realistic. Rather than using a 'cable', the systems being developed at the moment are based on a series of transmitter 'pads' in the roadway, which offer a far more even flux distribution (and one which can be better directed towards a similarly desinged receiver pad) than that of a cable. Moreover, they seem to show reasonable resilience to misalignment which is inevitable in a roadway setting.

Edit: Turns out that the systems I am thinking of (based on pads) are in the range of < 10 kW..

For a pertol vehicle at a steady 100km/hr doing 10L/100km this means assuming 20% efficiency and 35 Mj/L in petrol the such a car uses a steady 20kw.
A smaller vehicle like a leaf would do 6L/100km or so and would need just over 10kw.

Haven't got time to check my maths, may come back to general derision if I got it wrong.

merlin the happy pig · Thu Oct 25, 2012 1:35 am

bobbybobson wrote:
merlin the happy pig wrote:
bobbybobson wrote:
merlin the happy pig wrote:See here for wireless charging.

http://news.bbc.co.uk/2/hi/programmes/c ... 708468.stm

Though the video doesn't state it explicitly, using a little imagination you could install these at traffic lights.
As you approach an electronic handshake could tell the charger your details so that you can be debited while charging.
Efficiency of 90% which is as good as plugging a battery into the mains

Another possibility is for motorways and main trunk roads to include inductive charging while moving,
The induction efficiency is lower for a moving vehicle but a mitigating factor is that it can directly power the motor avoiding the losses in the charge/discharge cycle.
Not quite, a wireless charging system will have all of the losses of a system with a physical connection as well as the magnetic losses associated with transferring power across the air gap between the primary and secondary coils (basically like a transformer without a core). But the magnetic losses are not particularly substantial so the difference isn't all that great. I certainly think that wireless charging is something which will become a commercial reality in residential settings (i.e. instead of plugging your car into the mains, you park it over a charging pad), although roadway powered IPT systems are probably a lot further off for a number of reasons.

My apology, near as good as.

As for mobile recharging, there has been quite a bit of research into powered roads in Korea.
http://www.reuters.com/article/2009/05/ ... 7U20090517

This is the only one I could find at short notice.
The efficiency doesn't approach that of stationary charging at this point (I heard 75%) and I'm not sure what the theoretical limits would be.
However as I stated before, one advantage of drawing power directly from the road is that you don't have to charge/recharge the battery which itself is maybe 90% efficient resulting in .9 x .9 = .81 for charge/discharge.

The other upside of course would be smaller batteries, lighter cars.

Agree with your comment, this is a lot further in the future and may not occur at all.
There is actually a lot of research happening in NZ with regard to wireless power transfer for both EV and industrial applications - http://www.qualcommhalo.com/ - a spin-off company from the Univeristy of Auckland, was recently acquired by Qualcomm. I think they are looking into both stationary charging and roadway charged/powered vehicles..

Thanks for the linky

bobbybobson · Post by **bobbybobson** » Thu Oct 25, 2012 1:37 am

merlin the happy pig wrote:
bobbybobson wrote:
slow wing wrote:Bobby Bobson, so what would you use for highway charging? How much power could it transfer and at what efficiency?

The trouble with a 2-dimensional transmitter is that it is only going to be aligned for a small fraction of a second. Insurmountable?
I know that inductive power transfer systems have been developed for operation at ~100 kW, however that is with a significantly smaller airgap than for EVs. I believe that the systems currently being developed (for both stationary battery charging and roadway powered EVs) are in the range of 10s of kW. Because most of the losses occur in the power conversion rather than the power transfer I (roughly estimate) that efficiencies of 80 - 90% are realistic. Rather than using a 'cable', the systems being developed at the moment are based on a series of transmitter 'pads' in the roadway, which offer a far more even flux distribution (and one which can be better directed towards a similarly desinged receiver pad) than that of a cable. Moreover, they seem to show reasonable resilience to misalignment which is inevitable in a roadway setting.

Edit: Turns out that the systems I am thinking of (based on pads) are in the range of < 10 kW..
For a pertol vehicle at a steady 100km/hr doing 10L/100km this means assuming 20% efficiency and 35 Mj/L in petrol the such a car uses a steady 20kw.
A smaller vehicle like a leaf would do 6L/100km or so and would need just over 10kw.

Haven't got time to check my maths, may come back to general derision if I got it wrong.

Was actually looking at some numbers this morning and the Tesla Roadster consumes about 8 kW at 60 km/h - just checked the Leaf numbers and at 60 km/h consumes about 13 kW.

bobbybobson · Post by **bobbybobson** » Thu Oct 25, 2012 1:38 am

merlin the happy pig wrote:
bobbybobson wrote:
merlin the happy pig wrote:
bobbybobson wrote:
merlin the happy pig wrote:See here for wireless charging.

http://news.bbc.co.uk/2/hi/programmes/c ... 708468.stm

Though the video doesn't state it explicitly, using a little imagination you could install these at traffic lights.
As you approach an electronic handshake could tell the charger your details so that you can be debited while charging.
Efficiency of 90% which is as good as plugging a battery into the mains

Another possibility is for motorways and main trunk roads to include inductive charging while moving,
The induction efficiency is lower for a moving vehicle but a mitigating factor is that it can directly power the motor avoiding the losses in the charge/discharge cycle.
Not quite, a wireless charging system will have all of the losses of a system with a physical connection as well as the magnetic losses associated with transferring power across the air gap between the primary and secondary coils (basically like a transformer without a core). But the magnetic losses are not particularly substantial so the difference isn't all that great. I certainly think that wireless charging is something which will become a commercial reality in residential settings (i.e. instead of plugging your car into the mains, you park it over a charging pad), although roadway powered IPT systems are probably a lot further off for a number of reasons.

My apology, near as good as.

As for mobile recharging, there has been quite a bit of research into powered roads in Korea.
http://www.reuters.com/article/2009/05/ ... 7U20090517

This is the only one I could find at short notice.
The efficiency doesn't approach that of stationary charging at this point (I heard 75%) and I'm not sure what the theoretical limits would be.
However as I stated before, one advantage of drawing power directly from the road is that you don't have to charge/recharge the battery which itself is maybe 90% efficient resulting in .9 x .9 = .81 for charge/discharge.

The other upside of course would be smaller batteries, lighter cars.

Agree with your comment, this is a lot further in the future and may not occur at all.
There is actually a lot of research happening in NZ with regard to wireless power transfer for both EV and industrial applications - http://www.qualcommhalo.com/ - a spin-off company from the Univeristy of Auckland, was recently acquired by Qualcomm. I think they are looking into both stationary charging and roadway charged/powered vehicles..

Thanks for the linky

No problem

Geek · Post by **Geek** » Thu Oct 25, 2012 8:29 am

bobbybobson wrote:
Geek wrote:Maybe there's been a dramatic breakthrough in inductive charging in the last year but I doubt it. I saw some presentations from researchers working at the cutting edge of this technology in late 2010 and the efficiencies they were discussing for charging over these kinds of distances were certainly not in the 90% range. I wonder what power per unit area is being transfered at those efficiencies and how small the transmitter is relative to the receiver.
The problem is not efficiency but the fact that the power which can be transferred across an air-gap (especially a relatively wide one as is the case with EVs) is relatively low. The challenge is not to improve efficiency (wireless charging systems are actually very efficient) but to improve coupling between the primary and secondary.

Sorry are you talking about inductive or radiative charging? For inductive coupling is precisely the problem across an air gap of any significance, particularly as you increase the power per unit area - the efficiency of transfer plummets. Basically more of the magnetic field generated in the transmission coil will miss the receiver coil as the gap is increased. Has the University of Auckland published any papers which propose a solution to this problem?

merlin the happy pig · Thu Oct 25, 2012 8:40 am

Just an oddball thought with regards highway charging.
The contact area of a tyre is both in contact with the road and stationary in relation to it.

Maybe not really a feasible location for on part of an inductive coupling, but just throwing it out there..

Geek · Post by **Geek** » Thu Oct 25, 2012 8:42 am

merlin the happy pig wrote:
Geek wrote: It would be nice to believe you are right but I don't think you are. The ability to automatically drive a car and avoid randomly moving objects is extremely difficult to accomplish. There are various robotics programs aimed precisely in this area and they are currently limited to very slow speeds and very small operating distances. It is certainly not ready for primetime and won't be for many years yet.
The google driverless car use only LIDAR and hs driven 300,000km in normal traffic with no incidents.
(except the crash when the human occupant was in charge)
The environment is far from random, especially as it will become less so once most vehicles are automated. In the interim they can run automated on bus lanes and manual elsewhere unil they prove their reliability.

The task is not anywhere near requiring general intelligence, avoiding pedestrians and other vehicles is not really that hard.
They will also have access to information that humans don't. Other cars can transmit hazard data almost instantly, including pedestrians, potholes, crashes, snow an ice.

But don't take my word for it, read the article.

It also doesn't need to be perfect, only a lot better than human drivers which in truth isn't asking a lot.

The only downside I can see is an end to autodarwination of a few westies.

Just remembered this discussion.

The Google car does not just use LIDAR, it uses a variety of sensors and then compares the outputs. The "autonomous" driving of these cars has not been logged in traffic in inner cities as far as I know but on quiet roads in the middle of nowhere. Even the KPMG report you seem to love doesn't back up your suggestions - look at pages 12 and 13 in particular. As for hard science we are some way off having a system which can respond to randomly moving pedestrians (and more importantly predict them) in the same manner as a human mind. People massively underestimate the complexity of the human mind and the extraordinary job it does when planning automated systems to replace human thinking. Driving on a motorway is certainly far more achievable for these driver assist aids, but not inner city driving. Not for quite some time.

merlin the happy pig · Thu Oct 25, 2012 9:07 am

Geek wrote:
merlin the happy pig wrote:
Geek wrote: It would be nice to believe you are right but I don't think you are. The ability to automatically drive a car and avoid randomly moving objects is extremely difficult to accomplish. There are various robotics programs aimed precisely in this area and they are currently limited to very slow speeds and very small operating distances. It is certainly not ready for primetime and won't be for many years yet.
The google driverless car use only LIDAR and hs driven 300,000km in normal traffic with no incidents.
(except the crash when the human occupant was in charge)
The environment is far from random, especially as it will become less so once most vehicles are automated. In the interim they can run automated on bus lanes and manual elsewhere unil they prove their reliability.

The task is not anywhere near requiring general intelligence, avoiding pedestrians and other vehicles is not really that hard.
They will also have access to information that humans don't. Other cars can transmit hazard data almost instantly, including pedestrians, potholes, crashes, snow an ice.

But don't take my word for it, read the article.

It also doesn't need to be perfect, only a lot better than human drivers which in truth isn't asking a lot.

The only downside I can see is an end to autodarwination of a few westies.
Just remembered this discussion.

The Google car does not just use LIDAR, it uses a variety of sensors and then compares the outputs. The "autonomous" driving of these cars has not been logged in traffic in inner cities as far as I know but on quiet roads in the middle of nowhere. Even the KPMG report you seem to love doesn't back up your suggestions - look at pages 12 and 13 in particular. As for hard science we are some way off having a system which can respond to randomly moving pedestrians (and more importantly predict them) in the same manner as a human mind. People massively underestimate the complexity of the human mind and the extraordinary job it does when planning automated systems to replace human thinking. Driving on a motorway is certainly far more achievable for these driver assist aids, but not inner city driving. Not for quite some time.

I certainly don't uderstimate the Complexity of the human mind.
It is orders of magnitude more sophisticated/flexible than any computer/program synthesis we have devised.

Where I disagree is that this necessarily means the driving task is one that therefore requires such intelligence.
Birds for example are able to fly and maneuver at high speed in near unison as part of a flock while quite near to the ground and around variable terrain. Human intelligence could not achieve this despite our sophistication.
This could be argued is a more difficult problem than driving in city traffic.
Still even simulating a birds level of intelligence is far beyond us at this juncture, but the real question is whether there is another way to achieve this without the requirement of a general purpose intelligence.

As for humans dealing with/predicting paths of randomly moving pedestrians, we can't and pedestrians who walk suddenly onto the road are often hit before a human can even react.
The bar is not that high because driving is not a task we are well suited to.

Thanks for the correction re the LIDAR.
I'll check your other comments re the KPMG report when I have more time.
.

Geek · Post by **Geek** » Thu Oct 25, 2012 1:20 pm

Sorry, my last post was unnecessarily aggressive. My only excuse is I had an upcoming meeting on my mind and was in "attack mode". Sorry!

Regarding the requirements of artificial intelligence, I suspect the requirements are far higher than you believe. It's true that human drivers are far from perfect and yet without the human brain in conjuction with our sensors (eyes and ears) we would crash our cars almost immediately. Even driving down a relatively simple inner city street requires awareness of other vehicles and pedestrians. Flying in the air is relatively simple compared to this situation (in computational terms) since the variables (other random objects) are nearly non-existant. We already have autonomous flying drones for this very reason. The speed may be greater but the operations are far simpler. The more complex task of navigating in the changing landscape of our roads is something which is being worked on by various robotics programs but is not yet at an advanced stage. For my own money I don't see what the point of this would be for consumer vehicles, particularly if we are going to go fully electric. Just build rails and have our vehicles ride (and be powered) along them.

Incidentally the Google car has a driver at the wheel, for autonomous periods the driver simply doesn't take control of the vehicle. As soon as the driver (and their brain) is required the autonomous driving system is overridden. In some ways this is an advanced version of cruise control.

bobbybobson · Post by **bobbybobson** » Thu Oct 25, 2012 11:16 pm

Geek wrote:
bobbybobson wrote:
Geek wrote:Maybe there's been a dramatic breakthrough in inductive charging in the last year but I doubt it. I saw some presentations from researchers working at the cutting edge of this technology in late 2010 and the efficiencies they were discussing for charging over these kinds of distances were certainly not in the 90% range. I wonder what power per unit area is being transfered at those efficiencies and how small the transmitter is relative to the receiver.
The problem is not efficiency but the fact that the power which can be transferred across an air-gap (especially a relatively wide one as is the case with EVs) is relatively low. The challenge is not to improve efficiency (wireless charging systems are actually very efficient) but to improve coupling between the primary and secondary.
Sorry are you talking about inductive or radiative charging? For inductive coupling is precisely the problem across an air gap of any significance, particularly as you increase the power per unit area - the efficiency of transfer plummets. Basically more of the magnetic field generated in the transmission coil will miss the receiver coil as the gap is increased. Has the University of Auckland published any papers which propose a solution to this problem?

Inductive. I think you are equating the degree of coupling with efficiency, which is not quite correct. Of course, the power which can be transferred in a loosely coupled (i.e. wireless, coreless) system is constrained, but it is not necessarily inefficient. Imagine your 'receiver' being completely decoupled form your 'transmittter' (both of which in an inductive power transfer system are essentially inductors designed for high frequency operation). In this scenario all you are doing is delivering an AC current to an inductor (the transmitter) - no losses here aside form those as a result of copper losses in the windings and eddy current losses in the core/shielding material. The amount of power which is transmitted, so to speak, is dependent on the coupling - it isn't as if the transmitter radiates a fixed amount of power irrespective of the position of the receiver.

A number universities are beginning to take wireless (particularly inductive) power transfer seriously for a variety of applications - not just EVs - if you have access to a good electrical engineering database you should have no problem finding plenty of publications..

slow wing · Post by **slow wing** » Wed Oct 31, 2012 10:05 am

Video: Jay Leno Drives The Tesla Model S

Interesting video because Leno is a real petrol head but also very practical and knowledgeable about EVs. He asks great questions.

slow wing · Post by **slow wing** » Wed Oct 31, 2012 10:20 am

Tesla Produces 1000th Completed Model S Body

Recall that Model S production started slowly after the 22 June 2012 launch in order to ensure top build quality. After originally hoping for 5000 cars produced in this calendar year, Tesla is now looking good for around 3000. They are looking to build 20,000 Model S's in 2013. Currently they have around 13,000 deposits in their order books.

slow wing · Post by **slow wing** » Fri Nov 02, 2012 5:44 am

2013 Automobile of the Year: Tesla Model S
From our January, 2013 issue / By David Zenlea / Photos by A. J. Mueller

...this year's field was the strongest in recent memory. We weren't expecting much from the Tesla other than some interesting dinner conversation as we considered "real" candidates like the Subaru BRZ and the Porsche Boxster. In fact, the Tesla blew them, and us, away. ...

Link: Automobile Magazine

Video

slow wing · Post by **slow wing** » Tue Nov 06, 2012 10:10 am

Tesla has just released its 3rd quarter earnings statement. Their stock surged 9% as they are in great shape!

Tesla reaches ‘turning point,’ promises mass production for 2013

November 5, 2012 6:56 PM
Dylan Tweney
...

Link

Their production rate for the Model S has surged. They are now making 200 a week, which is sufficient for financial break-even. Elon Musk predicts they will be up to 400/week in a month's time - that's already the pace they need for their predicted 20,000 Model S sedans in 2013. With the Model S now Car of the Year, surely they will exceed that now!

ONWARDS THE GREEN REVOLUTION!!!

slow wing · Post by **slow wing** » Sat Nov 10, 2012 10:06 am

Ford is now getting serious about plug-in hybrids. It has just released the 2013 Ford C-Max Energi, which has an all-electric range intermediate between the plug-in version of the Toyota Prius and the Chevy Volt:

11 miles, 4.4 kWh battery - Prius plug-in
21 miles, 7.6 kWh battery - Ford C-Max Energi
38 miles, 16.5 kWh battery - Chevrolet Volt.

Notice that the Ford gets slightly better miles/kWh than the other two. Ford claim this is due to using Panasonic lithium-ion batteries, which Tesla also uses and which Ford claim to be the best in the world.

The Ford plug-in outsold the Prius plug-in last month in the US, in its first full month on sale. The base model retails at $32,950, which drops to just below $30,000 after the federal rebate for plug-ins.

This should prime people's interest for the mass-market Tesla all-electric vehicle that will be similarly priced but do away with the ICE. Expect that to be released in about 2 to 2.5 years and have a range of about 150 miles\250 km. Expect it to be called the Tesla Model T. Here. First.

Donger · Post by **Donger** » Sat Nov 10, 2012 3:50 pm

there is some sort of plug in (hybrd I guess) here in DC. it has this weird large attachment he hangs oout of the front grill when he is re-charging.

he parks on a very busy main road in a very rough area. it looks a little incongruous with the lead running across the footpath and into the junction box.

I fully expect the lead to be stolen....or for a few guys to set up a few camp chairs out there with a stolen TV and watch the Redskins.

DOB · Post by **DOB** » Sat Nov 10, 2012 4:17 pm

It's not really related to the technology itself, but Tesla may have to sort out some legal issues with their sales business model.

http://www.npr.org/2012/11/09/164736569 ... chise-laws

slow wing · Post by **slow wing** » Tue Nov 13, 2012 3:40 am

Cheers, DOB. This is has been in a lot of articles. Tesla has created a better sales model as part of their top to bottom review of how a car should be made and sold. They sell their cars like computers are sold - online. Instead of a network of sleazy dealers, they have welcoming booths in the shopping malls where people can ask questions and get familiar with their cars.

They are cutting out the middle men and, naturally, the car dealers don't like it. Tesla should prevail on this one. The laws are different in the different states of the US and I predict that Tesla stores in each state will back off just enough to satisfy the laws in that state. In some states, the prospective customer won't actually be able to order the car from an in-store computer, for example.