Will they build self-driving cars that obey the speed limit? Or will they follow Google's lead and program the cars to "keep up with traffic". That's the line they gave the NTSB a few years ago. But, the Feds came back and said "No you will not do that. Your cars will not be allowed to exceed the speed limit automatically." Apparently, excessive speed was partly to blame for Uber's car killing a pedestrian, after it's computer's memory revealed it was going 40mph in a 30mph zone. Why was a self-driving car going 40 in a 35 zone? What kind of self-driving car is that? The auto manufacturers are introducing unsafe products that haven't been tested in the snow, they haven't shown what happens when the camera lenses get mud or bird droppings on them, how the car's control system deals with worn out breaks or tire punctures, and their engineers and executives don't care about obeying traffic laws.

"energy density" They aren't "saying" that anywhere. They only say "energy", which doesn't necessarily translate into "energy density" in kilowatt-hours per kilogram. How do you know they don't mean energy per unit volume?

"automatically control the distance to the vehicle in front" if the equivalent distance is less than 2 seconds, the driver in front may slam on the brakes to repel the harassing tailgater.

What does "higher energy" mean? Their web site doesn't say anything more about it.

@Sheldon Harrison "There are many people who take 300, 400, 500 and more mile trips" You miss the point. The word "many" is a relative comparison. There are not many people, compared to the entire population of drivers, who regularly drive 300 miles in one day. "I would estimate at least 10%" Where did you get this estimate? It's about as reliable as Trump's estimate of his inauguration crowd size. They don't do it because it is dangerous because of fatigue, including in your case. One day it will catch up to you. The average daily distance travelled is a much better measurement than the maximum single-charge range required to sell a large number of BEVs, especially when millions of cheaper, more comfortable ICE cars are available for those who need long distance.

Self-driving cars will always be a thing of the future. No self-driving car company has shown they will function in snow. They all require a human behind the wheel, who hopefully is concentrating on the road, looking in the mirrors, and nearby cars, to be ready in an instant to take emergency action. So why spend a lot of money for all that automation, just to provide the manufacturer a nice profit? Most of the self-driving cars won't be self-driving most of the time because nobody will want to obey the speed limit, which the NTSA said the self-driving cars must do.

"pack the cars.." In a so-called car train with ten or twenty cars spanning about 50-100 meters in the right lane, how will a normal car, in the center lane be able to get through this gauntlet, onto the exit ramp? I've never heard of Google or any other self-driving car engineers showing how self- driving cars will work in the snow. The Google engineers have publically announced that their self-driving cars will always "keep up with traffic" no matter what the speed limit. Fortunately, the NTSA publicly ridiculed them, saying that Google's cars must obey the speed limit. "•concepts to allow the driver to safely take control of the driving" - How would a driver safely take control of the vehicle, if he hasn't driven in years because his car is always self-driving? Yes, lots of fun here!

More stuff, mechanical gimbals, cameras, microprocessors, communication busses, that people don't need, but will cost money and break down.

Why is 200 cycles unacceptable? Such a battery, with a 500 mile range, would last eight years, if it was cycled to deep discharge once every two weeks. If it was charged every night it would last much longer.

"This means drivers will not have to reach for their iPhone while operating their vehicle, allowing them to keep their eyes on the road and hands on the wheel." Stupid car makers haven't changed since the Ford Pinto was burning people alive. They know that talking on the phone while driving, hands-free or not, is as bad as being drunk behind the wheel. But they can't ignore any opportunity to scam their customers , in their quest to make a buck. They should be working with cell phone makers to integrate the phone and the car, so the driver's phone can't operate when the car is in motion.

"This is a flawed study with predetermined outcomes" ?

What I meant was that people will drive 40 miles each day and recharge at night. So for a 300 mile battery, the discharge is only 13%.

So .4 kwh/kg x 4 miles/kwh = 1.6 miles/kg. 300 miles would need about 187 kg. But this gives 10kw/kg x 187 kg = 1870 kw or 2,500 hp, more than matching the power of engine-driven systems.

@mahonj I think batteries for BEVs will always be sized for range. When they are cheap and small enough, people will want whatever gives them 300 miles range. But they will drive only 40 miles for 360 days out of the year. Just about every article on new batteries describes capacity lifetime in terms of deep discharge, but practical use will be different.

On-site PV. You mean a PV array that can put out 20kwh/day (with full sunlight), which would be about 15x15 meter array? Creating hydrogen like that for a car doesn't seem very practical. Probably the PV array wouldn't be backed-up, it would be supplemented by the grid most of the time because there's not enough space or sun to put the array.

No details about how the hydrogen will be created or transported. These problems have been studied with great intensity by governments and industries. The fact that Sandia needs to study it again means that practical solutions are still a fantasy.

30 kW/kg is great! I keep waiting for Li-Air battery to be perfected, but they seem to be plagued by low current density. So you need a 400 kg battery to get 100 hp, but 1,200 mile range. RPI's 600 wh/kg and 30kw/kg in a 100 kg battery would allow a BEV with about 250 miles range and 4000 hp!

Amazing breakthrough!

"One major cell phone manufacturer has already expressed interest in the system, according to the TUM team. " If the cell phone makers can do this, measuring the position of a cell phone to a few centimeters, then they could also disable the cell phone in a driver's hand.

Or long platoon of 20 semi trucks, six feet behind each other, hogging the right lane. How is that going to work?

"where they can join and leave platoons" How about where manually-driven cars can bust through a 20 car platoon hogging the right lane, when it needs to get off on an exit ramp?

It's not to surprising that a battery of $150 is available. But the 250 Wh/kg capacity is "not there yet". For a BEV, people need much more range. At 4 miles per kwh, you need about 50 kwh battery for a respectable 200 miles. But that's a big battery - 200 kg (440 lb) and still costs a lot, if you want good range. I think Envia's battery is "proven' because GM engineers say it has 400 Wh/kg ( 125 kg or 275 lb/50kwh), but the lifetime is a little less than advertised. I don't know how the weight affects milage, but I still think they are testing it for deep discharge. Anyway, the Envia problem is more about who stole the anode formula. They are using Argonne Lab's cathode, but the silicon composite anode is what the law suit is about and when that's settled, another battery maker will already have a similar anode on the market. What's the advantage of 5,000 cycles over 1,000? Tesla is pushing long-range vehicles, which means a big battery. This means most of the time the battery is being discharged only 20% and recharged every night. The 1,000 cycle battery on 20% DOD would have a practical lifetime of about 20,000 cycles / 365 = 54 years! But the 5,000 cycle battery on 20% DOD would have a real lifetime of about 100,000 cycles / 365 = 274 years! I don't see a practical advantage of 5,000 cycles for a long-range BEV.

...and the copyright lawsuits against Envia don't help them much.

One more item about the Envia battery. Their web site shows 400 cycle lifetime at 80% DOD. But they don't have any info about lifetime at 20%DOD. In addition, it seems obvious that GM is testing at deep discharge in a small battery designed to have high current output. The high current per kg also degrades the lifetime. So if they test a 200 mile battery, as designed for BEV, instead of a short-range battery designed for the Volt, the average current draw would be much smaller. Therefore, the lower current and the small DOD would extend the battery lifetime very much. I wouldn't give up on Emvia's battery yet, especially at $125/kwh, except there are plenty of competitors like CalBattery.

We used to talk about this genset style hybrid about twenty years ago. The key is constant speed duty cycle, which keeps the efficiency high. This is old technology and I'm surprised never to read that somebody has done it.