Autonomous vehicles have built-in fe disadvantage

[Frank Lee]

Driverless Cars Are Giving Engineers a Fuel Economy Headache

[Stubby79]

You'd think they'd be able to program them to be more fuel efficient....if a human can pulse and glide and do other such hypermiling techniques, you'd think a computer could master it and more than make up for extra power usage.

[Frank Lee]

Computer shouldn't suffer from epileptic gas pedal ankle seizures, pressing on gas and brake at the same time, and endless idling.

[ksa8907]

There are plenty of opportunities to save power. The article doesn't broach the subject of no longer needing to run an entire computer system to monitor the engine and transmission.

I don't think it will be as big of a problem as they let on.

[freebeard]

They make it sound like it's not about driving finesse.

That’s because self-driving technology is a huge power drain. Some of today’s prototypes for fully autonomous systems consume two to four kilowatts of electricity -- the equivalent of having 50 to 100 laptops continuously running in the trunk, according to BorgWarner Inc. The supplier of vehicle propulsion systems expects the first autonomous cars -- likely robotaxis that are constantly on the road -- will be too energy-hungry to run on battery power alone.

I call shenanigans.

https://www.nvidia.com/en-us/self-dr...cars/drive-px/

NVIDIA DRIVE™ PX is the AI car computer that enables automakers, truck makers, tier 1 suppliers, and startups to accelerate production of automated and autonomous vehicles. It scales from a single processor configuration delivering AutoCruise capabilities, to a combination of multiple processors and discrete GPUs designed to drive fully autonomous robotaxis.

The architecture is available in a variety of configurations. These range from one passively cooled mobile processor operating at 10 watts, to a multi-chip configuration with four high performance AI processors — delivering 320 trillion deep learning operations per second (TOPS) — that enable Level 5 autonomous driving.
....
DRIVE PX Xavier will deliver 30 TOPS of performance, while consuming only 30 watts of power. Packed with 7 billion transistors, a single Xavier AI processor will be able to replace today’s DRIVE PX configured with dual mobile SoCs and dual discrete GPUs — at a fraction of the power consumption. Available Q1 of 2018.

I would expect LIDAR, for instance, also following a development curve. I know the cost has been cut by an order of magnitude recently.

edit:
That's in the short term. Down the road () we can apparently blow right past Moore's law with neuromorphic (brainlike) silver wire mesh single-atomic-transistor brainz:

https://www.wired.com/story/brain-built-on-switches/

Gimzewski believes that the silver wire network or devices like it might be better than traditional computers at making predictions about complex processes. Traditional computers model the world with equations that often only approximate complex phenomena. Neuromorphic atomic switch networks align their own innate structural complexity with that of the phenomenon they are modeling. They are also inherently fast—the state of the network can fluctuate at upward of tens of thousands of changes per second. “We are using a complex system to understand complex phenomena,” Gimzewski said.

Do you remember Isaac Asimov's Positronic Brain? This is that, in a vegetative state. What happens when it wakes up?

[redpoint5]

Silly article. Maybe current prototype systems consume 2-4 kW since the focus is on proof of concept and not efficiency. The production versions will be much more power efficient. Just think of the processing power in a phone, which consumes just a couple watts.

Automakers don't seem to be that concerned with electrical efficiency. Just look at how common wasteful resistors are to control blower motor speed. Those can burn 100 watts alone.

[Frank Lee]

Originally Posted by fb

Quote:

Do you remember Isaac Asimov's Positronic Brain? This is that, in a vegetative state. What happens when it wakes up?

Then the smelly apes can be bypassed altogether so the computers can do their thing unimpeded.

[gone-ot]

Intelligence ain't electrically cheap...takes power, just like a light bulb.

One way to reduce the "Watts-per-thoughts" is to revert to burst-mode computation (think of it as "pulse-n-glide" for computer processing) where all the devices do something (think, measure, store, correct, etc.) and then go into brief hibernation (sleep mode) to reduce power consumption. Then, after a preset time, they all come back awake...and repeat the process. Sorta like the difference between a power-wasting LINEAR ANALOG power supply versus a DIGITAL SWITCHING power supply.

[redpoint5]

Human processing power accounts for approximately 1/5 of total energy use. That's about 20 watts of thinking power. Of course, some humans are more processing "efficient" than others.

[freebeard]

Then the smelly apes...

I prefer 'warm furry mammal'. But yeah, that's the game plan, at least for driving cars.