Ed Self-Driving Cars: Pros, Cons, and Predictions

[acbytesla]

There's plenty of blame to go around. You and I might want to pick and choose, but our choices are limited to what engineers have made available to us. In many case they were not just following orders working for hire, but developed innovations without being told to and even self-promoted them. Management will often go with it simply because the novelty may attract customers looking for something different, or because they respect the engineer's vision. OTOH management may have ideas of their own that don't come to fruition because the engineers don't like it. You may say that's the fault of management for not hiring the 'right' engineers, but when the industry is full of Luddite engineers what choice do they have?

Real engineers aren't just machines that just take in a specification and spit out a design, they are people just like you and me.Their own desires often factor bigly in what they come up with, and then we are stuck with those choices. Of course our decisions have impact too, but they are constrained by what is made available to us.

As in politics, it often comes down to the lesser of two evils. In the 1890's New York city had a big problem - horse manure. What ultimately solved it was the motor car, a far cleaner mode of personal transport. Electric cars were initially very popular in the city because they were quiet, low maintenance, easy to use (particularly for women who didn't want to risk breaking an arm or worse trying to start them) and didn't need an exotic fuel.

But then the electric starter motor was invented by Charles F. Kettering an American engineer who founded Delco and was head of R&D at General Motors from 1920 to 1947. Kettering didn't have to do that.
He could have worked on making electric vehicles better instead. But he was a petrol-head. Kettering's lack of concern about the environment led (pun intended) to him identifying tetraethyllead as an anti-knocking compound, precipitating one of the largest environmental disasters iin the history of mankind. How many of us used leaded petrol in our cars, blissfully unaware of the damage it was causing? An engineer was responsible for that.

OK, in all fairness BS.

Kettering made an electric, motor which is what an electric starter is. Electric cars actually predate Kettering's tenure at GM by 30 years. The problem in developing a practical electric vehicle wasn't solved by making better electric motors, but because better batteries were built. And that wasn't until Stanley Whittingham developed the first lithium intercalation battery at Exxon in 1972, and John Goodenough later discovered that lithium cobalt oxide could serve as a stable cathode material. And it stll took 19 years before a practical rechargeable lithium battery was first put into production by Sony.

Kettering was an engineer. Not really a research scientist which is necessary for engineers to build on.

 

[bruto]

Why are you blaming the engineers for this? They perform work for hire. Who is hiring them? Who is giving them the specifications?

We're not in the situation we're in because of engineers. We're in this situation because you and I want to pick and choose our conveniences, and lay off the moral weight of our choices on those we hire to fulfill our dreams. If engineers are perfecting a car that shouldn't be there at all, then what should we make of everyone who possesses a car, and vituperates the engineers for improving on it at our behest?

Ok, so perhaps it's misdirected to say it's engineers who are at fault, but engineers are hired by people who see engineering as the answer to needs, and I think in a more general sense society is seeking an engineering solution to one problem that is going to contribute to a greater one in the long run. Obviously, we still need cars, and I'm certainly guilty of using one, since no alternative exists here. Maybe it's too late already, but I just think that improving cars, while an admirable goal seen through a narrow lens, ignores the issue of whether we should be putting more resources into mass transit.

 

[acbytesla]

Ok, so perhaps it's misdirected to say it's engineers who are at fault, but engineers are hired by people who see engineering as the answer to needs, and I think in a more general sense society is seeking an engineering solution to one problem that is going to contribute to a greater one in the long run. Obviously, we still need cars, and I'm certainly guilty of using one, since no alternative exists here. Maybe it's too late already, but I just think that improving cars, while an admirable goal seen through a narrow lens, ignores the issue of whether we should be putting more resources into mass transit.

I don't disagree with any of that. In fact, I am 100 percent convinced we have to science and engineer ourselves out of the mess we have created.

I am a huge energy science geek. Every other day I read about some energy breakthrough in science laboratories. And almost none of them make the leap from the lab to practical products that might contribute to society. So much of developing practical products is the work of engineers

What's wrong with improving cars? It's not like people tomorrow are going to start taking the bus.

It's not a zero sum game. Let's improve transit, let's make gasoline cars more efficient, let's make better ebikes and e-trikes. Let's make 500wh or 1000wh per kilogram batteries. I like the idea of self driving cars because it solves a problem. I see electric Robotaxis taking me from my home to the mass transit line. So I don't need to have a car.

Science and engineer the crap out of this.

 

[Rolfe]

Let's not make ICE cars more efficient. Let's send them the way of the steam engine.

 

[acbytesla]

Let's not make ICE cars more efficient. Let's send them the way of the steam engine.

I couldn't disagree more. Eventually maybe. It's going to take decades to make that transition. Improving fuel efficiency in ICE engines means two things. One, is less money for the same amount of work. And two, less CO2 emissions. Better for the environment.

 

[The Great Zaganza]

If ICE spends its entire budget on EVs and pass them on to more deserving agencies, I'm all for it.

 

[arthwollipot]

Let's not make ICE cars more efficient. Let's send them the way of the steam engine.

*thinks in Australian*

There will always be a place for ICE vehicles in locations where EV charging infrastructure can't go. We need to make them as efficient as possible for those times where they are actually needed.

 

[Ivor the Engineer]

I think Adams is being too hard on engineers here. The examples of engineering I've ever looked at suggests that engineers by and large prefer to have exactly the features to meet the spec; no more, no less.

It's actually the customers - the normal people - who are fond of ignorantly assuming that there's room for more features.

When you find that They are breaking your product that just works, it's not because the engineers got greedy. It's because some other, more important customer got stupid, and/or marketing and management got greedy.

One problem is that other than for trivial systems, nobody can express what they want a system to do (and not do) in sufficient detail to reliably engineer it. Eager to please (and keep their jobs), engineers fill in the blanks after realising that no (sensible) answers to their requests for clarification or more detail will be coming. Once a system becomes too complex for a single human brain to comprehend at a useful level of detail then it's implementation is going to be a buggy bloated mess.

Another problem is that modern engineering is largely built on pillars of sand. For example, we have countless programming languages and the vast majority of them are, for want of a better word, crap. We even celebrate the "features" of our tools with things such as the International Obfuscated C Code Contest. The primary concern of most engineers appears to be avoiding using languages that are too verbose. I.e., languages where you have to be precise and explicit when doing things such as converting a floating point value into an integer, or an integer into a vector of bits.

In general software and electronic engineering have completely ignored Niklaus Wirth's Plea. We now live in a world where we use multiple complex and buggy software stacks implemented in non-strictly typed languages running slowly on ridiculously powerful processors just to (unreliably) turn lightbulbs on and off. It is the polar opposite of elegance and sophistication.

 

[Ivor the Engineer]

*thinks in Australian*

There will always be a place for ICE vehicles in locations where EV charging infrastructure can't go. We need to make them as efficient as possible for those times where they are actually needed.

ICE vehicles also have limited range and need to be refuelled via complex infrastructure. HVDC links can transmit electrical energy efficiently over 3000km. The largest coast to coast distance in Australia is 4000km, so it would only need a 2000km cable from the centre of the country to be able to reach any point on the coastline.

Another thought is that perhaps there are places that should be mostly out of reach of the vast majority of humans, such that the use of ICE vehicles by the small minority of people with a valid and useful reason to go to those places is insignificant.

 

[Ivor the Engineer]

I couldn't disagree more. Eventually maybe. It's going to take decades to make that transition. Improving fuel efficiency in ICE engines means two things. One, is less money for the same amount of work. And two, less CO2 emissions. Better for the environment.

I don't think so. See Jevons Paradox. Examples: Now we have multiple and bigger TVs since the CRT went the way of the Dodo. Software has gotten bloated and slower as processors and memory have become faster and larger. Also see this. The only way to reduce CO2 emissions from ICE is to reduce the number of ICE on the planet. Making the ICE more efficient will increase the number and use of them by more than the individual reduction in each ICE's CO2 emissions.

 

[acbytesla]

I don't think so. See Jevon's Paradox. Examples: Now we have multiple and bigger TVs since the CRT went the way of the Dodo. Software has gotten bloated and slower as processors and memory have become faster and larger. Also see this. The only way to reduce CO2 emissions from ICE is to reduce the number of ICE on the planet. Making the ICE more efficient will increase the number and use of them by more than the individual reduction in each ICE's CO2 emissions.

Piss poor analogy. There are presently constraints that make EVs not a fit for every application. There are no benefits for cathode ray tubes. Incrementally more efficient ICE engines will not prevent the transition to EVs. It will however reduce fuel consumption and emissions of the few ICE engines that will continue for those applications.

 

[theprestige]

One problem is that other than for trivial systems, nobody can express what they want a system to do (and not do) in sufficient detail to reliably engineer it. Eager to please (and keep their jobs), engineers fill in the blanks after realising that no (sensible) answers to their requests for clarification or more detail will be coming. Once a system becomes too complex for a single human brain to comprehend at a useful level of detail then it's implementation is going to be a buggy bloated mess.

Another problem is that modern engineering is largely built on pillars of sand. For example, we have countless programming languages and the vast majority of them are, for want of a better word, crap. We even celebrate the "features" of our tools with things such as the International Obfuscated C Code Contest. The primary concern of most engineers appears to be avoiding using languages that are too verbose. I.e., languages where you have to be precise and explicit when doing things such as converting a floating point value into an integer, or an integer into a vector of bits.

In general software and electronic engineering have completely ignored Niklaus Wirth's Plea. We now live in a world where we use multiple complex and buggy software stacks implemented in non-strictly typed languages running slowly on ridiculously powerful processors just to (unreliably) turn lightbulbs on and off. It is the polar opposite of elegance and sophistication.

My bad. I thought you were talking about engineering, not software development. Carry on.

 

[Ivor the Engineer]

Piss poor analogy. There are presently constraints that make EVs not a fit for every application. There are no benefits for cathode ray tubes. Incrementally more efficient ICE engines will not prevent the transition to EVs. It will however reduce fuel consumption and emissions of the few ICE engines that will continue for those applications.

Why bother? Which company is going to invest significant sums in eking out a few more percent efficiency from the ICE to sell fewer and fewer of them over time?

 

[Ivor the Engineer]

My bad. I thought you were talking about engineering, not software development. Carry on.

How quaint! A typical engineering decision is do we implement an algorithm entirely in hardware, in software with hardware acceleration, or entirely in software. FPGAs blur the boundary even further.

 

[acbytesla]

Why bother? Which company is going to invest significant sums in eking out a few more percent efficiency from the ICE to sell fewer and fewer of them over time?

That's an investment quandary. Maybe because there are still hundreds of millions of ICE vehicles and devices that are still. operating and serving a function. An̈d Honda, GM, Toyota, Ford and Cummins to name a few that continue to invest billions in engine development.

 

[theprestige]

How quaint! A typical engineering decision is do we implement an algorithm entirely in hardware, in software with hardware acceleration, or entirely in software. FPGAs blur the boundary even further.

Don't misunderstand me; I think you're wrong about software developers, too.

 

[theprestige]

Dup

 

[arthwollipot]

ICE vehicles also have limited range and need to be refuelled via complex infrastructure. HVDC links can transmit electrical energy efficiently over 3000km. The largest coast to coast distance in Australia is 4000km, so it would only need a 2000km cable from the centre of the country to be able to reach any point on the coastline.

And through what would that cable pass? Who's putting it there? Where are they sleeping and what are they drinking?

 

[Dabop]

*thinks in Australian*

There will always be a place for ICE vehicles in locations where EV charging infrastructure can't go. We need to make them as efficient as possible for those times where they are actually needed.

That's precious little ICE needed at all then- hell I live in a tiny rural town (1500 people) in rural Qld (so think ultra rightwing for Australia lol) and we got an EV charge point AND seven EV's in town.... (mind you, even one of the locals that relies purely on solar for their offgrid power STILL have an EV and charges at home... and soon to be two offgrid homes with an EV charging off solar- and that will be eight EVs in town, and still no Teslas lol)

I mean its literally an hours drive to even GET to the next town- so all the antiEV crowd would have you believe that EVs would never work here, let alone take off....

And yet here we are- not a single new car thats been bought in my town in the last two years has been an ICEr- literally every new car bought was an EV (and not one Tesla in the entire lot lol)

If they can work here, then they certainly can work for 99.99999% of the Australian population lol

 

[arthwollipot]

I mean its literally an hours drive to even GET to the next town- so all the antiEV crowd would have you believe that EVs would never work here, let alone take off....

Yeah, I'm talking about places where it's 14 hours drive to the next town. And yes, there are people out there.