Thermodynamic illiteracy and the dawn of the driverless car

Imagine you are a movie producer. Someone is pitching a screenplay to you. It’s April 1912. There’s the freezing north Atlantic. There’s stars. There’s people in lifeboats. Bodies in the water. 1,500 dead. 700 survivors. Tragedy. Heartbreak. Heroism. Ships steaming to the rescue. An iceberg.

“And the Titanic?” you prompt the screenwriter.

“The … what?” says the screenwriter.

It felt a bit like that watching the recent documentary, Dawn of the Driverless Car on the BBC. All the ingredients were there except for the biggest one of all: the elephant on whose back the whole concept of ‘car’ stands or falls – energy.

Was the omission deliberate? I’d say not. We are still, crazily, at a point in economics where an awful lot of clever technologists, skilled programme makers and experienced motoring pundits can devote hundreds of man-hours to a documentary and still not notice the street sign telling them they’re driving up a thermodynamic cul-de-sac.

It’s not as though the programme was a simple puff job for autonomous vehicles. The Beeb even deliberately undercut the inevitable tendency to over-hype the technology by having Sara Pascoe do the narration.

It also lined up a Greek chorus of motoring and tech writers to break into the action at intervals, commenting on what issues they thought AVs might solve. These folk all seemed pretty open-minded about AVs, as well as seemingly well-able to curb whatever enthusiasm they might personally feel for driverless travel.

DotDC was one of a set of programmes made by the BBC in a series on artificial intelligence. Maybe that’s why it focused on efforts to get cars to ‘read’ the road as they go along, from live data inputs alone, rather than each vehicle carrying a pre-mapped and scanned model of their surroundings around with them.

The whole thrust of the programme was that, given enough time and money, the developers and the AI will between them achieve a Level 5 Fully Autonomous Vehicle. And although the current state of the AV art is decidedly and somewhat dangerously Heath-Robinson, who is to say that it will prove technically impossible?

No-one had a very firm grasp on why it would be so good to have AVs, though. The Greek chorus liked the idea of being able to Skype on the way to work but were rather ‘meh’ about fully trusting AI to do the job of a human driver.

Some of the boffins thought that, by eliminating idle time (your average private car spends 95% of its life parked and unused), AVs would lead to fewer cars on the road. Others thought AVs might increase traffic. Instead of finding a parking place when popping into a shop, people would set their AV driving around the block until they came back out, for example.

And of course there were the usual claims that AVs will completely eliminate road deaths. Although, as always in this vexed topic, the promise is undercut by the amount of research having to go into whether pedestrians (and roads themselves) will have to be massively reprogrammed to create an environment where AVs can be 100% safe. Cart before the horse and all that.

Oh yes, cart before horse. Here’s where Titanic heaves into view.

The horse and cart of transportation is that automobiles enable the liquid fossil fuel industry. It’s not the other way round. We have hundreds of millions of private automobiles in the world because we have a huge global economy. We have the huge economy because we have colossal amounts of fossil energy. A third of that energy is oil, which needs to be turned somehow into that enormous economy. For that, we use cars, trucks, ships and planes. Mostly cars.

Even more to the point, each AV has a far greater footprint within the massive, fossil-fuelled economy than a conventional car and driver. The latter essentially only needs a printed route map. The AV needs GPS satellites, server farms, cameras, sensors, data networks, cellphone signals: the list goes on.

Every AV had better do the job of many conventional cars, because each one will be so damn energy intensive. But like we said, the more cars AVs gets rid of, the less able to support AVs will be the total economy as a whole.

Naturally, all the prototype AVs in DotDC were electric. Actually that is somewhat good for AVs, economically. EVs are basically fossil-fuelled cars where the fossil fuel is in one place and the car is in another. So electric AVs could work on a thermodynamic economic level, providing there are a lot of them and people can still find a way to use loads of oil (see air travel).

You still have the problem that AVs cannot function without the support of a complex, energy-intensive IT infrastructure – even with in-vehicle AI doing a lot of the legwork. Conventional cars on the other hand… Well, as they say, all you need is a mummy and a daddy and after a few years you get a pilotage system that can operate on only potatoes and water if it has to.

The whole AV thing is posited on continuing economic growth and increasingly complex systems. The energy inputs required are massive and are only possible due to fossil fuels, which are finite and already long past their best days in terms of energy return on investment.

We’re already entering the phase of automotive civilisation where AVs and EVs are starting to look like a cargo cult – earnest offerings made to a god of technology in the hope that she will send the days of happy motoring back to us.

It’s a cult led by high priests of techno-hubris. The guys who are so blind to thermodynamics that they think technology – aka Progress – is the answer to everything.

Unfortunately, the BBC couldn’t stop itself giving the last word in DotDC to one of those cultists, the computer scientist and statistician Sebastian Thrun:


“I can’t foresee the future but I can build it,” he began, enigmatically.

“I believe it is good to be a technology optimist because throughout the entire history of the human race, technology has empowered us. From the very early days, the bronze age, the stone age to the day of the smart phone and modern medicine, it has freed us, it has levelled the playing field for everybody. It has empowered the human race. Why stop that?”

Fine words. But worthless in the sense that his premise about ‘technology’ empowering us is at best only partially correct. Technology applied to available net energy is what’s done the trick. If Mr Thrun seriously believes our current massively technological civilisation will outlive fossil fuels, I’ve got a bridge to sell him. In short, he’s putting his computer before the science – more specifically the physics.

Mass motoring will persist only as long as there’s a global oil industry, i.e. for as long as there’s a worthwhile positive Energy Return on Energy Invested in extracting oil.

After that, when we’re all trying to make the best of far smaller flows of renewable energy and petroleum is for food production, mass transport and military use only, private cars, driverless or not, will again be the luxuries of the mega-rich they were 120 years ago.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s