Brad Templeton Home

Main Page

Brad Ideas
(My Blog)

The Case



Car Design






End of Transit







Robocar Blog

How Computer Geeks can Save the Planet

How Computer Geeks can Save the Planet

Why are Robocars (autonomous computer driven vehicles for ordinary roads) the best shot for computer people to make a difference in the world -- both saving lives and making a really meaningful dent in oil consumption and greenhouse gas emissions?

Computer nerds -- and I'm one of them -- often work for the chance to make a difference. And our inventions, from PCs to games to the internet itself, have made a big difference. Is it possible to make one that's even bigger?

Robocars are essentially a computer project. There's lots of cool automotive engineering to be done, but the real wizardry will come from hardware and even moreso from software.

The hardware needed includes cheap, powerful sensors, faster processors and specialized hardware for machine vision and pattern recognition. The software needs to drive all that, and combine machine vision, awareness and navigation with many other factors to produce the solution.

It's an Artificial Intelligence (AI) problem, but it's what is sometimes called a "weak AI" problem. It doesn't need a self-aware human equivalent being. It needs the more basic tools we're already getting good at. It's the sort of problem skilled computer wizards were born to solve.

(Note that this is a quick summary of a much more detailed argument and roadmap you may want to look into for a fuller understanding.)

Step One -- Save a million lives a year

That's how many are killed around the world each year by human drivers in car accidents. 45,000/year in the USA. Most caused by inattention, which computers don't have.

Millions more are injured or maimed or suffer from the loss of loved ones. Car accidents are the leading cause of death (among standard major categories) for ages 5-45. This is on par with curing a major disease like Azheimer's or stomach cancer. Getting here is not trivial, and there are many roadblocks, but it's worth doing not because it's easy, but because it's hard.

Step Two -- Save trillions of dollars

Car accidents also cost $230 billion/year in the USA, trillions worldwide, due to damage, suffering, congestion, wasted fuel and more.

Step Three -- Save hundreds of billions of hours

Americans spend 50 billion hours/year driving cars. It's not totally wasted, as they can listen to music and information, but even at minimum wage this is trillions of dollars globally.

Step Four -- Who saved the Electric Car?

I explain in detail in the main article how the ability of robocars to deliver themselves, park themselves and refuel/recharge themselves solves the big barrier that sits in front of electric cars -- having enough battery capacity for long range at a moderate cost, and being able to recharge as fast as you can fill a gas tank.

Robocars enable on-demand robotaxis that deliver the right car for your trip in seconds. After you're done it worries about the recharging. Nobody cares about the range of their taxi. You just want it to get you there. When nobody cares about range, or whether refueling is convenient, suddenly all sorts of alternative fuels (not just electricity) are enabled.

Switching urban trips (about half our driving) to electric cars or anything that's not gasoline cuts oil use by cars in half.

Step Five -- Welcome the ultralight car

With almost no accidents, and electric power, we enable the ultralight car for single person trips. That's 10 to 20 times more efficient than the typical car or transit system today. Now we're not just getting off of oil, we're reducing the true energy need, which cuts back the coal too.

These ultralights also should take far less energy to manufacture, without all that expensive steel and that complex gasoline engine and drive train.

Step Six -- Save the planet

By switching a large fraction of driving to ultralight electrics by making them better than today's cars in every way, you've reduced the environmental footprint of cars by more than half. Combine that with all the other efforts underway in that area, and it's the leading potential for a serious dent in emissions. And you saved lives and much more while doing it.

I don't see a lot of other computer engineering projects that could have this effect on the world. Full AI would, but it's much further down the pipeline; it would also solve the robocar problem after all. Nanotech would have a bigger effect too, though like Full human level AI, it would have both good and bad consequences.

This project seems to offer the greatest good combined with the highest near-term tractability.

If this set of steps intrigued you, now start exploring the full depth of the essays on how this all happens.

Why are robocars so important to going green?

There are only a few "big" areas where serious redictions in energy usage and pollution can be made. Those big areas are personal transportation, electrical generation, agriculture and possibly building heating/cooling. While industrial energy usage is quite large, it consists of a large number of different processes for which there is no one magic bullet.

Cars are the big one. But people won't get more efficient vehicles simply because they "should" or even because the law demands it. People buy what they want. (Indeed, even while auto engineers made cars 30% more efficient since the 80s, the fleet average MPG stayed flat as people bought bigger cars with more horsepower.) In order to change cars, you must make the alternative both cheaper and better in just about every way. A bit cheaper is not enough. A bit better is not enough, except over a long time. A technology has to be marketable to win, not just workable.

Robocars offer the chance to be both much cheaper and much better. And by enabling experimental fuels, they offer a number of ways to that path.

There are a few other options -- revolutionary battery technologies, for example, or some amazing new renewable liquid fuel so good that people will switch quickly from gasoline. We should continue to work on these. But robocars will bring them about even faster by changing how people buy and hire cars.