A Week of Robocars


This special chapter in my series of essays on Robocars describes a fictional week in the Robocar world, with many created examples of how people might use Robocars and how their lives might change.

If you haven't been following my essay on Robocars, this may be a good alternate entry to it. In a succinct way, it plays out many of the technologies I think are possible, more about the what than the how and why.

A Week of Robocar Stories

This ends this week-long series of postings on the Robocar essays. Though I have some new sidebars already written which I will introduce later. I realize this set of essays has been more longer than one typically sees in the short-attention-span blogosphere, but I think these ideas are among the more important and world-changing I've covered. I hope I'll see more comments from the readers as you get more deeply into it.


I have just discovered your robocars essays. I have been following the darpa challenges for a while and have thought a bit about the benefits of autonomous vehicles though not nearly as extensively as you have.

I think you make an excellent case for the benefits of this technology. Many similar arguments are used by the skytran people, if you are familiar with this concept: http://www.unimodal.com/. (I am not affiliated with them in any way)

Robocars provide all the benefits but don't require the huge capital cost of special tracks like skytran does.

I'm glad to see such a thorough exploration of this idea, i can't wait to get my robocar.


I was for some time a fan of PRT. But after the results of the Darpa Urban Challenge I realized that computers, with Moore's law behind them, should be able to solve the "track" problem of PRT long before government transit planning commissions would build PRTs.

In short, if you think robocars will come within the next 15 years, you should abandon work on PRT, since it will be obsolete only a few years after it goes into operation, if not before it goes into operation. If you feel robocars are multiple decades away, you might be able to squeeze some life out of PRT.

It's pretty simple. When it comes to software and computers, large central planned projects will never beat bottom-up consumer chosen technology. Never.

I don't have anything against top down or bottom up planning. The key is having the right balance between the two. Leadership works best when it provides a clear focus and allows the bottom to succeed, and the bottom works best when people aren't fighting among themselves and work towards a common goal. I've been reading your BSG stuff and haven't paid much attention to anything else, but the underlying issue of approach seems a common enough point. It's a bit orthoganal but how deal with the world and globalisation are similar issues. Who knows, maybe, the evolution of the car has a significant role beyond itself in how we deal with that.

Things just happen because the people want to do them. People will want to buy robocars because they are safer and pleasant to ride, and eventually because they are cheaper to operate (lower energy) and after that cheaper to buy. Plain old selfish reasons. Vendors will compete, making their models better every year. Much better. Like computers rather than cars because what's core to a robocar is computing.

New electric planes are emerging now with 438 mpg equivalent and 100mph.

UAVs are more advanced and proven than robotic cars.

Combining seems like the perfect way to transform long and mid-range commuting and eventually travel between LA and San Francisco. Even if range was only improved from 100 miles to 200 miles. Making a few stops for battery swaps would be faster and cheaper than driving.

But of course only go between airports unless they are VTOL. Driving such a long haul trip is indeed not very productive. There are a number of ways I can see robocars changing things here, some of which are already on the site:

  • Air travel of any sort can be made much easier if a robocar is allowed to take you directly to and from your plane. This is certainly doable physically, it's the security issues that would stop it. That doesn't apply nearly so much to a private plane, though.
  • This also applies to train or bus travel. Robocar takes you to a small parking lot or station where you board a bus or train, which goes non-stop to another station where you walk out into a waiting robocar.
  • The ability to get quick, reliable, cheap transportation on demand makes not having your car with you in a remote city much more palatable. Why would you want to haul your own car down?
  • Computerized coordination could create ad-hoc transit. If many people announce their travel plans to a central server, it can notice, "Hey, we've got 60 people all going from X to Y at the same time. Let's try to get them to use a bus or train or plane." Such vehicles might offer a cheaper trip, and comfort like a larger seat, ability to stand up etc.
  • This could actually happen during a trip. You start out in a private car but the systems notice a group of people on the road with a common goal, and you pull off to join one another.

On-demand pooling may also allow more efficient scheduling of aircraft, including unscheduled aircraft that "go when a full load can be booked." The aircraft might also be robots.

The planes cannot carry much weight. If one were to load them with explosives then one would get similar effect with larger, cheaper UAV model planes.

Security is not radically reduced.

The light gliders could have assisted short takeoff launching systems, which would be easy because they are so light.

Brad, you've probably already thought of this, but being my pet peeve, I just had to bring it up. I *hate* being behind folks who slow down moving past accidents. Now, I know some slowing is necessary, but most folks do it even if it's across the divider of an interstate! Robocars would simply navigate around the spot, and even perhaps accelerate to keep the same traffic volume. Yet another benefit.

Thanks for the series of essays. I would love to participate in this!

Indeed, the congestion they cause is one of the cost factors in the 230 billion dollar cost of accidents.

However, robocars may not looky-loo, but that won't stop their occupants from commanding them to slow down, short of a draconian regulation system. But they'll pack tighter when they do, and move to the right, leaving the outer lanes for those who did not command theirs to slow.

However, immediately after the accident, the cars themselves will slow down to maintain safety, as there might be debris on the road or hurt people. Remember we design our devices to act more safely than we do. Only after something (emergency crews) send out a message declaring it safe to go through the area at full speed will the cars not want to slow somewhat.

However, the long term goal is just to have fewer accidents, and not worry about this as often.

Add new comment