Keeping Autonomous Vehicles Away from People

If we have Autonomous Vehicles navigating around people, can they keep from running into each other? The Autonomous Vehicles can be very good at avoiding people, but people can do very unpredictable things, like tripping, dropping things they are carrying, running into each other. And even if the Autonomous Vehicles aren’t going to run into you, just having them around your feet will be very unnerving, and even cause you to tip. Remember Chewy and the mouse droid in Star Wars.

We have many models for how servants avoid the people they are serving, notably Downton Abbey, with: separate doors, hallways, rooms, floors, back stairs, and my favorites, dumbwaiters (Thomas Jefferson had one of the first in Monticello).

We can follow those models by providing separate, out of the way, “Autonomous Ways”, or “A-Ways” J. In addition to keeping Autonomous Vehicles from under foot, they eliminate noise, and provide a controlled environment for the Autonomous Vehicles: no people children, animals, debris, wind, rain, and other obstructions. In coming posts I will describe many more advantages and innovative features for A-Ways and how they will affect future buildings and communities.

Autonomous Vehicles and Intersections

As we’ve seen, there is no question that Autonomous Vehicles can already navigate and keep from colliding. They can even do it in 4-Dimensions: the usual 3 in space plus Time. Let’s see what that means for our current approaches to traffic control, starting with intersections.

With cars controlled by people, crossing roads are a major source of crashes and injuries. So we have invented a whole range of control mechanisms: stop signs, traffic lights, traffic circles, clover-leafs, and jug-handles (a New Jersey oddity – I love this Wikipedia entry because the language is almost as confusing as the intersections J). These do a reasonable job of reducing crashes (euphemistically called “accidents”, but I believe that is just an attempt to hide the seriousness of this national calamity, 40,000 deaths/year and 1 million injuries; they may not be “on-purposes” but that doesn’t make them accidents, at least one person or vehicle was at fault, so I will avoid the euphemism).

However, these mechanisms also delay and annoy us, and have their own share of crashes. (Have you noticed that lately there seem to be more people running red lights, creeping ahead after stopping, and other egregious behavior?)

Anyway, if you think about what the copter swarms are doing, it’s clear that Autonomous Vehicles don’t need any such controls at intersections! They don’t even need to slow down! They can just time their crossings so they don’t run into each other. Later I’ll talk about my proposals for intersections and traffic management in more detail, but you get the basic idea.

Unless you love roller coasters and bumper cars, as I do, you may not want to be watching out the window while your Autonomous Vehicle whizzes through an intersection at 120 mph, only feet away from a vehicle crossing in front of you. But I predict very few people will be looking out the window anyway, they will be too busy reading, writing, texting, dictating, watching videos, talking, and other more physical activities.

To get from my house to the mall (hey, it’s New Jersey) I go through 3 stop signs and 8 traffic lights, one of them twice because of the jug handle. So it takes 10-15 minutes to go 5 miles, even though the speed limit on most of the distance is 50 mph. And at busy times it takes a lot longer.

And of course the speed limits are another mechanism for managing crashes. So the 25 mph, 40 mph, 45 mph, and 50 mph speed limits I encounter on that trip to the mall can be replaced with whatever the Autonomous Vehicles can safely manage in each situation.

Thus I predict that one of the big attractions for Autonomous Vehicles, and my proposed system, is how fast you get where you want to go. And also how fast you can get things delivered to you.

Can Autonomous Vehicles Keep from Running into Each Other?

Serendipity is awesome! It always makes me believe I am on a right track. The topic for today was set 3 posts ago, and just this morning I found an item in the April 2013 Wired issue about exactly this topic! (I wrote this 2 weeks ago, but we are moving and life got a bit hectic for a while, sorry for the hiatus.) It’s a response to a reader’s question about Autonomous Vehicle safety:

• One key point in favor of Autonomous Vehicles is that “30% of traffic fatalities involve drunk drivers, while 10% are due to drivers distracted by phones, chatter, or food.” Autonomous Vehicles won’t be drunk or “unwrap hamburgers while merging.”

To see a good description of the power of coordinated Autonomous Vehicles, watch this TED2012 talk Swarms of helicopters. To see an example of what they can do, here is an excerpt from that video play music. If you Google “copter swarm” you will see amazing examples of: Coordination, Communication, and Autonomy.

Chris Anderson, editor-in-chief of Wired Magazine, Curator of the TED Conference, and innovator in drones, predicts that these helicopter drones will solve some of the challenges I have been discussing. In the article How I Accidentally Kickstarted the Domestic Drone Boom, Chris points out that you can buy a quad-coptor on Amazon, and “For less than $30, you can buy a little circuit board that can connect it to an autopilot.”

But I don’t want those rotors anywhere near my face, and I really hate to think of 3-dimensional traffic jams, so let’s see what else we can do with these same technologies with the approach I’m describing.

Can Autonomous Vehicles Navigate?

Isn’t navigating without a person complex and expensive? Seems like an obvious question.

It may be complex, but it it’s no longer expensive :-)

When I was a boy, my uncle worked for Lockheed and loved airplanes. He took me to a remote control airplane show and we watched a model B-36  take off and fly. Today this B-36 remote control plane shows 50 years of technology advances, but the concept is the same.

These hobbyists have been working on these models, and today you can buy model planes and even helicopter “drones” that essentially fly themselves. You can buy them as kits or pre-assembled.

What about the technology that makes this possible, and inexpensive?

Think about what an iPhone processor does, with GPS, accelerometers, and a high-resolution camera. Think about the user interface, you see graphics, you touch it for input, and even talk to it. Think about the battery keeping all this going for days.

Now apply all this technology to the challenge of navigation. Chris Anderson in the June 2012 issue Wired Magazine has given some insight into this world of progress: How I Accidentally Kickstarted the Domestic Drone Boom. Here’s his description of how available the navigation technology is:

“Today, all the sensors required to make a functioning autopilot have become radically smaller and radically cheaper. Gyroscopes, which measure rates of rotation; magnetometers, which function as digital compasses; pressure sensors, which measure atmospheric pressure to calculate altitude; accelerometers, to measure the force of gravity—all the capabilities of these technologies are now embedded in tiny chips that you can buy at RadioShack. Indeed, some of the newest sensors combine three-axis accelerometers, gyros, and magnetometers (nine sensors in all), plus a temperature gauge and a processor, into one little package that costs about $17.”

Hopefully this convinces you that it is inexpensive to solve these complex navigation challenges. Next we’ll talk about completely Autonomous Vehicles.