Autonomous Vehicles: Show Me The Efficiency

Autonomous vehicles are coming. I don't want to start out this blog by shoveling yet another semi-skeptical post atop of the steaming pile of articles already written by design professions on AVs. The autonomization of vehicular traffic is a topic on which everyone has something to say, and too much of what is said boils down to either unquestioning exuberance or neo-Luddite nervousness. I do think that a healthy dose of critical thinking is needed when it comes to the conversation around AVs and technology fads in general. I will attempt here to highlight one issue I have with the presentations put forth by the pro-AV crowd, and explain why I think some assumptions about future autonomously flowing traffic are flawed.

I don't buy the predictions of massive efficiency gains in traffic intersections as a result of AVs. The diagrams and animations of free-flowing traffic simply defy logic. The laws of physics cannot be changed, and no matter how skilled at driving they are, AVs cannot magically stop on a dime. Take this example, and pay attention to the woman crossing the street at 0:48.

https://www.youtube.com/watch?v=4SmJP8TdWTU

Now, imagine that the aforementioned woman slips on the sand on the crosswalk, and imagine that the truck that passed in close proximity, traveling at 25 miles per hour, with a new obstacle just four feet away. You can picture what would happen. Autonomously controlled or not, an 8,000 pound truck cannot stop on a dime. The margin of error in this scenario is too thin.

Now, I'm not denying that AVs will save lives by their uncanny ability to anticipate crashes. It has already happened, and will continue to happen. But we have to envision a margin of error that is realistic. In reality, to prevent the possibility of a fatal crash due to unforeseen circumstances, the animated cars in this video would have to be traveling much slower. Slower to the point that those great gains in traffic efficiency won't be so great anymore. Slower to the point that it might look suspiciously like traffic does today, with us dumb humans behind the wheel. Touting the same flawed logic in a TED Talk with almost 2 million views, another fishy animation comes up at 10:28.

https://www.youtube.com/watch?v=OlLFK8oSNEM

Interesting how there are sidewalks, but no people on them. How would one walk across the street with this traffic which is "liquid like our blood"? The answer is that this so called "liquid" traffic would have to come to a standstill, in order to allow people to cross. Again, this is starting to look a lot like the same old intersections. Not so exciting.

There are plenty of other things to get excited about when it comes to autonomous vehicles. The convenience, safety, and comfort of "hands free" driving is enough for us to welcome AVs into our society. We should want AVs to come, and at this point there is no stopping them if we didn't. However, we cannot delude ourselves with empty promises of a world where steel and carbon fiber intelligently whiz by us at 70 miles per hour while we comfortably go about our day. The margin of error is too thin, and no matter how well you can animate a scenario that suggests otherwise, the physical reality remains.