What Happens After the First Pedestrian Death from an Automated Vehicle?

Certain things are inevitable. We're going to have automated vehicles within my lifetime. We are going to have a pedestrian killed by being hit by a self driving car within my lifetime.

I believe some inevitable changes occur generationally. While the track record of the small fleet of automated vehicles currently on the road shows them to be orders of magnitude safer than human drivers, I find myself right around the age line of people that tend to be pro-automated vehicles and people that express trust issues.

A common criticism I've heard from naysayers is that they don't trust the machine and want a human behind the wheel because they're somehow intrinsically more capable of making decisions. My own perspective is data driven. The data shows automated vehicles to be safer than human drivers.

One possible criticism of that result could be that the driverless cars opperate in cherry picked white room conditions. That is a criticism worth taking seriously and studying. I'm open to evidence to the contrary, but I consider this claim to be unsupported. If, as I expect, the safety record of driverless cars continues as wider rollouts occur, then I even believe there is a strong case for making human driving illegal.

If affordable, wide-scale automation is available and demonstrably safer than human drivers, then for me, restricting the ability for humans to drive cars is the right course of action. I say restricting, because I am open to some special licensing system (presumably much more discriminating than our current licensing system) to enable some people to opperate cars. This ought to work similar to licenses for other potentially deadly activities like handgun ownership or pilot licenses.

The other interesting concern naysayers express is liability. However rare, there will be accidents, injuries, and deaths which occur when two or more driverless vechicles collide with one another. Indeed, the legal liability questions here are interesting, and something better left to lawyers.

I'd like to speculate in a slightly more unexpected direction. I believe driverless cars will change the way we consume transportation in ways that most people aren't yet thinking about.

It's no great prediction to state that private vehicle ownership will likely become unpopular in favor of on-demand fleets owned by companies like Uber and Lyft. There's no need to own a machine that sits idle the majority of the time. Especially when a fleet owner could be staffed for peak time and leveraging their fleet for alternative purposes like package delivery when demand allows.

A less discussed possibility is the fact that when a critical mass of vehicles are driverless, the opportunity for large multi-agent collaboration becomes possible. Imagine vehicles that are communicating upstream conditions to their peers a few miles behind. Consider cars that calculate the minimum following distance for a safe stop down to the millimeter, and cars that tighten that gap by having a direct line of communication and information sharing with the vehicles in front of them to better determine the likelihood of sudden braking.

The opportunity for vehicles to travel more efficiently and at faster velocities will surely be one outcome. But if many people object to driverless cars to begin with, how many more would object to noticable increases in velocity. As a frequent pedestrian, I have to admit I'm a bit irrationally afraid of this.

But if people are granted new free time during their commutes, looking out the window becomes significantly less important. Of course, we still might enjoy that or even require that to avoid motion sickness. But if that issue can be solved, some of us would be happy to engross ourselves in work behind a computer screen during our travel times. That being the case, perhaps these vehicles might run more efficiently underground.

When dangerous machinery is created, we tend to place it in special non-accessible areas. This is of course in the case of nuclear power plants but is also the case for more banal equipment. Try to get yourself near an industrial lumber planar. It would take a certain amount of trespassing and, I expect, result in a security guard asking you to leave.

If society is truly afraid of driverless cars, perhaps they need to be long in a specialized, even enclosed, area of their own. I expect that accident between driverless cars and pedestrials will frequently be due to negligence on the part of the pedestrian. Maybe we need to protect people from themselves.

Whether my idea about an enclosed system pans out or not, I am quite confident the fleet model will take over for most people. Although overall safety should improve with automation, there are going to be unfortunate cases we as a society will have to face. My hope is that we do so in the interest of expanding protection and safety, rather than eliminating innovation.