Video of fatal Uber accident in Arizona, authorities suggest company not at fault

Police in Tempe, Arizona have looked at video of the fatal accident involving an autonomous Uber vehicle and a pedestrian and concluded the car probably isn’t at fault. From the San Francisco Chronicle:

“The driver said it was like a flash, the person walked out in front of them,” said Sylvia Moir, police chief in Tempe, Ariz., the location for the first pedestrian fatality involving a self-driving car. “His first alert to the collision was the sound of the collision.”…

From viewing the videos, “it’s very clear it would have been difficult to avoid this collision in any kind of mode (autonomous or human-driven) based on how she came from the shadows right into the roadway,” Moir said….

“I suspect preliminarily it appears that the Uber would likely not be at fault in this accident, either,” Moir said…

“I won’t rule out the potential to file charges against the (backup driver) in the Uber vehicle,” Moir said.

I’m going to tell you what I see but first, watch the clip yourself. This is two videos, one showing the forward view from the car and the other showing the driver. Both clips stop an instant before the collision.

My first reaction is that I’m not sure I could have avoided this accident under the same conditions. The road where the victim, Elaine Herzberg, was walking is so dark that I didn’t see her until her shoes appeared in the headlights. From that point, it was just over a second before the car reached her. The car was speeding slightly but even at 35mph, there wasn’t much time to do anything except maybe swerve left and hope for the best.

In the second clip, the one facing inside the car, the driver appears to be looking at a phone, maybe texting or reading something. She’s clearly not paying full attention to the road. Her eyes are down for about five seconds just before the crash. Again, I’m not sure she could have prevented it anyway, but if she’d been watching more closely maybe she could have done something. It seems at least possible that human eyes would have been more sensitive than the camera that recorded this. Maybe she’d have seen the woman’s shape in the road a second before she became visible in the clip. But she wasn’t really looking.

When fully autonomous vehicles become available, there will probably be all sorts of requirements that the human driver pay strict attention to the road at all times. But in reality, people text and drive all the time now. Once cars start driving themselves, I suspect most people are going to be doing exactly what this driver was, i.e. paying some attention to the road, but only occasionally. Of course, this is a pilot program, which means this driver should have been aware this is not (yet) a perfect system. She’s a test pilot of sorts which means paying attention is pretty important.

There’s something else to consider here. Maybe the fault isn’t just with the car or the driver or even the pedestrian. Maybe part of the problem is poor road design. Popular Mechanics highlighted a series of tweets by Eric P. Dennis, a Transportation Systems Analyst at the Center for Automotive Research.

Dennis told Popular Mechanics, “Human beings have certain things that they do that are somewhat predictable—cutting corners when you’re walking is a very predictable human thing to do—and there are many elements of right-of-way design that could either exacerbate those risks or mitigate them.”

It sounds trivial but I think he has a point. These look like bike lanes but no one is supposed to ride or walk on them. But if you were looking to cross the road you can see how this might seem tempting, though obviously not as safe as heading to the crosswalk. And as you zoom out, you can see that once you are in that space between the lanes of the road, there’s really no path to get to the crosswalk:

I think it’s possible the human driver could have tried some extreme (and probably dangerous) maneuver to avoid the pedestrian if she’d been paying more attention to the road. That said, it also seems as if poor design may have played a role here.