And it confirms all of our worst suspicions about the incident.

Sunday night’s crash in Tempe, Arizona involving an Uber in “autonomous mode” has caused a whirlwind of speculation. How in the world was this crash not avoided? Did the pedestrian suddenly jump out from between some parked cars? Was there some other obstruction to visibility or other factor that made this crash truly unavoidable?

Now we can (sort of) see for ourselves.

Tonight, a crappy-resolution video of the crash was finally made available to the public. At a mere 22 seconds, only the most relevant footage is shown. And it tells us exactly who dropped the ball here.

Advertisement

Everybody. Everybody dropped the fucking ball.

There’s no mistake about it; pedestrian Elaine Herzberg was in clear violation of right-of-way, jaywalking across multiple lanes without using a nearby crosswalk. She doesn’t appear to change her pace or even bother to look in the direction of the oncoming vehicle until the very last split-second. Nor does she bear any lamps or reflective clothing to make herself more visible. It’s almost like she wanted to get hit.

And the Uber doesn’t hesitate to fulfill such a death wish. Right up until the footage cuts away to spare us the graphic part, the car (in autonomous mode, mind you) maintains speed and trajectory, never attempting to slow or swerve. The car doesn’t react to her or her bicycle, which is fully perpendicular to the car, a sizeable profile for any cameras or sensors to pick up. Yet every system in the car ignores her presence.

Advertisement

And then there’s the final fail-safe: Rafaela Vasquez, a “safety operator”, there to make sure that the vehicle performs as it ought. She has pedals at her feet, a steering wheel right in front of her, and bright headlights to show the way. She even has the luxury of street lamps lining the road in question. And she can see far more than what any of us can make out in the video... if she would just LOOK.

WE are first able to see Elaine’s shoes at the video clip’s 2-second mark. Vasquez would have had a much better view, though.

The poor video resolution doesn’t allow us to see Herzberg until 2 seconds into the clip. By that point, the impending crash is borderline avoidable. Hard braking might not even be enough, but a quick swerve might have saved her.

Advertisement

#ThingsVasquezWouldRatherLookAt

But Vasquez wasn’t blinded by the limitations of 480p. She was blinded by her own choice to trust the car and give her attention to what seems to be a phone instead. It’s estimated that had she been watching through the windshield, Vasquez should have been able to see Herzberg and her bicycle at LEAST 2-4 seconds before we are able to spot her in the low-res video. That’s more than enough time to keep this from happening.

Alas, by the time she looks up to see Elaine through the windshield, it’s already too late. She’s already squandered every opportunity she had to prevent this.

Advertisement

She just let it happen. As did the car. As did Elaine.

The pedestrian could easily have waited, stopped, and/or jumped out of the way. The car should have been able to see her much sooner than the video suggests, and hit the brakes and/or swerved in time. And the safety operator could have been looking at the road. Is that too much to ask?