Video showing moments leading up to self-driving Uber crash released

Image 1 of 2

Photo Courtesy: Tempe Police Department

A video that shows the moments leading up to a fatal crash involving a self-driving Uber has been released.

The crash, which happened Sunday, took the life of 49-year-old Elaine Herzberg. The person who was inside the car at the time was identified as Rafaela Vasquez, 44. The car, according to officials, was in autonomous mode at the time.

The video, released by Det. Liliana Duran with Tempe Police Wednesday afternoon, combines video taken by an exterior camera with video taken by an interior camera. The video does not show the collision itself, as the video taken by the exterior camera stopped moments before the impact.

(Can't see this Facebook video? Click here)

Video taken from an interior camera shows the driver, as well as her reaction upon seeing the pedestrian. In the video, Herzberg was seen suddenly emerging, walking a bike across Mill Avenue. She was not in a crosswalk, and had nearly made it across the street when she was struck.

Inside, Vasquez appears to be looking down at something, but it is unclear what she was looking at. She looked up for a few moments, and then took her eyes off the road again for a full five seconds before she realizes she was about to hit Herzberg.

Speed at the time was 40 miles per hour, and the Uber car, according to Tempe Police, was headed north on Mill Avenue.

According to a statement released by Det. Duran, Tempe Police Vehicular Crimes Unit is actively investigating the accident, and when the investigation concludes, they anticipate submitting the case to the Maricopa County Attorney's office for review.

Meanwhile, Uber released a statement on the incident, saying, in part:

"The video is disturbing and heartbreaking to watch, and our thoughts continue to be with Elaine's loved ones. Our cars remain grounded, and we're assisting local, state and federal authorities in any way we can."

Expert reviews video

"There will never be perfect driving situations. We just want to reduce and mitigate the risks associated with it," said Ashraf Gaffar, an assistant professor at the Ira A. Fulton Schools of Engineering at Arizona State University.

Gaffar said the camera doesn't show how a human would see the shot, but it does indicate some telling signs, including the woman in the roadway and the reaction from the driver.

""Even human drivers would probably not do better," said Gaffar. "We don't know, but in similar situations, human drivers would still have the accident."

Gaffar says the lasers and sensors in the car may need to be altered to catch movement in the distance. Timing and reacting, however, can only be so fast.