Aug 14, 2019

Waymo's self-driving vans learn the struggles of parking

A Waymo van learns how to navigate parking lots. Photo: Waymo

Waymo's self-driving vans are learning to share the human driver hatred of shopping mall parking lots.

Why it matters: Parking lots are one of the most difficult environments for a self-driving car to master: unruly vehicles, darting pedestrians and the occasional runaway shopping cart contribute to a Wild West atmosphere.

What's happening: Waymo, a self-driving development company, is teaching its automated test vehicles how to navigate this chaotic environment by practicing on 91 acres at the former Castle Air Force base near Merced, Calif., Popular Science reports.

The state of play ... Engineers can dial up the complexity in a variety of ways:

  • They can control cars reversing out of parking spots, traveling the wrong way down a lane or cutting across empty spaces.
  • They can add people carrying large packages — which affects how the car's perception system sees them — and cause them to dart in front of the car's path.
  • They taught Waymo's vans to be cautious around dumpsters, which can obscure a shopping cart rolling out into traffic.

The bottom line: Just like on roadways, Waymo must map out the parking lot, including details like the orientation of angled parking stalls for cues about the direction of traffic.

  • Parking lots are uniquely challenging, Waymo's Stephanie Villegas tells Popular Science.
  • “They don’t really have standardized rules for how people should and can move about within them. They’re just kind of lawless."

Go deeper: Companies get innovative to fill in urban transportation gaps

Go deeper

Why self-driving car companies are spilling their secrets

Illustration: Lazaro Gamio/Axios

Self-driving technology is hard — so hard that even the industry front-runner is showing its cards to try to get more brainpower on the problem.

Driving the news: Waymo announced Wednesday it's sharing what is believed to be one of the largest troves of self-driving vehicle data ever released in the hope of accelerating the development of automated vehicle technology.

Go deeperArrowAug 21, 2019

Former Google and Uber engineer indicted for trade secret theft

Illustration: Rebecca Zisser/Axios

Former Google and Uber engineer Anthony Levandowski was charged on Tuesday by the U.S. Attorney for the Northern District of California on 33 counts of theft and attempted theft of trade secrets, the New York Times reports.

Why it matters: Levandowski’s actions became well known during the now-settled civil case between Uber and Alphabet, but the criminal charges represent a new wrinkle.

Go deeperArrowAug 27, 2019

Cities are rewriting zoning and land use rules to usher in AVs

Illustration: Rebecca Zisser/Axios

A growing number of U.S. cities, including Las Vegas, Minneapolis and Chandler, Arizona, are re-examining their zoning, land use, and transportation regulations to ease the way forward for AVs.

Why it matters: Cities are exploring changes to decades-old laws in the hopes of attracting new technologies and investment as well as the economic and quality-of-life gains that come with them.

Go deeperArrowAug 21, 2019