Adventures in Google Self-Driving Cars: Pizza Delivery, Scavenger Hunts, and Avoiding Deer
After 500,000 miles of road tests, Google’s self-driving car team gave New Yorker writer Burkhard Bilger unusually deep access for a profile this week.
But self-driving cars are what put Google on the map as a company that tries to make science fiction into reality, so the tale of how they came to be is compelling.
The hero of Bilger’s story is 33-year-old engineer Anthony Levandowski, who joined Google after building a self-driving motorcycle for the DARPA Grand Challenge (though it sounds like that didn’t work all that well). Levandowski was toiling away on Google Street View with the more-famous inventor/professor Sebastian Thrun before the two of them got the go-ahead from Google co-founders Larry Page and Sergey Brin to start working on self-driving cars.
What made the difference? A TV show producer’s wacky idea for a stunt, believe it or not.
From Bilger’s piece:
“In February of 2008, Levandowski got a call from a producer of ‘Prototype This!,’ a series on the Discovery Channel. Would he be interested in building a self-driving pizza delivery car? Within five weeks, he and a team of fellow Berkeley graduates and other engineers had retrofitted a Prius for the purpose. They patched together a guidance system and persuaded the California Highway Patrol to let the car cross the Bay Bridge—from San Francisco to Treasure Island. It would be the first time an unmanned car had driven legally on American streets.”
The successful bridge crossing earned Page and Brin’s go-ahead within a few months, according to Thrun.
Then the two Google co-founders, “like boys plotting a scavenger hunt,” gave the self-driving car team a set of 10 100-mile itineraries. “The roads wound through every part of the Bay Area — from the leafy lanes of Menlo Park to the switchbacks of Lombard Street. If the driver took the wheel or tapped the brakes even once, the trip was disqualified.” The team completed all 10 in a year and a half.
There’s a lot more to the story, but the real question is, what’s next for the self-driving car? There are challenges on multiple fronts, now that the scavenger-hunt phase is over. There’s jumping over legal hurdles and figuring out how to bring the cars to market, given carmakers are allergic to the word “self driving” (though they’re okay with smaller and subtler tweaks, where machine smarts help drivers out), plus making the next technological leaps forward in sensors and machine learning. Even if autonomous cars are statistically safer, any mistake will tarnish the entire endeavor.
As Bilger reports, “The car has trouble in the rain, for instance, when its lasers bounce off shiny surfaces. (The first drops call forth a small icon of a cloud onscreen and a voice warning that auto-drive will soon disengage.) It can’t tell wet concrete from dry or fresh asphalt from firm. It can’t hear a traffic cop’s whistle or follow hand signals.”
At the same time, the car is often smarter than a human. For instance, it can prepare to brake preemptively based on traffic data about a slowdown coming ahead, or slow to a crawl at nighttime on a wooded road when it senses a deer walking on the shoulder.
And improvements are coming, said Bilger: “At the tech meeting I attended, Levandowski showed the team a video of Google’s newest laser, slated to be installed within the year. It had more than twice the range of previous models — eleven hundred feet instead of two hundred and sixty — and thirty times the resolution. At three hundred feet, it could spot a metal plate less than two inches thick. The laser would be about the size of a coffee mug, he told me, and cost around ten thousand dollars — seventy thousand less than the current model.”
The rest of the longish read is here.