For slightly more than a year, 400 volunteers have tested out Waymo’s self-driving car service in Phoenix, Arizona, for free, letting the driverless vehicles whisk them to work, shopping centers, the bar, or anywhere else within a 100-square-mile area.
In that time, these riders have been filing the non-technical equivalent of bug reports, using the cars’ rider support call buttons and in-app feedback forms to point out issues with the service and highlight use cases that Waymo researchers might have missed.
Through their experiences, Waymo has learned a few things:
- It needs to get better at designating specific pick-up entrances at a store so that frustrated riders won’t have to lug shopping bags through the hot sun to reach a car
- On narrow streets, riders prefer to cross the road to reach a car, instead of having it drive to the end of a road, turn around and come back
- It needed to figure out how to accommodate people with service animals (it figured this out after a query from a passenger)
- The best way to wake sleeping passengers is with a little chime sound.
Not all the rides have been seamless, which is kind of point of this experimental phase.
“We’re taking each element of their feedback and weaving it into the car’s design and its behavior,” Saswat Panigrahi, a Waymo product manager, tells CNBC.
Waymo, which started as Google’s self-driving car project and is now an independent part of parent company Alphabet, is expected to launch a paid self-driving car service to the public in Phoenix before the end of the year, although it hasn’t announced an official date or price. This pilot is part of the tech giant’s ambitions to use its AI chops to get an early lead in self-driving vehicles, a space expected to generate $800 billion per year in revenue by 2030.
While only 400 people have been accepted into Waymo’s early rider program so far, the company has received more than 20,000 online applications. Its passengers have ranged from age 9 to 69 and Panigrahi says they were deliberately chosen to represent a wide spectrum of use cases. There are parents who want to cut out rides to school, single people running errands and riders who don’t have drivers licenses for medical reasons.
Waymo is closer than any other company to making fully autonomous vehicles a real product, even as the industry is facing increased scrutiny.
Americans are more afraid of riding in self-driving cars this year than last year, according to recent AAA surveys, and Uber temporarily halted all of its self-driving tests after one of its cars hit and killed a pedestrian in Arizona in March.
After that accident, Waymo’s CEO said that its own technology would have prevented the crash and reiterated that safety is the company’s top priority.
Panigrahi emphasizes that its early rider tests in Arizona have proven how much time driverless cars can save for people who usually commute.
“Once the excitement of being in the car wears off, people realize how taxing driving actually is,” he says. Parents have reported feeling like they could focus more fully on their childrens’ stories while riding together and one woman discovered a neighborhood park that she’d driven past for years but never noticed.
But for ride-hailing services like Waymo’s, that time savings could come at the cost of a human driver’s job. In a Phoenix-specific Reddit thread about the service’s upcoming launch, many expressed hesitation about using it over Lyft or Uber, unless they saw a significant cost difference.
“I’m not going to take someone’s livelihood away and not save money,” one user wrote.
Besides Arizona, Waymo has also filed a permit for driverless testing in California.