Riding in Waymo One, the Google offshoot’s very first self-drivin …


It is late afternoon in Chandler, a suburb of Phoenix, Arizona, and I’m getting jittery waiting for my self-driving car to make a left turn before arriving at the coffee shop. Seconds tick by, and the vehicle — an autonomous Chrysler Pacifica minivan owned by Waymo — is letting too many opportunities pass by without turning. But before I can get truly annoyed, the Waymo vehicle makes the left, and my mid-afternoon caffeine fix is soon satiated.

Waymo, the self-driving subsidiary of Alphabet, launched its first commercial autonomous ride-hailing service here in the Phoenix suburbs on Wednesday — a momentous moment for the former Google self-driving project that has been working on the technology for almost a decade. I was one of the lucky few to test out the company’s robot taxi experience a week before the launch. And I say “lucky” because to ride in one of Waymo’s autonomous minivans, not only do you have to live in one of four suburbs around Phoenix, but you also have to be in a very exclusive, 400-person club called the Early Riders.

It’s not a motorcycle club for morning people, but rather Waymo’s year-and-a-half-old focus group for its self-driving cars. To start out, Waymo’s new self-driving taxi service — dubbed “Waymo One” — will only be available to “several hundred” members of the Early Rider program, said Dan Chu, head of product at Waymo. That’s exclusivity within exclusivity.

Since April 2017, the Early Riders have been testing the company’s autonomous vehicles for trips to work, school, and various errands. Some of these people will now migrate to Waymo One, while others will stay in the Early Rider program. Those who move to Waymo One will continue to use the company’s self-driving Chrysler Pacifica minivans in much the same way they did while in the Early Rider program — except now they will be charged for the rides.


Photo by Vjeran Pavic / The Verge

Waymo declined to comment on how quickly it plans to open up its service to the rest of us. “We want to understand each step,” Chu said. “How are people responding? How are people feeling? Then, as people are comfortable with that, it means we can broaden out.”

The cars aren’t fully driverless yet: they will include “trained drivers” behind the steering wheel until Waymo decides to pull them out. Chu says it will test a variety of “configurations;” the company says it will eventually offer driverless rides, but it declined to give an exact date.

The presence of these trained drivers can be both practical and psychological. While Waymo insists its cars will drive themselves a majority of the time, the human driver can take control should the vehicle become confused. Plus, initial riders may be more comfortable getting into a car with a human in the driver seat.

In November 2017, the company shocked the world when it first demonstrated its fully driverless vehicles in Chandler. The company sent out a video clip of happy riders taking selfies in the backseat, while a minivan with no human behind the steering wheel drove around slowly. It sent a powerful message: driverless cars are real, and they’re coming.

But those trips occurred in just a small, largely residential portion of its service area, and they were never a majority of the rides provided by Waymo. According to a recent report in The Information, Waymo’s most advanced vehicles are still occasionally confounded by certain traffic situations, which suggests the tech — while incredibly advanced — is still not quite ready for the real world. In recent months, the company began putting trained drivers back in the vehicles, and it looks like the drivers will stay for the time being.

Taking human monitors out of the vehicle, though, is still part of Waymo’s ultimate “vision,” Chu said. But those trips will continue to be tested on Early Riders before being offered to Waymo One customers. Chu framed it as a customer service issue rather than a technical problem. “That way we really kind of soak up the user feedback from the Early Riders before we bring it to Waymo One,” he said.

The Ride

Over the course of three separate trips in Chandler, the trained drivers in my Waymo vehicles never take control. I’ve ridden in a Waymo vehicle without a human being in the driver’s seat once before, but it was not on public roads. I was fully prepared to experience a fully driverless ride while in Chandler, but, alas, Waymo rejected my request.

The rides are uneventful, but it is exciting to experience the little flourishes that have been added for ride-hailing customers. The minivans still smell new, or at least recently cleaned. The screen on the back of the driver’s headrest features a large blue “start” button that I could press to initiate the ride. (There’s also a physical button in the headliner of the vehicle that performs the same task.) After pressing the button, a musical chime sounds and a robotic-sounding woman’s voice says, “Here we go.”

As I said, I’m an experienced Waymo rider — three trips and counting — but this one feels more mature. Before, it felt like you were being driven by your half-blind grandmother, but now, riding feels… mostly normal. The car slows down for speed bumps, accelerates for lane changes, and handles a number of difficult maneuvers like unprotected left turns. And it even surprises me a couple of times, like when it ended up braking too far into the crosswalk at an intersection, and then reversed back a few inches to make room for pedestrians. Of course, it probably shouldn’t have stopped so abruptly in the first place, but it is still comforting to see the car correct its mistakes in real time.

Waymo reportedly has had some trouble in the past with unprotected left-hand turns, and the extra seconds it took to make the turn from North Alma School Road into the parking lot confirms it. The company maintains these turns are advanced and inherently dangerous maneuvers, which is why UPS drivers aren’t recommended to do them. But they are a normal part of driving, and it is something Waymo will need to figure out as it expands its horizons.

As expected, Waymo One will only be available in four Phoenix suburbs where the company has been testing its vehicles for the last two years: Chandler, Mesa, Tempe, and Gilbert. The service area roughly equals 100 square miles, Chu said. Territory expansion will also occur first in the Early Rider program before shifting over to Waymo One. The company’s vehicles have traveled 10 million miles autonomously on public roads, with an additional 7 billion miles in virtual simulation.


Photo by Vjeran Pavic/ The Verge

The App

How will these Phoenicians summon a Waymo vehicle? Through an app, of course. Waymo’s app looks like a minimalist version of Uber’s app. Verify your pickup and drop-off locations, and the app gives you an estimate for your car’s arrival. The app will be available in iOS and Android, but it won’t appear in Apple or Google’s respective app stores until the service becomes more widely available.

Chu said the company is trying to manage demand so riders would never be forced to wait more than a few minutes. Occasionally a blue-and-green “W” may appear on the map to indicate a more accessible pickup or drop-off location. That means riders may be prompted to walk a little bit so Waymo’s vehicles have an easier time locating them.

Much like Uber or Lyft, Waymo One riders can rate the quality of their trip on a scale of one to five stars. They can also elaborate on what made the ride great by selecting from a list of canned responses like “route choice,” “driving,” and “car condition.” A support function allows riders to get an immediate phone call from a Waymo representative or engage in an in-app chat with them. This is meant to better prepare riders for when the cars arrive without a trained driver in them.


When I use it, the app is a little buggy, which is typical of most beta software. After I hail my first Waymo, the car that is supposed to pick me up vanishes from the app entirely before eventually arriving in real life. Clem Wright, Waymo’s product manager for the app, said he hadn’t seen that happen before, but he insisted it was something that was easily fixed. (It turned out a previous user had toggled a setting in the app that caused the vehicle to drop off the screen.) It was a reminder that Waymo is likely to encounter similar hiccups going forward as it pivots from “moonshot” project to a full-grown business.

Let’s talk money

Okay, but how much will it cost? After all, Waymo One is just a ride-hailing service like Uber or Lyft, but with vehicles powered by highly advanced AI brains instead of normal human ones. In order to be competitive, Waymo will need to take into account what people are willing to pay for their transportation. Thanks to intense competition between Uber and Lyft, the answer to that question these days is probably not much. While not as inexpensive as it was in the early days, ride-hailing is still incredibly cheap.

The ride that I took from downtown Chandler to the coffee shop roughly three miles away took around eight minutes and cost a little more than $7, according to the app. That’s about what I would pay using Uber or Lyft.

Waymo says it is continuing to experiment with pricing, but it will be using fares that are based on the time and distance of the trip’s route. As with ride-share apps, prices will vary based on demand, which has essentially become the industry standard, Chu said. “Thinking about the distance that you’re traveling, the time it takes, those will definitely be part of [pricing],” he said.

Once the trained drivers are gone and Waymo figures out how to charge for in-ride entertainment and advertising, analysts expect base fares to drop. Of course, no one knows when that will be. But despite these uncertainties, being the first autonomous taxi has its advantages on Wall Street. Last August, a team of Morgan Stanley analysts concluded the company’s ride-hailing service is worth about $80 billion — even before the service launches. Trucking and technology licensing add another $96 billion in current value, the firm concluded.

“We see this business as being somewhat similar to search in its early days when Google paid other publishers to use its search technology,” a Morgan Stanley analyst wrote in his research note. “We believe Waymo is in discussions with ‘more than 50%’ of the global auto industry by volume as it relates to the personal vehicle opportunity.”


“It’s like riding in a ride at Disneyland,” said Diego Vera, a Waymo Early Rider.
Photo by Vjeran Pavic / The Verge

Early Riders

One question for the success of Waymo One is whether Early Riders will be willing to start forking over money for the service. Diego Vera, who has been using Waymo to commute from his home in Chandler to his job in Gilbert as an account manager at GoDaddy.com since July 2017, said yes. “As long as I think the pricing stays comparable to other ride sharing services, I would absolutely take it,” he said in an interview set up by Waymo. (Early Riders are required to sign non-disclosure agreements with Waymo.)

Vera said he’s noticed a lot of changes in the quality of Waymo’s driving since he first started riding. The cars used to be a lot more defensive, but now are more aggressive and drive more similar to a human, like running yellow lights or speeding up in certain situations. “So those kind of behaviors I think, once they start normalizing, you’re not going to be able to tell the difference really between a machine driving versus a person driving,” Vera said.

This matches up with the public comments of other Early Riders, some of whom have recalled early moments of confusion that later were smoothed out after providing feedback. Alex Hoffman, a cybersecurity engineer from Mesa, recalled three separate times when the Waymo car would get “confused” by a large bush near his office. “I gave feedback each time, and by the fourth time, the car went around it with no problem,” Hoffman said in an interview published by Waymo.

Waymo subjects its Early Riders to “heavy engagement,” Chu told me, which includes surveys, interviews, panels, and even ride-alongs with Waymo researchers. Still, Vera said he’s eager to use the vehicles as a Waymo One customer. He’s already paid for a few rides, and he prefers his trips with Waymo to other ride-hailing services. He’s also interested in riding in the fully driverless vehicles without trained drivers behind the wheel, which he has done only once before. Without that safety net to fall back on, he said he feels more present during the trip.

“Right now, riding with a safety driver, you kind of don’t pay attention much, you just kind of do your own thing,” he said. “I saw myself without a driver not really paying attention to my phone at all, being fully present, and kind of paying attention to what the car was doing, watching the pedals move on their own and the steering wheel on its own. So that caught me a little bit off guard, and then just really seeing people’s reactions around, that was kind of cool.”

Vera added, “It’s like riding in a ride at Disneyland.”



What’s next?

For the undisputed leader in autonomous technology, Waymo One’s launch is tiny: exclusivity to a few hundred riders, operating in just a handful of suburban towns, trained drivers in the driver seat. But it’s on this modest perch that Waymo hopes to build a global business.

The company plans to buy up to 62,000 plug-in hybrid Chrysler Pacifica minivans and 20,000 all-electric Jaguar I-Pace SUVs to build its fleet over the next few years. Waymo was recently approved for a permit to operate fully driverless vehicles in Silicon Valley, and recently, it held its first town hall meeting with residents. (The meeting was contentious, according to Palo Alto Online. One sample reaction: “We’re going to storm City Hall if these cars come to Palo Alto.”)

Congress is on the cusp of passing a bill that could authorize hundreds of thousands of self-driving cars without traditional controls like steering wheels and pedals to be deployed as taxis or sold as personal vehicles. This could benefit Waymo’s plans to expand its territory across the US, but it would also give a similar advantage to its competitors like Uber, GM’s Cruise, and Ford.

So Waymo will need to move fast — but not too fast. One of its biggest self-driving competitors, Uber, killed a 49-year-old pedestrian who was crossing the street in Tempe, Arizona, in March 2018. That self-driving car’s automatic emergency brakes were disabled; police later said the safety driver at the wheel was streaming The Voice on her phone at the time of the crash. Uber has since shut down its autonomous vehicle testing program in Arizona, laid off a majority of its testers, and is now engaged in delicate negotiations with Pennsylvania officials to restart its public testing.

Ride-hailing companies have earned ill-will in other ways, too. After marketing campaigns that suggested Uber and Lyft would decrease car use, recent studies have found they poach riders from public transportation. That means ride-sharing is actually increasing traffic congestion.

Congestion is certainly a major problem for many US cities, including Phoenix. It’s the 25th most congested city in the US, according to INRIX. But the real danger is for the people walking around those cars. Arizona has the highest pedestrian death rate in the nation. Between 2014 and 2017, fatal collisions killed 271 pedestrians in Phoenix. Almost three dozen people have been killed while walking in 2018 so far. Sean Sweat, executive director of advocacy group the Urban Phoenix Project, said some local elected officials use self-driving cars as an excuse to oppose common-sense street improvements.

“They say that we don’t need transit because once driverless cars are here, everyone will just be in a driverless car,” Sweat told me as we stood in one of downtown Phoenix’s beautiful sculpture-strewn plazas. “I don’t understand how they don’t understand that just because a car doesn’t have a driver doesn’t mean that you suddenly have all this extra capacity because a car is still a car. The geometry doesn’t change.”

Waymo CEO John Krafcik has said “safety is urgent,” and that the company’s motivation is to eliminate the 1.2 million people killed on the road globally each year. Chu adds, “We want to be a part of that solution as soon as we possibly can. But on the other hand, we also know that for such a new technology rolling it out incrementally, being very methodical and careful about it is the job at hand.”

Back in Chandler, I step out of Waymo’s minivan for the last time and watch as it slowly drives away into the gathering dusk. There will soon be thousands more of these vehicles on the road, not just here in Arizona, but also California, Michigan, Georgia, Washington, and Texas — all states where Waymo has been testing its self-driving cars.

They’re coming. There’s no going back now.

Correction December 6th, 3:30PM ET: A previous version of this story misstated Diego Vera’s name as Diego Rivera. We regret the error.

David Cartu Custom Site Search

Leave a Reply

Your email address will not be published. Required fields are marked *