Why some robotaxi companies are looking for their Goldilocks cities

This article is part of our series on smart cities—click here to visit the interactive hub and explore other stories.

One sunny day in 2016, Chris Ludwick and his coworkers spent eight hours driving around the Phoenix, Arizona, metro area. With Taco Bell in hand, they talked through traffic scenarios and vehicle maneuvers, taking notes on construction zones, roundabouts, and road laws.

They worked for Waymo, and after weeks of considering and visiting potential places to establish the world’s first commercial robotaxi program, the company was ready to make its decision: Phoenix was a go.

Waymo gradually pushed the autonomous ride-hailing service forward over the next three years, beginning with an early rider program, then partnering with Lyft to scale up, and finally, in October 2020, expanding to the general public, with anyone able to book a robotaxi ride via a mobile app.

Once upon a time, many US cities were predicting they’d have robotaxi networks just like this, but two years after Waymo One’s public release, it remains one of just two fully driverless services with commercial operations in the US. Part of the challenge comes with the territory—literally—as robotaxi companies struggle to find the right environments, not only regarding compliance with local laws but also due to the intrinsic properties of cities, from climate to population density to road infrastructure.

After a few years of setbacks, some AV companies are becoming more intentional about how they select cities to partner with, according to Philipp Kampshoff, a senior partner with McKinsey who co-founded the firm’s Center for Future Mobility.

“We’ve seen some of the shared mobility players…sort of take a step backward and say, ‘Look, why don’t we concentrate more on the cities that are already in the money and profitable?’” he said. “I think the robotaxi players have learned that lesson…They will focus on, ‘Let’s unlock a city like SF, then maybe a city or two after that’—go bit by bit and, rather, concentrate on scaling up an individual city rather than trying to be in 15 cities at the same point in time.”

Slow and steady

Waymo’s initial goal was to find a city that made it semi-challenging, but not impossible, to roll out a commercial service. And for those purposes, Phoenix had “the perfect sweet spot of difficulty”: medium-high speed limits across many roads; a friendly regulatory environment; a fast-growing population; challenging maneuvers like unprotected left turns across three lanes of traffic; neighborhoods with varying population density; and a climate that allows for shorts year-round (i.e., no snow or ice on the roads).

Earlier this year, the company started a pilot program in downtown Phoenix, expanding its service beyond the suburbs of Chandler and Tempe. The downtown service offers “‘rider-only’ (with no one behind the wheel) trips” to both Waymo employees and “Trusted Testers,” Katherine Barna, head of PR at Waymo, told us in an email.

This year, Waymo also began testing its fully driverless robotaxis in a second city: San Francisco, near its headquarters in Mountain View.

“We’re currently offering autonomous rides to Trusted Testers across the majority of the city of SF, 24/7, and fully autonomous rides to Waymo employees,” Barna said in the statement.

The city provides a markedly different training ground for its vehicles compared to Phoenix, with different weather, geography, and more population density and pedestrians. The company announced Los Angeles as its third city in mid-October.

Kampshoff expects the robotaxi space to grow gradually, taking a similar city-by-city approach over the next five years.

But what makes sense for a city?

The viability of robotaxis depends on a municipality’s transportation goals, according to Alex Engel, senior communications manager for the National Association of City Transportation Officials.

But, he said, the integration of AV tech becomes “a lot more complicated when you think about city streets” and their complex environments.

“One thing that we’ve seen is that there is a possibility of making our transportation systems a bit more efficient, potentially, with AVs,” Engel said. “That said…some of the simulations and some of the other ways we’ve seen it, sometimes it comes at the cost of other ways we have of moving around that are even more…attainable.” Think: walking, using a scooter, biking, dedicated transit lanes, and more.

Kampshoff said that some of the cities McKinsey has worked with tend to dislike the possibility of empty robotaxis clogging up roadways, so as a result, there may eventually be a “zombie tax” for empty robotaxis on a per-mile basis, Kampshoff added, or even a carpool-based ride-hailing service akin to Lyft Line or UberPool.

Kampshoff also echoed the idea that complexity can make testing difficult but pointed out that it can also provide a strong opportunity for companies—making high-density cities the best case for profitability but also the most complicated case for safety.

“Imagine [a] robotaxi in downtown Manhattan with all the traffic that’s going on, all the pedestrians…this is probably, in the US, the hardest operational design domain you can go after,” Kampshoff said. “At the same time, once you unlock it, it’s also probably one of the most profitable ones. So there’s…sort of this balancing between how quickly can you get to market because you solve the technology, but also how attractive is the market.”

But whatever cities decide with respect to adopting AVs, Engel said the tech isn’t a silver bullet.

“Something that will not improve our transportation outcomes, if you’re looking at things like sustainability…equity, [and] access…is just replacing one-to-one all the vehicles we have now with AVs,” Engel said.

This complex balancing act—weighing company goals and technical needs versus a city’s goals as well as its physical, regulatory, and social environment—may help explain why robotaxi companies are beginning to find and focus on goldilocks cities, per Kampshoff.

And ultimately, that’s the approach Waymo took five years ago, when it decided to stake out its future in Arizona and, per Ludwick.

“Some of the things that make Phoenix hard: It’s got fast roads, so 45 mph boulevards are very common in Phoenix. Almost every major piece of the grid [has speed limits of] 35, 40, or 45 mph, [and] often are three lanes in each direction,” Ludwick said.

“But it didn’t have some of the complexities of older cities in the US, like the density, the number of pedestrians and cyclists,” he added. “It has new infrastructure. And so this allowed us to basically take a big bite [of] the problem and solve some of the hardest pieces, but also not have to solve everything in order to get to a commercial service.”

Update: This piece was updated 11/3 to note that Waymo has announced a third city for robotaxi testing.